<%BANNER%>

Characteristics of Phonological Processing, Reading, Oral Language, and Auditory Processing Skills of Children with Mild...

Permanent Link: http://ufdc.ufl.edu/UFE0021918/00001

Material Information

Title: Characteristics of Phonological Processing, Reading, Oral Language, and Auditory Processing Skills of Children with Mild-to-Moderate Sensorineural Hearing Loss
Physical Description: 1 online resource (221 p.)
Language: english
Creator: Park, Jungjun
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: auditory, awareness, binaural, congenital, ctopp, disorder, dyslexia, evt, extraphonological, gort, hearing, literacy, loss, memory, mild, moderate, morphosyntactic, naming, peripheral, phonological, ppvt, processing, rapid, reading, sensorineural, short, specific, spelling, vocabulary, wrat, wrmt
Communication Sciences and Disorders -- Dissertations, Academic -- UF
Genre: Communication Sciences and Disorders thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The principle aim of the study was to examine the effect of mild to moderate SNHL on 19 children's performances on a range of spoken language, phonological processing skills, and literacy (reading and spelling). Performances of two controls groups (29 normally developing and 30 dyslexic children), matched for age, grade, and non-verbal intelligence were also analyzed to investigate the HI group's relative weaknesses and strengths in phonological processing ability and its relation to their literacy skills. It is well known that a lack of natural and complete linguistic input in early childhood contributes to significant language delays even in children with mild to moderate hearing loss. Although they may be helped by hearing aids, these children usually remain unable to extract enough auditory information to develop spoken language with the ease and efficiency of a normally hearing child. Due to this, the development of appropriate phonological processing skills which are necessarily required for the early reading development is a considerable challenge for children with hearing loss. This study was planned with an assumption that literacy development would require the same acquisition of phonological processing skills, whether a child's hearing is impaired or normal. A set of MANCOVAs revealed that HI group's performances on the measures of spoken language (receptive and expressive vocabulary and morphosyntactic knowledge) and phonological processing tasks were significantly delayed when compared to their normal controls. The HI group?s phonological processing skills were selectively depressed such that only phonologically-based tasks (phonological awareness and phonological short-term memory) were seen significantly lower than the normal controls but their rapid naming skills were seen preserved despite hearing loss. This finding suggests that auditory perceptual distortion due to SNHL can only affect phonologically-based component and would not have negative impact on extraphonological processing ability (RAN). It was found that the HI group's literacy skills were significantly lower than their normally hearing controls. The necessary role of phonological processing skills was also confirmed by the correlational and regression analyses which suggest that phonological processing skills are important correlates and predictors of hearing-impaired children's reading and spelling as well. The strongest predictor of reading and spelling was the Elision subtest on the CTOPP. These findings contradict previous studies which argued that reading skills of hearing-impaired would not necessarily require the support from phonological processing ability. Impressively, RAN measures were not associated with any of untimed reading tasks, but had significant correlations only with timed reading tasks. RAN also additively contributed independent variance to timed reading measures after phonologically-based variables were controlled for. Lastly, assuming that hard-of-hearing individuals, like hearing individuals, will also benefit from direct phonological instruction, the findings imply that effective programs and strategies for teaching hearing-impaired children these skills may be the key to obtaining higher levels of reading achievement for this population.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jungjun Park.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Lombardino, Linda J.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021918:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021918/00001

Material Information

Title: Characteristics of Phonological Processing, Reading, Oral Language, and Auditory Processing Skills of Children with Mild-to-Moderate Sensorineural Hearing Loss
Physical Description: 1 online resource (221 p.)
Language: english
Creator: Park, Jungjun
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: auditory, awareness, binaural, congenital, ctopp, disorder, dyslexia, evt, extraphonological, gort, hearing, literacy, loss, memory, mild, moderate, morphosyntactic, naming, peripheral, phonological, ppvt, processing, rapid, reading, sensorineural, short, specific, spelling, vocabulary, wrat, wrmt
Communication Sciences and Disorders -- Dissertations, Academic -- UF
Genre: Communication Sciences and Disorders thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The principle aim of the study was to examine the effect of mild to moderate SNHL on 19 children's performances on a range of spoken language, phonological processing skills, and literacy (reading and spelling). Performances of two controls groups (29 normally developing and 30 dyslexic children), matched for age, grade, and non-verbal intelligence were also analyzed to investigate the HI group's relative weaknesses and strengths in phonological processing ability and its relation to their literacy skills. It is well known that a lack of natural and complete linguistic input in early childhood contributes to significant language delays even in children with mild to moderate hearing loss. Although they may be helped by hearing aids, these children usually remain unable to extract enough auditory information to develop spoken language with the ease and efficiency of a normally hearing child. Due to this, the development of appropriate phonological processing skills which are necessarily required for the early reading development is a considerable challenge for children with hearing loss. This study was planned with an assumption that literacy development would require the same acquisition of phonological processing skills, whether a child's hearing is impaired or normal. A set of MANCOVAs revealed that HI group's performances on the measures of spoken language (receptive and expressive vocabulary and morphosyntactic knowledge) and phonological processing tasks were significantly delayed when compared to their normal controls. The HI group?s phonological processing skills were selectively depressed such that only phonologically-based tasks (phonological awareness and phonological short-term memory) were seen significantly lower than the normal controls but their rapid naming skills were seen preserved despite hearing loss. This finding suggests that auditory perceptual distortion due to SNHL can only affect phonologically-based component and would not have negative impact on extraphonological processing ability (RAN). It was found that the HI group's literacy skills were significantly lower than their normally hearing controls. The necessary role of phonological processing skills was also confirmed by the correlational and regression analyses which suggest that phonological processing skills are important correlates and predictors of hearing-impaired children's reading and spelling as well. The strongest predictor of reading and spelling was the Elision subtest on the CTOPP. These findings contradict previous studies which argued that reading skills of hearing-impaired would not necessarily require the support from phonological processing ability. Impressively, RAN measures were not associated with any of untimed reading tasks, but had significant correlations only with timed reading tasks. RAN also additively contributed independent variance to timed reading measures after phonologically-based variables were controlled for. Lastly, assuming that hard-of-hearing individuals, like hearing individuals, will also benefit from direct phonological instruction, the findings imply that effective programs and strategies for teaching hearing-impaired children these skills may be the key to obtaining higher levels of reading achievement for this population.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jungjun Park.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Lombardino, Linda J.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021918:00001


This item has the following downloads:


Full Text





CHARACTERISTICS OF PHONOLOGICAL PROCESSING, READING, ORAL
LANGUAGE, AND AUDITORY PROCESSING SKILLS OF CHILDREN WITH MILD-TO-
MODERATE SENSORINEURAL HEARING LOSS




















By

JUNGJUN PARK


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2008

































2008 Jungjun Park



























To my Godly wife, Mina;
and to a bright future for my precious little princess, Hayoung.









ACKNOWLEDGMENTS

This research, from start to finish, spanned over approximately three years. First of all, I

would like to acknowledge my wife (Mina) and daughter (Hayoung), precious family and friends,

for their tireless support. Especially, Mina was more a part of this research than she realizes.

I also acknowledge my four parents. I can never thank them enough for their support,

encouragement, prayers, and deep faith in me. I cannot imagine how I could have finished this

challenging task without their love.

I would especially like to thank my Chair, Dr. Linda Lombardino for her countless hours

of brainstorming, editing, encouragement, and professional advising she offered so patiently. I

extend sincere gratitude to Drs. Bonnie Johnson, Alice Holmes, and Ratree Wayland, my

committee members, for their professional support and guidance.

I also would like to acknowledge several people who contributed a great deal of love,

support, and encouragement. The faculty and staff of the University of Florida's Department of

Communication Sciences and Disorders provided remarkable education, financial support, and

training that were necessary for conceptualizing and completing this research project.

I would especially like to acknowledge my department chair, Dr. Chris Sapienza, for her

remarkable patience with me during this unexpectedly lengthy process and for her incessant

encouragement. I extend special thanks to outstanding teachers that I met during my studies in

Gainesville, including Drs. William Brown, Howard Rothman, Kenneth Logan, Scott Griffiths,

Kenneth Gerhardt, Lori Altmann, Rahul Shrivastav, and Mini Shrivastav.

I thank Drs. Shannon Brumfield and Judith Wingate, my clinical supervisors, for providing

such a wonderful opportunity for me to finish the practicum hours. In addition, I am also

thankful for the friendship and support of my colleagues in the Reading Research Laboratory at

the University of Florida: Sue Ann Eidson, Rebbecca Wiseheart, Cynthia Puranik, Heeyoung









Park, and Sunjung Kim. They have made the experience bearable, and often times even

enjoyable. I am also indebted to many of the staff members, Ms. Idella King, Mrs. Debbie

Butler, Ms. Cassie Mobley, Mrs. Addie Pons, Mr. David Fleming, and Mr. Neal Musson. I also

thank three research assistants, Heeyoung Park, Nicole Leshin and Taprini Spence, for their

valuable assistance in completing a wide variety of tasks associated with the research. Thank

y'all!

Financial assistance for my doctoral study and this research were graciously provided by

the Department of Communication Sciences and Disorders at the University of Florida, Doorae

Academic Foundation, and the Dissertation Research Grant by the College of Arts and Sciences

of the University of Florida. I also would like to express my sincere appreciation to the children

and parents involved in this research for giving their time and energy so that others may benefit.

I thank Ms. Lori Lazarus, a fascinating speech pathologist serving a lot of hard-of-hearing

children, for her initial supports for this study. Also, I want to thank Drs. Melissa Riess and

Ronald Kelley for their great devotions to data collection in the areas of audiology. I also

acknowledge the staff members of the Exception Education Departments whom I worked with

for data collection. This degree would not have been possible without the helps of all of them.

The following space has been saved for me to thank a special group of friends that God has

brought into my life: my brothers and sisters at North Central Baptist Church in Gainesville. I

thank them for the special love and prayers for my family and this uneasy project. Especially, I

extend my deepest thanks to Mr. Jerre Brannen, Mrs. Pat Brannen, and my pastor Calvin Carr for

their continuous words of wisdom, caring heart, and prayers. Last and yet greatest, I thank Jesus

Christ for His amazing mercy.









TABLE OF CONTENTS



A C K N O W L E D G M E N T S ..............................................................................................................4

L IS T O F T A B L E S ................................................................................................. ..................... 9

LIST OF FIGURES ........................................................ ............................ 11

A B S T R A C T .......................................................................................................... ..................... 12

CHAPTER

1 INTRODUCTION .................................... ........... ..................................... 14

B ack g rou n d of th e Stu dy ..................................................... .. ............................................ 14
R ationale and Significance of the Study............................................................ ................ 16
Study O objectives ......................................................................... .... ............. .................. 18
B rief D definition of T erm s .............. ..................................................................... 19

2 L ITE R A TU R E R E V IE W .............. ..................................................................... 2 1

Introduction ............................................................................ ..................... 21
Phonological C ore D deficit Theory .......................................... ........................ ................ 21
Com ponents of Phonological Processing .......................................................... ................ 23
Phonological A w areness ..................... ................................................................. 24
D im tensions of phonological aw areness .............................................. ................ 25
Factors related to order of difficulty ................................................... ................ 26
Phonological M emory ............... .. .............................. ........ .... ............... 27
R a p id N a m in g ..................... .... .. ............................ ..................................................2 8
Cognitive M odel of Phonological Processing .................................................. 29
T heories on Im paired Phonology................ ........................................................ ................ 31
Phonological R presentation H ypothesis................................................... ................ 32
L exical restructuring theory .............. .................................................. ................ 33
A application of representation hypothesis............................................ ................ 35
A challenging observation ........................................ ........................ ................ 36
A critical summary ........................... .......... ........................ 38
D ouble-D deficit H ypothesis ......................................................................... ................ 39
RA N as a phonological processing skill ............................................. ................ 39
T h re e su b ty p e s ..........................................................................................................4 0
D ifferentiation from phonology .......................................................... ................ 41
Linkage between RAN and reading process ....................................... ................ 43
A uditory Perceptual D deficit H ypothesis..................................................... ................ 45
H hearing loss, Phonology, and Literacy ....................................................... ....... ................ 46
Phonological Processing, Language Skills, and Hearing Loss...................................47
O titis m edia w ith effusion (O M E) ...................................................... ................ 48
Sensorineural hearing loss........................................ ........................ ................ 52


6









H hearing L oss and L iteracy. ................................................................... ................ 54
C onductive hearing loss .......................................... ......................... ................ 54
Sensorineural hearing loss and literacy ............................................... ................ 57
Impaired phonology in children with hearing loss.............................. ................ 58
Su m m ary ............................................................................... .... ............. .................. 59
C onductive hearing loss .......................................... ......................... ................ 59
Permanent SNHL ................................. ................60

3 M ETH OD S AN D M A TERIAL S .......................................... ......................... ................ 66

Introduction ................................................ ............... .................. 66
Setting and P participants ................................................... ............................................... 67
R ecruitm ent S ettin g ......................................................................................................... 6 7
S e le ctio n C rite ria .............................................................................................................6 8
P artic ip an ts ........................................ ...................................................... ..................... 6 9
Participants with normal hearing and reading skills (NH group)..........................69
Participants with hearing impairment (HI group) ............................... ................ 70
D yslexic group (R D group).................................... ....................... ................ 71
M watch in g V ariab les ............... ............................... ..............................................7 3
P ro c e d u re ..................................................................................................... ..................... 7 3
M a te ria ls ....................................................................................................... ..................... 7 4
A u d io lo g ic M easu res .......................................................................................................7 4
Puretone and speech audiom etry ...........................................................................75
Auditory processing tests ........................................76
M iddle ear function (tympanom etry)....................................................................77
Literacy (Reading and Spelling) ................................................................................78
Phonological Processing Skills .....................................................................................81
Standardized Oral Language Tests...............................................................................83
In terrater R eliab ility ................................................................................................................8 5
Research Questions and H ypotheses ...................................................................................86
Category I (Group Effect) ...........................................86
Category II (Relationships among Measures).................. ...................................86
Category III (Regression Question)..............................................................................87
Treatm ent of the D ata ...................................................................................................... 87

4 R E S U L T S ..................................................................................................... ..................... 9 3

D e scrip tiv e S statistic s ..............................................................................................................9 4
D em graphic D ata ..................................................................................................... 94
A udiologic A ability M measures ..................................................................... ...............94
Oral Language, Phonology, and Reading..................... ...................................95
In feren tial Statistics .............. .... ............................ ..... ..... .............................. 9 5
Question Category I: Group Comparisons of Language, Phonology, and Reading........95
MANCOVA for oral language measures ............................................... ...............97
MANCOVA for phonology ................. .. ............................. ............... 99
MANCOVA for literacy measures (reading and spelling) ..................................... 102
Question Category II: Correlations among Measures .......................... 105


7









Question 1: Phonology and reading ....... ... .......................... 105
Question 2: Phonology and hearing loss...... ........... ...................................... 108
Question 3: Phonology and oral language............... ........................ ............... 109
Question 4: Auditory processing, phonology, and reading measures................. 109
Question 5: Reading and oral language.........................................110
Question Category III: Exploratory Hierarchical Regressions ...................................111
Regressions on word-level reading ....... ... ...... .....................1... 12
Regression on passage reading rate and accuracy...... ................. .................. 115
R egression on spelling ........ ............. ................................. ............... 116
Role of rapid naming in further regression analyses................... ...................117
S u m m a ry : ...............................................................................................................1 1 8

5 D IS C U S S IO N ....................................................................................................................... 1 5 5

Introduction ............................................... ................. .....................155
O ral Language Skills Findings ................... .............................................................. 155
Group Comparison ........ .... .. ............ ............... 156
Interrelationships between Phonological Processing and Language Skills.................158
Phonological short-term memory and vocabulary ...................... ...................158
Grammatical knowledge and phonological processing................ .................. 160
Phonological Processing Findings................................................................... ............... 162
Hearing Loss and Phonological Representation......................................165
Phonological Representation and Short-term Memory Skill...................................166
Intact Rapid Naming .................... ............. ............................... 167
L ite ra c y F in d in g s ..................................................................................................................1 6 9
Group Comparisons ................... ............ ............................ .. ........... 169
Correlation and Regression Findings ....... ....... ........ ...................... 172
S u m m ary ..................................................................................................... ..................... 17 6
Clinical Implications .............................. ........... .............................. 179
L im stations .................................................................................................... 180

APPENDIX

A INFORMED CONSENT LETTER FOR PARENTS AND GUARDIANS ........................ 182

B Q U E ST IO N N A IR E F O R M .................................................................................................. 186

C SC O R IN G SH E E T S ....................................................... ................................................ 193

L IST O F R E F E R E N C E S ....................................................... ................................................ 196

B IO G R A PH IC A L SK E T C H .................................................... ............................................. 221








8









LIST OF TABLES


Table page

3-1 Matching variables (Grade, Age, Gender, and non-verbal intelligence). ........................89

3 -2 L ist o f te sts u sed ................................................................................................................ 9 0

3-3 L ist of research hypotheses .............................................................................. ................ 9 1

4-1 Background information and basic auditory skills of individuals in the HI group..........119

4-2 Descriptive statistics for the NH and HI groups on PTA, SRT, and WRS scores........ 120

4-3 Mean pure-tone thresholds for all tested frequencies (NH and HI groups only)............121

4-4 Descriptive statistics of auditory processing variables. ....................... ...................124

4-5 Descriptive statistics and estimated adjusted means of oral language measures
(M A N C O V A ). ................................................................................................................ 12 5

4-6 ANOVA tables for univariate ANCOVAs for each oral language measure. .................126

4-7 Pairwise comparison results based on adjusted group means................ ...................126

4-8 Descriptive statistics and estimated adjusted groups means for phonological
processing m easures............. .. ..................... ................ ............ .... ...... ... .......... 128

4-9 Summary of three univariate ANCOVAs on the phonological measures from the
C T O P P .................................................................................................. .................... 12 9

4-10 Summary of post hoc pairwise comparisons of phonological measures (CTOPP
subtests) ............................................................................................... .. 130

4-11 Descriptive statistics and estimated adjusted group means for literacy measures.........134

4-12 Summary of univariate ANCOVAs on the reading and spelling measures.................. 139

4-13 Summary of post hoc pairwise comparisons of eight reading and spelling measures.....140

4-14 Correlations between the degree of hearing loss and phonological measures,
partialing out for age and grade. ................. .......................................................... 141

4-15 Partial correlation matrix between phonology and reading (HI group)........................ 141

4-16 Partial correlation matrix between phonology and reading (NH group). .......................142

4-17 Correlation matrix between better ear's PTA and phonology, partialing out age (in
m o n th s). ......................................................................................................... .......... 14 3









4-18 First-order correlation matrix between oral language and phonology measures.......... 144

4-19 Correlation matrices among auditory processing measures, and phonology and
reading scores, partialing out for age, grade, duration of hearing aid use, and
nonverbal intelligence .... ............... .. ........... .............. ............... 145

4-20 Correlation matrix between oral language and reading measures ..............................146

4-21 Hierarchical regressions of variables separately predicting untimed word reading
(the Word Identification subtest on the WRMT-R). .................................... 147

4-22 Hierarchical regressions of variables separately predicting untimed nonword reading
(the Word Attack subtest on the WRMT-R). ......... .........................148

4-23 Hierarchical regressions of variables separately predicting timed word reading (the
Word Efficiency on the TOW RE). ........................................................ 149

4-24 Hierarchical regressions of variables separately predicting timed nonword reading
(the Word Decoding on the TOW RE). ...... ........... .......... ...................... 150

4-25 Hierarchical regressions of variables separately predicting timed passage reading
rate on the GORT-4 ........................... ........... ........................... 151

4-26 Hierarchical regressions of variables separately predicting timed passage reading
accuracy on the G O R T-4. .................... ................................................................ 152

4-27 Hierarchical regressions of variables separately predicting scores on the Spelling on
th e W R A T ...................................................................................................... .......... 153

4-28 Summary of hierarchical regressions on timed reading measures, controlled for
phonological awareness and short-term memory scores (showing the results from
S te p 3 o n ly ) .................................................................................................................. .... 1 5 4









LIST OF FIGURES


Figure page

2-1 Three different areas of phonological processing......................................... ................ 61

2-2 Phonological loop. .................................. ........... ........................... 61

2-3 W ord recognition through phonological route.............................................. ................ 62

2-4 Model of (underspecified) phonological representation...............................................63

2-5 Two different aspects of phonological representations ................................ ................ 64

2-6 Sim plified m odel of visual nam ing...................................... ...................... ................ 65

4-1 M ean PTA thresholds for the left ear...... .......... .......... ..................... 122

4-2 Mean PTA thresholds for the Right ear. .................. ......................122

4-3 Histogram of 19 hearing impaired children by better ear pure-tone average ................123

4-4 Clustered box plot of oral language measure for the NH and HI groups. ......................127

4-5 Clustered boxplot for phonological awareness measures for the NH, HI and RD
groups ........................................................................................... ........ .. 13 1

4-6 Clustered boxplot for phonological short-term memory for the NH, HI, and RD
g rou p s ...................................................................................... .................... .................. 132

4-7 Clustered boxplot for rapid naming for the NH, HI, and RD groups ............................133

4-8 Clustered boxplot for untimed word-level reading (Word Identification and Word
Attack subtests on the WRMT-R) ..................................................... 135

4-9 Clustered boxplot for untimed word-level reading (Word Efficiency and Word
Decoding subtests on the W RM T-R)...... .......... ......... ..................... 136

4-10 Clustered boxplot for connected-text reading fluency (GORT-4 Rate and Accuracy).. .137

4-11 Clustered boxplot for passage comprehension (WRMT-R) and spelling (WRAT). .......138









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

CHARACTERISTICS OF PHONOLOGICAL PROCESSING, READING, ORAL
LANGUAGE, AND AUDITORY PROCESSING SKILLS OF CHILDREN WITH MILD-TO-
MODERATE SENSORINEURAL HEARING LOSS
By

Jungjun Park

August 2008

Chair: Linda J. Lombardino
Major: Communication Sciences and Disorders

The principle aim of the study was to examine the effect of mild to moderate SNHL on 19

children's performances on a range of spoken language, phonological processing skills, and

literacy (reading and spelling). Performances of two controls groups (29 normally developing

and 30 dyslexic children), matched for age, grade, and non-verbal intelligence were also

analyzed to investigate the HI group's relative weaknesses and strengths in phonological

processing ability and its relation to their literacy skills.

It is well known that a lack of natural and complete linguistic input in early childhood

contributes to significant language delays even in children with mild to moderate hearing loss.

Although they may be helped by hearing aids, these children usually remain unable to extract

enough auditory information to develop spoken language with the ease and efficiency of a

normally hearing child. Due to this, the development of appropriate phonological processing

skills which are necessarily required for the early reading development is a considerable

challenge for children with hearing loss. This study was planned with an assumption that literacy

development would require the same acquisition of phonological processing skills, whether a

child's hearing is impaired or normal.









A set of MANCOVAs revealed that HI group's performances on the measures of spoken

language (receptive and expressive vocabulary and morphosyntactic knowledge) and

phonological processing tasks were significantly delayed when compared to their normal

controls. The HI group's phonological processing skills were selectively depressed such that only

phonologically-based tasks (phonological awareness and phonological short-term memory) were

seen significantly lower than the normal controls but their rapid naming skills were seen

preserved despite hearing loss. This finding suggests that auditory perceptual distortion due to

SNHL can only affect phonologically-based component and would not have negative impact on

extraphonological processing ability (RAN).

It was found that the HI group's literacy skills were significantly lower than their normally

hearing controls. The necessary role of phonological processing skills was also confirmed by the

correlational and regression analyses which suggest that phonological processing skills are

important correlates and predictors of hearing-impaired children's reading and spelling as well.

The strongest predictor of reading and spelling was the Elision subtest on the CTOPP. These

findings contradict previous studies which argued that reading skills of hearing-impaired would

not necessarily require the support from phonological processing ability. Impressively, RAN

measures were not associated with any of untimed reading tasks, but had significant correlations

only with timed reading tasks. RAN also additively contributed independent variance to timed

reading measures after phonologically-based variables were controlled for.

Lastly, assuming that hard-of-hearing individuals, like hearing individuals, will also

benefit from direct phonological instruction, the findings imply that effective programs and

strategies for teaching hearing-impaired children these skills may be the key to obtaining higher

levels of reading achievement for this population.









CHAPTER 1
INTRODUCTION

Background of the Study

The purpose of this study is to investigate the strengths and weaknesses of skills in

reading, phonological processing, and oral language of children with congenital mild to moderate

sensorineural hearing loss in comparison to those of dyslexic and normal readers with intact

hearing ability.

It is widely acknowledged that the critical requirement of written language skills is the

acquisition and development of language and concomitant cognitive skills at as early an age as

possible. Among various tiers of linguistic knowledge, the phonological component is now

known as the most important foundation for the development of early literacy skills. Over the

last several decades, an intense research focus has been placed on relations between phonological

processing capacities and reading performances measured by word recognition, reading

comprehension, reading fluency, and spelling. Evidence from both typically developing and

atypically developing children demonstrates that the quality of a child's phonological

representations is important for their subsequent progress in literacy. This relationship has been

found across all languages studied, for both persons with typical reading ability (e.g. Bradley &

Bryant, 1983; Hoien, Lundberg, Stanovich, and Bjaalidet, 1995; Siok & Fletcher, 2001), and

persons with reading disabilities (e.g. Bradley & Bryant, 1983; Bruck, 1992; Landerl, Wimmer,

& Frith, 1997; Porpodas, 1999). It is generally accepted that dyslexia is characterized by

developmental weaknesses in establishing phonological representations of speech sounds.

The 'phonological core deficit' theory (Stanovich, 1988) argues that children with dyslexia

find it difficult to represent mentally the sound patterns of the words in their language in a

detailed and specific way. Research to date has supported the causal connections between









phonological representation and reading acquisition. Thus, research supports that the presence of

normal phonological processing capacity is a hallmark characteristic of good readers while its

absence is a consistent characteristic of poor readers (Hurford, Darrow, Edwards, Howerton,

Schauf, & Coffey, 1993; Mann, 1993).

Based upon the phonological core deficit hypothesis of literacy development, if children's

phonological representations are adversely affected for some reason, we can hypothesize deficits

in their phonological skills will result in deficiencies in reading skill. Many students with

congenital hearing impairment can have problems with reading and writing because of their

difficulty in acquiring and/or manipulating the structural characteristics of phonological

component of language. It is assumed that hard-of-hearing children have constrained access to

the input of speech sounds of their first language during the early critical period of language

acquisition. Because of the possible distortion of the incoming acoustic signal, it is hypothesized

that the phonological processing capacity of the children with hearing loss and the emerging

literacy-related cognitive skills resulting from it would also be negatively affected (Briscoe,

Bishop, & Norbury, 2001; Nittrouer, 1996; and Nittrouer and Burton, 2005).

According to Nittrouer and Burton (2005), early auditory experience associated with

normal hearing capacity facilitates the development of language-specific perceptual weighting

strategies, which are believed to be critical for accessing phonetic entities of a given language

and constructing a language-specific phonological structure. In turn, this knowledge of a

language's phonemic inventory and phonological structure, based on normal auditory perception,

will allow for the appropriate development of phonological processing skills such as efficient

temporary storage (verbal working memory), lexical retrieval from long-term memory, and a

variety of meta-phonological skills. In this respect, peripheral hearing loss which is assumed to









result in alteration or distortion of auditory input, can provide a way of testing the phonological

processing deficit hypothesis. That is, can the theory of phonological core deficit be applied to

hearing impaired children, for whom possible phonological depression may be associated with a

peripheral auditory deficit?

Rationale and Significance of the Study

Links between hearing loss, language, and reading performance have been widely

investigated in children with conductive hearing loss caused by otitis media with emission

(OME) (Friel-Patti & Finitzo, 1990; Mody, Studdert-Kennedy, & Brady, 1997). Researchers

tried to address the effects of frequent OME on children's speech, language, and reading

abilities. However, findings are inconsistent across studies. Also, there is a large body of

research investigating oral and written language of children with SNHL, but most of this

research focuses on children with severe to profound losses who rely heavily on visual

information. Thus, scant research efforts have been directed to studying the contribution of

phonological processing skills to literacy development in children with mild to moderate levels

of SNHL.

To date, only a few studies have investigated the effect of potentially impaired

phonological skills in children with mild-to-moderate SNHL on reading and related cognitive

skills (Briscoe et al., 2001; Delage & Tuller, 2007; Gibbs, 2004; Halliday & Bishop, 2005).

Furthermore, the results of previous research are contrary to our expectations based on findings

from other populations. That is, hard-of-hearing children with depressed phonological processing

skills do not show the pervasive difficulties with language and literacy. For example, Briscoe et

al. (2001)'s study shows that even with significantly depressed phonological skills associated

with hearing loss, reading skills are possible to some degree. In Briscoe et al.' study, nineteen

children with mild and moderate SNHL were merged together to comprise a single experimental









group and their reading and phonology skills were compared with those of children with specific

language impairment (SLI). However, among the total 19 children in this group, 13 children had

mild hearing losses and only three children had moderate SNHL. This skewed distribution of

hearing loss levels would not adequately represent the population of children with mild-to-

moderate SNHL. As expected, this might be the reason the subjects' reading performance was

comparable to that of normally hearing children.

Seemingly, due to this limitation, statistical analyses conducted for group comparisons

may not have captured the potential adverse effects of phonological depression associated with

hearing impairment. Gibbs (2004) also reported a similar pattern. Fifteen children with mild-to-

moderate bilateral SNHL were compared to normally hearing controls on reading performance.

The reading abilities of the children with SNHL were indistinguishable from their hearing peers

and reading abilities were not significantly associated with the degree of hearing impairment.

However, their phonological skills were inferior to those of the normally hearing controls even

though they had a significant negative correlation with hearing loss level. Gibbs concluded that

reading might be possible even without well developed phonological skills. Thus, further

research strictly focusing on the literacy and reading-related cognitive skills in children with

mild-to-severe SNHL with appropriate methodology is warranted.

In these studies, the hearing-impaired children's success with reading in spite of depressed

phonological processing skills was notable and unexpected in the theoretical framework of

phonological core deficit. These unexpected findings raise a critical question regarding the

precise role of phonological skills in written language acquisition. That is, according to the

'phonological processing deficit hypothesis', impaired phonology would lead to compromised

literacy ability since phonological capacity is regarded as a necessary, although not a sufficient









condition for reading. Moreover, no previous studies compared hard-of-hearing children's

phonological and literacy skills to the same skills in children who have dyslexia. Such a

comparison could provide insight into the cognitive skills of two groups of children with reading

disabilities due to different etiologies.

In addition, very few studies investigated the effects SNHL has on phonological, language,

and reading skills using a wide range of tests. To provide a more precise view of the

developmental characteristics of hearing impaired children's reading and phonological skills, a

comprehensive battery of tests was used in this study. That includes: (1) oral language (receptive

and expressive vocabulary, grammatical knowledge), (2) phonological processing (phonological

awareness and short-term memory, verbal working memory, rapid naming), (3) reading skills

(word recognition and nonword decoding in both timed and untimed manner, paragraph reading

accuracy/speed, reading comprehension, and spelling). Finally, auditory processing tasks were

also used to determine if hearing impaired children's reading skill is associated with low-level

auditory perceptual processing, a skill needed to perceive speech events for brief temporal

durations.

Study Objectives

* Objective 1: To provide a comprehensive data set on a range of oral language, phonology,
auditory processing, and literacy skills in hard-of-hearing children.

* Objective 2: To investigate the strengths and weaknesses in reading, phonological
processing, and oral language of hearing-impaired children and to compare their abilities to
those of children who are typical readers and children who have dyslexia, a specific
reading impairment.

* Objective 3: To overcome some methodological limitations mentioned earlier by collecting
a set of data which well represent the population and by using a better statistical approach
for analysis.

* Objective 4: To address the universality of phonological core deficit theory. The theory is
that the contributive and necessary role ofphonological processing capacity for literacy
found in normal hearing children will also be observed in children with hearing loss, that









is, the theoretical framework known as the phonological deficit hypothesis predicts that the
more severe the initial hearing loss, the more impaired reading skills will be. This issue is
of great theoretical importance in understanding associations between impaired
phonological processing skills and reading achievement in children with hearing loss.
Clinically, if this hypothesis is supported in this study, the results will serve to guide
principles for reading programs for hearing-impaired children in school settings.

Brief Definition of Terms

Research literature in the area of phonological processing and phonological awareness

entail highly technical language. The following definitions can be used as a guide for the

subsequent discussion of these complex concepts.

* Phonological representation: Use of arbitrary symbols (oral or written) to represent
experience or concepts (e.g., words or graphic symbols like "$").

* Phonological processing: The use of phonology or sounds of language to process verbal
information in oral or written form in short- and long-term memory (Wagner & Torgesen,
1987). Components include awareness and coding (i.e., coding sounds for storage in
memory and retrieval of sounds from memory codes) of verbal information only
(Cornwall, 1992; Hurford et al., 1993; Torgesen et al., 1990; Vellutino & Scanlon, 1987a;
Wagner & Torgesen, 1987).

* Phonological coding: "The representation of information about the sound structure of
verbal stimuli in memory" (Torgesen et al., 1990, p. 236).

* Phonological recoding: Translation from either oral or written representation into a sound-
based system to arrive at the meaning of words in the lexicon (stored vocabulary) in long-
term memory (Wagner & Torgesen, 1987).

* Phonological units: Refers to the size of the sound (e.g., phonemes, onset-rimes, syllables,
word).

* Phonetic recoding: Translation of verbal information into a sound-based system for
temporary storage in working memory for processes such as decoding unfamiliar words in
fluent reading, or during the beginning reading processes of blending and segmenting.

* Phonological awareness: Conscious awareness of the sound segments in language (e.g.,
syllable, shared rimes, or phonemes) and ability to manipulate sound (e.g., move, combine,
and delete).

* Phonemic awareness: Awareness of phonemes, discrete individual sounds that correspond
to individual letters. Spector (1995) pointed out that many terms have been used for this
ability, including phonemic awareness, phonetic analysis, auditory analysis, phonological
reading, phonological processing, and linguistic awareness. Typically, phonological









awareness is used as a general term and phonemic awareness is used to refer specifically to
awareness at the phoneme level.

* Decoding: Translating individual letters and/or groups of letters into sounds to access the
pronunciation of a word.

* Lexical access: Access to internal dictionary in memory.

* Retrieval: Accessing coded information from short-term or long-term memory.

* Phonemes: Individual sounds, smallest unit of sound.

* Grapheme: Written symbols or letters of the alphabet; arbitrary, abstract, and usually
without meaning; the written equivalent of phonemes.

* Grapheme-to-phoneme correspondence: Linkages between discrete phonemes and
individual letters or graphemes.

* Onset-rime: Two-part division of words into units that are smaller than syllables; onset is
the first division of a single phoneme or consonant cluster (e.g., /br/ in bright), rime is the
last division with multiple phonemes (e.g., /ight/ in bright).

* Alphabetic understanding: Understanding that letters represent sounds and that whole
words have a sound structure consisting of individual sounds and patterns of groups of
sounds.









CHAPTER 2
LITERATURE REVIEW

Introduction

The primary goal of this chapter is to provide a review of previous studies related to the

phonological core model of reading disabilities is provided.

The nature of weakened phonological processing and its possible causal role to later

literacy development is considered through a review of current explanatory models or theories of

the interface between reading and phonology. Theories reviewed will include: (i) Phonological

representation hypothesis and lexical restructuring hypothesis; (ii) double-deficit hypothesis, and

(iii) auditory processing deficit hypothesis. Integrated summary of each theory will be provided

in each section. The chapter also reviews studies dealing with language and reading development

of hearing impaired children. Before that, a brief theoretical framework of phonological

processing is given in the first section.

Phonological Core Deficit Theory

Developmental dyslexia (DD) refers to the inability to acquire proficient reading skill and

is a prevalent learning disability affecting between 5 and 15% of children in school. Vellutino

(1979), one of the earliest studies in reading disorder, reported that children with reading

problem have systematic difficulties on tasks with verbal demands, whereas they performed at

the similar level with normal readers on non-verbal tasks.

Over the last two decades, a large body of converging evidence now indicates that dyslexia

stems from an underlying deficit in the phonological processing system, suggesting that deficits

in processing the sounds of language explain a significant proportion of beginning reading

problems and correlated problems with older readers even though debates still remain regarding

whether a single, phonological core deficit or other cognitive deficits lead to reading failure









(Beitchman & Young, 1997; Mody, 2003; Shaywitz, 1998; Snowling, Nation, Moxham,

Gallagher & Frith, 1997) or underspecified, poor phonological representations (Elbro, 1996;

Fowler, 1991; Hansen and Bowey, 1994; Metsala, 1997; Snowling, Goulandris, Bowlby, &

Howell, 1986; Swan and Goswami, 1997a,b).

The strong association between the phonological deficits and dyslexia led Stanovich

(1986) to propose that dyslexia should be defined as a core phonological deficit. In the

phonological core-variable difference model of dyslexia (Stanovich & Siegel, 1994), poor

phonology is related to poor reading performance, irrespective of IQ. An important advantage of

the core phonological deficit definition of dyslexia is that it makes sense in terms of what is

known about the normal acquisition of reading. That is, it has been known that phonological

awareness skills measured in preschool is an excellent predictor of later literacy performance,

even after the substantial effects of IQ are controlled. It is well understood that the ability to

reflect upon the sound structure of words at the phonemic level is critical to the development of

the alphabetic principle that allows children to decode novel words that they have not seen

before.

There is now converging evidence that the core deficit in reading disability is at the level

of phonological awareness, letter-sound decoding and limitations of verbal short-term memory

(Ehri, Nunes, Willows, Schuster, Yaghoub-Zadeh, & Shanahan, 2001; Fletcher, Shaywitz,

Shankweiler, Katz, Liberman, Stuebing, Francis, Fowler, & Shaywitz ,1994; Foorman, Francis,

Beeler, Winikates, & Fletcher, 1997; Morris, Stuebing, Fletcher, Shaywitz, Lyon, Shankweiler,

1998; Scanlon & Vellutino, 1997; Shaywitz et al., 1999; Stanovich, 1988, 1993; Stanovich &

Siegel, 1994; Torgesen, Wagner, & Rashotte, 1997; Vellutino, Fletcher, Snowling, & Scanlon,

2004). Shaywitz (2003) describes phonological awareness as an inclusive term that "...includes









all levels of awareness of the sound structure of words. It also is used to refer to the earliest

stages of developing an awareness of the parts of words, such as sensitivity to rhyme or noticing

larger parts of words such as syllables" (p. 144). Letter-sound decoding is the process of

converting the written symbols on the page to the smallest unit of speech sounds called

phonemes (Shaywitz, 2003). There is also evidence that dyslexic children have trouble with

long-term verbal learning and the retrieval of phonological information from long-term memory.

Word-finding difficulties are often seen clinically and deficiencies in lexical retrieval can be

manifested by rapid naming tasks.

In a recent summary of what has been learned about dyslexia in the past four decades,

Vellutino et al. (2004) reviewed the support behind a number of theories that have been proposed

as the underlying cause of dyslexia. Citing findings from the research literature, Vellutino et al.

found that there is "...growing consensus that the most influential cause of difficulties in

learning to read is the failure to acquire phonological awareness and skill in alphabetic coding"

(p. 12). More specifically, weak phonological coding has been identified as the central cause of

reading disability in most impaired readers (Archer, Gleason, & Vachon, 2003; Ehri et al., 2001;

Lyon, Shaywitz, & Shaywitz, 2003; Ramus, Rosen, Dakin, Day, Castellote, White, & Frith,

2003; Vellutino et al., 2004).

Components of Phonological Processing

Phonological processing is defined as the use of phonological information (i.e., the sounds

of one's language) in processing written and oral language. Phonological processing is defined

as the use of phonological information (i.e., the sounds of one's language) in processing written

and oral language (Wagner & Torgesen, 1987). It sometimes has been used interchangeably with

phonological awareness, but these two concepts are best seen as distinct from one another. Three

major components of phonological processing deficits have been identified: (a) phonological









awareness, (b) phonological recoding in lexical access, and/or (c) phonetic recoding to maintain

information in working memory (Wagner & Torgesen, 1987).

More specifically, research with school-age children has identified three interrelated

phonological processing abilities that are important for reading and writing: phonological

awareness, phonological memory, and efficiency of phonological access to lexical storage. How

the various phonological processing abilities are related to each other and what roles they play in

literacy development are issues of considerable theoretical and practical importance.

Phonological awareness (PA) refers to one's ability to detect or manipulate the sounds in

his or her oral language (for review, see Anthony & Francis, 2005). PA encompasses phoneme

awareness, the ability to manipulate individual sounds (phonemes) in words, and rudimentary

phonological awareness skills, such as judging whether two words rhyme. Phonological memory

(PM) refers to the coding of information in a sound-based representation system for temporary

storage. PM is utilized during all cognitive tasks that involve processing sound information.

Individuals' PM capacity is often operationalized by auditory span tasks, like digit span. Rapid

naming (RAN) refers to the efficiency of retrieving phonological codes from memory. Individual

differences in efficiency of retrieving phonologically stored information from memory are

typically measured by performance on rapid autonomic naming tasks in which individuals

verbally identify common objects, letters, or numbers as quickly as possible.

Phonological Awareness

Stanovich viewed phonological awareness, as "conscious access to the phonemic level of

the speech stream and some ability to cognitively manipulate representations at this level" (1986,

p 362). Basically, phonological awareness is a multilevel skill of breaking down words into

smaller units (Hoien, Lundberg, Stanovich, & Bjaalid, 1995). It refers to an individual's

awareness of or sensitivity to the sound structure, or phonological structure, or segments of









differing level in spoken words. Torgesen (1997) defines phonological awareness as the ability to

notice, think about, or manipulate the sounds in language.

Dimensions of phonological awareness

Based on recent phonological theories focusing on the hierarchical sound structure of a

word, phonological awareness can be described in terms of different phonological tiers such as

syllable, onset-rime, and phonemes. Stanovich (1994) viewed phonological awareness as the

ability to deal explicitly and segmentally with sound units ranging from syllable, onset-rime, and

phonemes. Gillon (2004:5-9) subdivided it into three subtypes; syllable awareness, onset-rime

awareness, and phonemic awareness (Figure 2-1).

A variety of measures have been used to assess individual's knowledge of these three

differing levels such as auditory discrimination, blending, segmenting, deletion, isolation,

rhyming, substitution, sound categorization, tapping, reversing order of sounds, and word to

word matching. Especially, regarding phoneme-level of awareness, Adams (1990) describes five

different types of manipulative skill in terms of abilities: (i) To do oddity tasks (comparing and

contrasting the sounds of words for rhyme and alliteration); (ii) to hear rhymes and alliteration as

measured by knowledge of nursery rhymes; (iii) to blend and split (segment) syllables; (iv) to

perform phonemic segmentation (such as counting out the number of phonemes in a word), and

(v) to perform phoneme manipulation tasks (such as adding, deleting a particular phoneme and

regenerating a word from the remainder).

Similarly, research has shown that phonological awareness dimensions can be validly and

reliably measure through a variety of tasks (Yopp, 1988). According to Yopp, the dimensions of

phonological awareness are represented by a range of difficulty. From easiest to hardest the

range of difficulty is as follows: (a) rhyme, (b) auditory discrimination, (c) phoneme blending,









(d) word-to-word matching, (e) sound isolation, (f) phoneme counting, (g) phoneme

segmentation, and (h) phoneme deletion.

Factors related to order of difficulty

A number of properties of phonological units have been found to affect the degree of

difficulty of phonological awareness tasks including: (a) the position of the unit in words (i.e.,

first, middle, or last); (b) degree of abstraction; (c) characteristics of tasks used; and (d) size of

sound unit and related acoustic features.

* Position: Research points to the differential difficulty for initial, medial, and final
positions, with initial and final positions being easier than middle (Byrne and Fielding-
Barnsley, 1989).

* Presence of semantic content: Degree of semantic abstraction also affects difficulty.
Different from phonemes or rhymes, real words are less abstract entity due to their
semantic entity. Thus, words are recognized and manipulated naturally without less
instruction. Instead, phonemes are: (a) the smallest phonological unit, (b) not acoustically
pure (not easily isolated), (c) independent of meaning, and (d) abstract and arbitrary.

* Task characteristics: The characteristics of tasks involved decide difficulty as well. Adams
(1990) indicated that most young children can rhyme but not delete.

* Size of unit: Differential difficulty among units of varying size can be explained by their
acoustic features. Spector (1995) conjectured that we do not hear discrete pure phonemes
because they overlap in speech chain; rather, we hear in syllables. Therefore, tasks that
require sensitivity to phonemes may be more complex and necessarily more difficult than
those that require manipulation of syllables. Similarly, syllable segmentation is easier and
often develops without instruction in contrast to phoneme segmentation. Liberman &
Shankweiler (1985) reported that in groups of four-year-old children, none could segment
by phoneme whereas about 50% could segment by syllables

* Processing memory: Memory capacity also contribute to difficulty related to phonological
awareness since each phonological awareness task requires a certain amount of memory
capacity to temporarily hold material for later processing. For example, phonemic
awareness tasks can be divided into different categories depending on the memory
processes and the number of operations required. For example, when asked what sounds
are heard in fish (segmentation), only one operation of pulling apart sound is needed (i.e.,
/f/, /i/, and /sh/) while holding the word in short-term memory. In contrast, when asked to
delete the first sound from 'fish' (deletion), a child should (1) segment the sounds, (2)
identify the beginning sound, and (3) creating a new chain of phonemes using the
remaining sounds filling the memory slot previously occupied by the omitted sound.









Phonological Memory

Phonological memory refers to recoding visually (letters) or orally (speech) presented

words into a temporary phonological buffer to maintain efficiently for subsequent verbal tasks

such as rote repetition of words/nonwords or conscious manipulation of parts of a word. This

process is called verbal short-term storage or temporary retention of phonological representation.

Baddeley and his colleague figured out human memory model which best fits various cognitive

behaviors. They extended the previous short-term memory, a 'passive storage of verbal input', to

a more active working space (i.e., working memory) serving lots of cognitive tasks such as

phonological awareness, sentence analysis, retrieval of lexical information from a LTM, or

decoding of words.

Old research on short-term memory (STM) has focused on temporary holding of

information, rather than on the processes or transformation of information in general cognition.

However, for tasks which require more than simple retention of information, a different memory

module is needed. For Baddeley and Hitch (1974), the working memory is a 'workspace' and its

capacity can be divided between 'processing' and 'storage'. The storage portion was referred to

as the 'phonemic buffer' or 'phonological loop' and the processing portion was referred to as

'central executive'. Tasks such as verbal rehearsal which extracts the information from the

phonemic buffer, segmenting, blending, reordering, and substitution were all regarded as the

functions attributed to the central executive, a processing system. Figure 2-2 is a simplified

display of Baddeley (1986)'s tripartite framework, which had consolidated the position of the

working memory model in cognitive psychology. The frame consists of two modality-specific,

limited-capacity storage systems called 'slave systems' and a limited-capacity general processing

system called 'central executive'. Repetition of words, nonwords, or sentence and digit span

tasks (i.e., serial recall of digits) are used to measure the efficiency of phonological recoding in









working memory (i.e., they measure phonological short-term memory). Among these, nonword

repetition is seen as the best measure of short term memory since semantic cues would be

available to facilitate rote repetition of nonsense words.

Rapid Naming

A great deal of research over the past 30 years on children's phonological awareness has

brought us virtually everything we need to know about identifying children who lack

phonological awareness and teaching them to develop this knowledge. However, another

capacity was shown significantly compromised in dyslexic children; slow lexical access (rapid

automatized naming, RAN in short). Wagner and Torgesen (1987) used a specific terminology

for this process: phonological recoding in lexical access. It is defined as getting from a written

word to its lexical referent by converting the graphemes into sound-based representation system.

In general, coding process involves translating stimuli from one form to another (e.g., from

letters to auditory). For efficient word recognition, phonological coding ability should be

automatized or fluent.

Since Wagner and Torgesen (1987)'s study, a new line of research have suggested that

automatized lexical access and retrieval may significantly affect ease of reading acquisition.

Rapid naming and lexicality tests are two tasks commonly used to measure ability to fluently

code letters into phonological representations. O'Connor, Jenkins, Leicester, & Slocum (1993)

reported that significant difference on reading and spelling measures between low- and high-

skilled readers could be explained by differences in rapid letter naming. Cornwall (1992)

similarly indicated that students that had rapid rates of letter naming did better in word

identification and passage reading speed and accuracy than those with lower rates of rapid

naming. Wren (2005) indicated that impaired skills in RAN is not easy to identify at younger

age, and improving RAN and visual processing speed is considerably more difficult than helping









children develop PA. To the theorists, it has been an important question whether the

phonological awareness and the impaired RAN represent the so-called double-deficits or whether

they are two manifestations of the same underlying disorder (Tijms, 2004). Unfortunately, less is

known about cognitive processing model of RAN.

First of all, visual stimuli (letter sequence or pictures) should be converted into a sound

representation to get access to permanent lexical item (retrieval). According to Ramus (200 ib),

after visual or auditory signal is processed to form sub-lexical phonological representation, with

which no lexical, semantic information is yet linked, retrieval system will start to find a lexical

target of which the phonological form is well matching with the sub-lexical phonetic entity.

After retrieval is successfully executed, semantic component of the target item will be activated

or pulled out to temporary working memory space and used for subsequent verbal tasks such as

comprehension.

Cognitive Model of Phonological Processing

Figure 2-3 provides is a cognitive model of phonological processing and word recognition,

which is a refined version of Levelt (1989)'s model of speech production and Ramus' (2001b)

model of lexical access. This model integrated orthographicall lexical access' (reading), 'visual

lexical access' (object recognition), and 'auditory lexical access' (speech-based word

recognition) to explain reading.

Each box stands for distinct levels of representations and arrows stand for a specific

conversion, or translation between different modules. Mental lexicon is divided into three parts

(lexical meaning, phonological form, and orthographical form). When children perceive ambient

speech signal, it is first encoded as non-specific manner, that is, as non-speech signal (arrow 1).

In the model, this non-specific signal is called 'acoustic representation'. At a later stage, this

must be encoded as a speech signal (arrow 2). Since no lexical entry is accessed yet, the model









uses a term of 'sub-lexical phonological representation' at this level. This sub-lexical

representation has also to be converted into a phonological representation for a specific lexical

item. So, the arrow 3 between sub-lexical and phonological form represents 'auditory word

recognition'. Auditory word recognition requires the finding of lexical item whose phonological

form matches the sub-lexical sound representation. As indicated by Ramus (2001a:201), the

phonological lexicon is a permanent, long-term storage for word forms (LTM), whereas the sub-

lexical phonological representation is a short-term storage for whatever can be represented in a

phonological manner, that is, words or non-sense words. In this regard, Ramus' model can be

extended into a new framework, in which the loci of the phonological memory (STM/LTM) and

awareness can also be established.

When a word is very familiar to a listener (i.e., cat), a procedure called 'whole-word

recognition' will happen without any rehearsal in working memory area, that is, the child may

not have to manipulate or rehearse the input sequence of sounds to figure out the ultimate target

in lexicon and will instantly map the sub-lexical sound form with the corresponding

phonological form (arrow 3). However, when the incoming sub-lexical phonological

representation is new or unfamiliar (e.g., peruse), the child needs to analyze it into smaller parts

(syllable, intra-syllable, or phonemes) and use this parsing information (e.g. per + use) for

lexical retrieval by manipulating or processing them. Since these manipulation tasks are quite

different from rote repetition usually done in STM, we need working or processing memory, as a

functionally separate module. Working memory is a locus where retention and phonological

manipulation (awareness tasks) occur. The input from the STM in sub-lexical component can be

sent to WM for further manipulation or phonological rehearsal need to process unfamiliar words

or non-words (arrow 4). WM also receives its input from the lexical representation. From the









mental lexicon, phonological form of a lexical item can be sent to WM for further awareness

tasks (arrow 5), but WM is not a place for permanent storage like lexical storage. For

phonological awareness tasks, WM consists of its own 'retention resource' needed for temporary

processing (storage capacity) and 'manipulation function' (processing capacity). In this respect,

WM is displayed as separated from sub-lexical and lexicon. Of importance is the fact that

phonological processing components (sub-lexicon, WM, phonological awareness skill) are the

main gate which incoming auditory signal must enter for successful lexical access and

acquisition.

Theories on Impaired Phonology

One of the robust findings in literacy development research is that children with impaired

reading skill show concurrent weakness in meta-phonological skills (phonological awareness

tasks). Indeed, a child's PA knowledge has been described as the best single predictor of reading

performance (Lieberman, Shankweiler, & Liberman, 1989). Because of this, it was assumed that

weak phonological awareness is a 'causal' factor in reading and spelling difficulties. Logically,

however, a strong predictive power of a relevant factor does not necessarily imply its causal

relation to a resulting condition. Rather, it is still quite unclear in what way aspects of weakened

phonological awareness skills are causallyy' linked to reading impairment.

Of equal significance is the fact that deficits in other two areas of phonological processing

skills have been also identified in children with reading disorder: (1) inefficient activation of

long-term phonological coding in lexicon for rapid naming tasks, and (2) impaired short-term

memory skills necessary for immediate or rote repetition of sequences of words or sentences.

Considering these general impairments in phonological components, research pointed to a more

central phonological module as the possible source of dyslexia (Brady, 1997; Mody, 2003;

Snowling, 2001; Swan & Goswami, 1997a).









Thus, the precise nature of the phonological processing difficulties in reading disorder is a

central topic of current psycholinguistic research on dyslexia (Elbro & Jensen, 2005). Various

hypotheses have been suggested about the nature and origin of phonological processing difficulty

in dyslexia. How these problems relate to each other interrelationn), the extent to which

weakened phonology is a cause of reading difficulty (causality) continue to be explored, but the

exact mechanisms by which dyslexics' weakened phonological processor impact each relevant

skill of oral or written language skills such as impaired decoding or spelling, delayed word

recognition, and reading comprehension is not clearly documented.

In this section, three theoretical approaches will be shortly reviewed regarding the possible

causes of impaired phonological skills in dyslexic population: phonological representation

deficit, lexical underspecification hypothesis, double-deficit hypothesis, and auditory perceptual

deficit theory.

Phonological Representation Hypothesis

One suggestion is that the dyslexic readers' phonological representations of lexical items

may be less well specified than normal children (Katz, 1986; Fowler, 1991; Elbro, 1996; Foy &

Mann, 2001; Griffith & Snowling, 2002). Problems related to establishing 'complete', 'full',

'clear', or 'precise' phonological representations in children's speech-based coding system or

long-term lexical memory have frequently been mentioned as a possible cause of diverse

phonological difficulties of dyslexia.

This hypothesis derives from theories of the development of spoken word recognition and

production. When children first begin to acquire lexical items during infancy, each word is coded

in terms of certain semantic and phonological features. For example, for them, 'Daddy' may

refer to a person of a certain sex/size and it would appear different from 'doggy'. These features









will be more fine-grained, specified, and augmented over time. So, children will soon understand

the difference between 'Daddy' and 'doggy' or 'Debbie'.

Phonological distinctness refers to the magnitude of the difference between a

representation and its neighbors and it is an aspect of the static quality of phonological material.

This hypothesis states that a lack of distinctness and/or segmental specificity in dyslexic

children's developing phonological representations supporting spoken word recognition and

production is causally linked to their impaired phonological processing skills (Goswami, 2000;

Snowling et al., 1986, Elbro, 1996). Thus, it should be distinguished from dynamic phonological

processes such as conscious manipulation, verbal rehearsal, phonological retrieval, and

articulation, all using phonetic segments already represented.

For example, due to the poor phonological encoding skill, only parts of the phonetic

material of the input can be stored in lexicon (e.g. 'sub' for subway or 'croco' for crocodile), or

relatively full but not complete representation can be provided (e.g. 'cro?dile'), where the

question mark indicates that any unspecified segment can be inserted which fits with the

phonotactic rules of English (e.g. crowdile, cropodile). To sum, inaccurate, underspecified,

indistinct, or low-quality phonological representation of incoming sounds may hinder higher

level of phonological processing (Snowling, 2001; Swan & Goswami, 1997b; Tijms, 2004).

Lexical restructuring theory

A few evidence for weakened phonological representation come from a theory of lexical

learning strategy. Metsala and Walley (1998) and other studies suggested a theory of lexical

maturation, 'lexical reconstruction' theory (Fowler, 1991; Walley, 1993). This theory suggests

that children's phonological representation of words is re-presented a number of times in terms

of different aspects, that is, from whole words to syllables, and phonemes.









It is interesting to note that children's word recognition strategies change with the increase

of vocabulary; as vocabulary grows, initial 'holistic representations' are gradually re-structured,

and ultimately, phonemes. In this process, frequent or familiar or early-acquired words will be

encountered many times and so will experience more re-structuring than less frequent words.

Also, children should have more experience for phonologically ambiguous words, that is, words

with many similar-sounding neighbors than words with few phonological neighbors.

Another important aspect of this theory with regard to phonological under-specification is

that the phoneme would not be an integral aspect until certain period of language growth (Eimas,

1974), but rather it emerges as a representation unit via continuing spoken language experience

as children experience further lexical restructuring processes. So, the degree to which segmental

(i.e., phoneme-mediated) representation has taken place will be in turn thought to determine

children's abilities in phonological awareness, which is essential for phoneme-based decoding

process and resulting word-level reading. It is assumed that each child's lexical restructuring

would be different in terms of rate and accuracy. Thus, based on a huge body of research on

phonological awareness deficit in dyslexia, it is hypothesized that dyslexic children will show

delayed or slow developing lexical restructuring.

Similarly, Snowling et al. (1986) and Metsala (1997) provided a few evidences. The

former study reported that, when appropriate lexical contexts were provided, dyslexic children

were better at recognizing phonologically ambiguous words, suggesting a more holistic lexical

presentation. The latter study used a speech gating task, in which small segments from onset of

words are presented via headphones (e.g., /f/, /fu/, or /fud/ for fudge). Dyslexic children needed

more acoustic information than age-matched controls to recognize words when the target word

had few similar-sounding neighbors, suggesting that their phonological representation is not









totally decomposed into phonemes when compared to the controls. This was interpreted by

Metsala (1997) to evidence a delayed segment-based specification of such words.

Application of representation hypothesis

The explanatory power of the phonological representation hypothesis is pronounced when

it comes to weaknesses of other phonological tasks. First of all, poor phonological representation

would easily explain dyslexic children's typical problems in phonological awareness. Since the

child only knows that the sound after initial st- is consonant, but not which, he or she would not

efficiently delete [r] and say sting when asked to say string without saying [r]. Or the child

would not easily say ring by deleting st- when asked to say string without saying [st] since he or

she is now aware of the actual consonant after st- (Elbro & Jensen, 2005). In general, phonemic

manipulation tasks (segmenting, blending, etc.) would be very hard if incoming segments are

underspecified or incorrectly specified.

Second, the distinctness hypothesis (or representation hypothesis) may also account for the

picture naming speed deficits, a strong predictive factor of reading disorder. For successful

lexical access, it is easier to get access to a phonological representation that is well specified and

clearly separated from its neighbors than to unstable or underspecified representations (e.g.,

Katz, 1986, Bowers & Wolf, 1993). In this way, the representation hypothesis may explain the

naming speed deficit.

Third, low phonological distinctness may also be linked to the delayed decoding capacity

of dyslexic children. Decoding involves the following steps and each step has its own

requirement for successful decoding.









* Step 1. Consecutive grapheme-to-phoneme conversion: each letter should be converted into
its corresponding phoneme (i.e., p--[p], r--[r], i--[i], n--[n], t--[t]).

* Step 2. Representation accuracy (distinct coding): each sound linked to a certain
orthographical pattern should be accurately coded in the decoder's phonological
representation in a fixed order.

* Step 3. Short term memory: every phoneme obtained from step 2 should be held in a
phonological buffer temporarily, waiting to be blended to produce a whole word print.

* Step 4. Blending (phonological manipulation): each phoneme will be ultimately put together
to form a final representation of a word.

If any of these procedures are impaired, decoding problem can occur. Especially, efficient

short term retention and blending process would be totally dependent upon the quality of

phonological representation of each phoneme converted from input letters. It is expected that

step 3 and step 4 will be damaged by underspecified input. Likewise, all phonology-based tasks

utilizing underspecified input representation will be deteriorated in a differing degree. These

chain-like negative effects of inaccurate phonological representation on related phonological

tasks are summarized in the Figure 2-4.

A challenging observation

Based upon the fact that dyslexic children's low performances on phonological processing

tasks are related to processing incoming sound information, Goswami (2000) indicated that the

input-based representation is the major obstacle for further literacy development. So, low scores

on input measures of speech perception should differentiate dyslexic children from controls;

Dyslexic children should find it more difficult to discriminate between different phonemes in

speech (phoneme perception), they should find it difficult to recognize spoken words (auditory

lexical decision), and they need more phonological information or clearly articulated input for

accurate spoken words (speech gating).









However, lots of studies found that such speech recognition-related skills do not

consistently characterize individual dyslexic children even though all these difficulties have been

reported in group or population studies (Manis, Doi, & Bhadha, 1997; Mody, Studdert-Kennedy

& Brady, 1997). Instead, one consensus has been documented that dyslexic children showed

significantly impaired scores on tasks tapping their production skills such as word finding

difficulties (rapid naming), rote repetition of non-words (STM for nonsense word repetition), and

less distinctness of vowel production (See Figure 2-5).

This subtle observation raises some challenges to the theory of phonological representation

deficit. Is dyslexia an input-processing deficit (representation deficit)? Or, are other factors

necessary for phonological production such as insufficient verbal memory resource or

phonological awareness capacity associated with dyslexia (post-input processing deficit)? In

spite of the above observation, Goswami (2000) argued that it would be logically possible that

our current measures are failing to tap the main representation deficits causing underspecified

representation, or that compensation for early representation deficits have already occurred in

some children after remediation.

Even though this theory is right, the fundamental difficulties measuring children's inner

phonological representation in their mental system would make experimental design hard.

According to Elbro (1998), studies of the quality of phonological representations are scarce. In

addition, with a good deal of possible evidence for representational or auditory perceptual deficit

hypothesis, the fine-grained explanation of the relationship between underspecified phonological

representation and associated phonological tasks (phonological awareness, working memory,

lexical retrieval) still remains to be resolved to reveal the exact mechanism responsible for the

apparent impairment seen at the production level as indicated.









A critical summary

There is now strong and highly convergent evidence in support of weak phonological

coding as an underlying cause of dyslexia. Dyslexic children's impaired phonology-based skills

such as phonological awareness and alphabetic decoding are believed to be due to the weak

coding skill. Other problems such as slow lexical storing and/or lexical access, impairment seen

in short-term memory tasks are also viewed as stemming from the phonological

underspecification (Vellutino et al., 2004). In turn, difficulties in word storing and retrieval can

impair the beginning reader's ability to establish a stable connection between the sublexical

phonological codes constructed from oral or graphical stimuli and permanent phonological codes

in the lexicon, which will ultimately impede the reader's efficient word identification. Fluency in

word identification is a critically important prerequisite for adequate reading comprehension

(Perfetti, 1985). Compromised short-term phonological memory can also impair the reader's

decoding capacity, for which each phoneme converted from letters needs to be blended in an

orderly manner. Thus, phonological coding weakness was hypothesized to be a core deficit

which could cause a collection of phonology-based difficulties observed in dyslexics such as

word identification (retrieval), phonological awareness, decoding of letter sequences, rapid

naming, vocabulary learning, and nonword repetition.

This hypothesis now raises an essential question of why dyslexia is associated with

underspecified phonological representation. Very little work has been done on the possible

causes of poor representation. Metsala (1999)'s well-known 'lexical restructuring theory' is in

line with the phonological representation hypothesis. In this theory, the unit of phonological

representation gets smaller starting from whole utterance to phrase, words. For some unknown

reason, this lexical restructuring appears to take place at a lower rate in dyslexic children than

others. But, this hypothesis also does not readily explain why dyslexics' weak phonological









awareness is sustained after their vocabulary grows at a near normal rate and the theory itself

does not provide the root cause of delayed lexical restructuring in dyslexic children.

Double-Deficit Hypothesis

Deficits other than phonological ones characterize individuals with dyslexia. Impaired

performance on rapid serial naming tasks distinguishes individuals with dyslexia from those with

other developmental disorders, like attention-deficit disorder (Felton, Wood, Brown, Campbell &

Harter, 1987; Felton & Wood, 1992).

Research suggests that for most children, there are two main aspects that drive the

development of fluent decoding skills (Lovett, Steinbach, & Frijters, 2000). First, children

should have good meta-phonological awareness skill to identify or manipulate sounds within

speech. Secondly, children should be able to process visual or orthographical information very

rapidly to be a good reader. This skill is a part of rapid automatic naming (Torgesen, Wagner,

Rashotte, Burgess, & Hecht, 1997) and it is reported that children with slow retrieval time for

pictures or objects have similar delayed naming with letters or printed words. This slow

processing of visual information to activate lexical storage will put some students at a

disadvantage when it comes to reading (Lovett et al., 2000).

RAN as a phonological processing skill

Many research studies have demonstrated that RAN makes a contribution to reading that is

independent of the contribution of other predictors of reading ability such as phonological

awareness and memory (Bowers, 1989; Bowers, Steffy, & Tate, 1988). Nonetheless, RAN has

often been placed within the phonological processing domain, along with phonological

awareness (both synthesis and analysis) and verbal working memory (Wagner et al., 1994;

Torgesen, Wagner, Rashotte, Burgess, and Hecht, 1997). This idea is based on Denckla and

Rudel (1974)'s early suggestion that RAN is basically a phonological processing skill. Those









who believe that RAN is a component of phonological processing, alongside phonological

awareness and memory, define RAN as the "efficiency of phonological code retrieval" (Wagner,

Torgesen, Laughon, Sommons, & Rashotte, 1993; Vellutino, Scanlon, Sipay, Small, Pratt, Chen,

& Denckla, 1996). Torgesen and his colleagues regarded naming speed into "a part of the

phonological family" (Torgesen et al, 1997b). Researchers who conceptualized naming speed as

a phonological process suggested that naming speed, like any other linguistic tasks (e.g.,

expressive vocabulary), involves accessing and retrieving a phonological code stored in long-

term memory. One of the reasons for this approach is associated with sufficient amount of

correlation between serial naming and performance on other phonological processing tasks and

based on that, it is argued that serial naming tasks should be included as part of assessment of

children's phonological processing abilities.

However, others have argued that deficits in visual naming speed and phonological

processing are distinct and dissociable aspects (Bowers & Wolf, 1993; Wolf & Bowers, 1999).

For them, RAN is not a subsidiary component of phonological processing and the evidence

based on correlations between speed naming and other phonological skills is insufficient reason

to categorize and subsume naming speed under phonology. This opinion that RAN represents

functions separate from the phonological processing domain stems from the facts that: (1) RAN

consistently makes a unique contribution to reading beyond phonological awareness and

memory; (2) some group of poor readers show a specific clinical characteristics (i.e., adequate

decoding skill, but slow reading speed).

Three subtypes

Impaired RAN skills, about which there is more controversy, concerns deficiencies in

visual naming speed, that is, impairments in rapidly accessing and retrieving names for visually

presented symbols (letters, numbers, pictures of objects, or animals) even though the names are









quite well known to the child. Wolf and Bowers (1999) proposed a neurocognitive model of

reading disability, the Double Deficit Hypothesis (DDH), along at least 2 dimensions of dyslexia

subtypes with unique causes of reading failure. According to the DDH, these subtypes are

classified according to the presence or absence of phonological and naming-speed deficits, where

one subtype exhibits both deficits.

* Readers in i/hin the phonological deficit subtype are characterized by marked difficulty
decoding words and exhibit little phoneme awareness of the sound structure of words.
Such individuals have special difficulties with pseudo-word reading. They are typically
identified by significantly low scores on tests requiring blending phonemes into words or
pronouncing parts of words by removing one or more phonemes (phoneme elision tasks).

* Readers within the second, naming-speed-deficit subtype are markedly slower at serially
naming a sequence of visually presented stimuli, such as letters or numbers. This deficit is
believed to encompass problems with the automatic or rapid word retrieval necessary for
the development of fluent reading and ultimately comprehension.

* Readers with problems on both phonological and naming speed tasks (i.e., more than one
standard deviation below average) fall within a double-deficit subtype. These individuals
regularly represent the most severe cases of reading failure.

According to the DDH, as compared with phonological processing ability, rapid serial

naming speed contributes unique variance to reading rate and fluency measures (Bowers, 1995),

and is a stronger predictor of reading ability in more orthographically transparent languages like

German (Wimmer, 1993) and Dutch. These findings suggest a unique role for the processes

underlying the rapid recognition and retrieval of visually presented linguistic stimuli, and that

they may not be subsumed under the phonological core deficits of dyslexia.

Differentiation from phonology

Intrinsic cognitive complexity of RAN: The rationale for emphasizing the difference

between naming speed and phonology is based on the complexity of cognitive procedure

involved in them. Wolf, Bowers, and Biddle (2000), using a model of simple letter naming task,

argued that serial naming and its internal complexity go beyond phonological processes. For









them, visual naming represents a demanding array of attentional, perceptual, conceptual,

memory, lexical, and articulatory processes and all of these components places heavy emphasis

on precise time requirements.

The following is a simplified version of the model they used to illustrate the complex

processes involved in visual naming (Figure 2-6)

* Step 1 (Activation of attention system): Naming task requires the participants to focus on the
task with high level of attention.

* Step 2 (Visual processes): Based on adequate level of attention, the letter's visual
information will be processed and this will allow for identification process by integrating
information about the present stimulus (letter 'B') with known mental, visual representation.

* Step 3 (Integration processes): After Step 2, lexical information will be integrated with
accumulated visual information.

* Step 4 (Lexical access/retrieval): Lexical information, especially the target letter's
phonological code, will be accessed and retrieved.

* Step 5 (Production): Motor planning and articulation will translate this phonological
information into an articulated name.

This model exemplifies both the importance of access to the phonological code in naming

and the fact that phonological processes represent only one subset of the multiple processes

involved in naming. Also, the authors emphasized the importance of speed requirement needed

for fast naming and this is marked as PSR, which is required in every step of whole model.

Different from single letter recognition, serial letter naming add to this model the extra demands

of rapidity, maintenance of attention, and serial processing. In sum, the model's inherent

complexity, the extent of processing speed demands, and the addition of rapid rate and seriation

make naming-speed task a quite different cognitive task from phonology (Wolf, Bowers, &

Biddle, 2000:393).

Other reasons for differentiation: In addition to the cognitive difference, there are other

evidence supporting the differentiation between naming speed and phonology. First, the









correlation found between phonological measures and RAN tasks are modest. There is variability

in this finding; it is reported that correlation is higher in younger readers and it gets weaker in

relatively older readers since naming speed approaches somewhat automatic rates in average

readers between Grade 1 and 2 (Biddle, 1996).

Secondly, there are different patterns of relationships with reading subskills that

characterize naming speed and phonological variables. Bower and her colleagues found that

phonological tasks strongly predicted word/nonword decoding, but not word and text reading

speed. Rather, naming speed predicted latency (response time) of word identification and speed

of text reading (Bowers, 1993; Bowers & Swanson, 1991). In Bower (1995), only naming speed

contributed to speed on reading measures. After finding similar results, Cornwall concluded that

"naming speed may represent unique aspects of the reading process as opposed to an overall

phonological ability" (1992:537). Given this evidence, to subsume naming speed only under

phonology will obscure its separate role in predicting reading disorder.

Linkage between RAN and reading process

The relation between phonological processing skills (awareness, memory) is easily

understandable. Specifically, learning to decode through grapheme-to-phoneme conversion

requires an acute awareness of and an ability to analyze or blend the sounds within words. For

this to be successful, normally developing memory skill must be associated with age-appropriate

decoding skills. In contrast, no such straightforward conceptualization exists to explain how the

processes underlying naming speed affect word identification and nonword decoding. Wolf et al.

(2000) proposed two speculative hypotheses on the possible roles of orthographic processing.

* Orthographic pattern recognition: Slow naming speed can prevent the appropriate
amalgamation of the connections between phonemes and orthographic patterns at sublexical
level. When a reader does not efficiently merge phonemes decoded from each letter due to
slow access to phonological codes, it will surely reduce the efficiency of word recognition.









* Second route involves the importance of the learned association between letters in the
development of orthographic representation. That is, if a beginning reader is slow in
identifying individual letters (as indexed by RAN), then single letters in a word will not be
activated in sufficiently close temporal proximity to allow the child to become sensitive to
the knowledge about orthographic (letter) patterns (Bowers, Golden, Kennedy, and Young,
1994). Similarly, Bowers and Wolf (1993) and Wolf & Bowers, 1999) suggested that deficits
in visual naming speed disrupt reading acquisition by inhibiting growth in the connections
between phonemic and orthographic patterns at word and subword levels of representation
during word identification learning. The possible delay and inefficiency in establishing this
relationship is believed to reduce the quality of orthographic codes in memory.

Not all scholars seems to accept the view that reading-related cognitive deficits are caused

exclusively or primarily by phonological limitations. Double deficit theory suggests that naming

speed deficits are caused by disruption of a 'precise timing mechanism' that may influence

temporal association of visual and phonological counterparts, thereby impairing the child's

ability to detect and represent orthographic pattern. This compromised connection between

visual and phonological pattern of a word cannot be explained only by indistinct phonological

representation. As discussed above, more general cognitive problem such as visual pattern

matching may be involved in impaired RAN

Current research may not be able to establish a unifying framework with regard to the

differences in the psycholinguistic mechanism and underlying causes between these two

components. By positing two separate deficits (DDH) cognitive pathways (visual vs. decoding-

based), the single-deficit approach (phonological representation theory) may not be the ultimate

answer. It may have to be posited that there are multiple causes of reading disorder and there are

other cognitive areas that cross many domains including visual, auditory, and, possibly, motor

processing. All of these functions need to be developed simultaneously with a normal phonology

for successful literacy.









Auditory Perceptual Deficit Hypothesis

Researchers on the opposite side suggested that phonological deficit should be understood

in the context of more general 'sensory' or 'learning impairments' (Nicolson and Fawcett, 1990;

Tallal, Miller, & Fitch, 1993). This 'auditory perceptual theory' has been proposed by Tallal and

her colleagues (Tallal, 1980; Tallal et al., 1993). They have argued that dyslexic children have

particular difficulties in processing rapidly changing or transient acoustic events, and that the

ability to process rapid successive information is fundamental to setting up the phonological

system. It is posited that children with language-learning disabilities and dyslexia process sounds

more slowly than the average children and this may diminish their ability to distinguish

phonemes.

Evidence for this theory comes from studies that measure how much time (inter-signal

interval, known as ISI) children need between two sounds before recognizing that there is more

than one sound. The children showed deficits in comparison to controls when one stimulus

rapidly followed another in both a temporal order judgment paradigm (TOJ) and a same-

different discrimination paradigm. Similar deficits were then observed in 8 out of 20 dyslexic

children (Tallal, 1980). ISI is called 'timing threshold' and it is interpreted as the time it takes for

nerve cells to fire and process a sound's acoustic features and then recover to pick up the next

sound. While the average child has an average timing threshold in the tens of milliseconds range

for simple tones, children with reading disabilities have time thresholds measuring hundreds of

milliseconds. For example, because the difference between 'ba' and 'ad' occur within tens of

milliseconds, children who need longer time to detect changes may not hear the difference. If

they cannot discriminate phonetically similar sounds, this may cause underspecified

phonological representations of words in lexicon. Theoretically, it was argued that a rapid

processing deficit could affect literacy because efficient processing of transient information is









critical for phoneme perception and fully specified phonological representation of words in

permanent lexicon. This stable, correct phonological specification is regarded as the basis of

competent phonological processing (phonemic awareness, short tem retention, and lexical

access), which are necessary for successful reading.

This 'auditory processing deficit' theory has become so dominant that a remediation

package based on the elongation of brief perceptual cues has been developed and is administered

to thousands of children (Merzenich, Jenkins, Johnson, Schreiner, Miller, & Tallal, 1996).

However, this theory has lost favor despite its logical appeal (Goswami, 2003). Common

criticisms are that: (i) positive findings are difficult to replicate; (ii) that only sub-groups of

dyslexics are affected; (iii) that when positive relationships are found they are more robust in

control groups, (iv) and that when auditory deficits are found they tend to be small (Ramus,

2003; Rosen, 2003). In response to the criticism, the proponents of this theory argue that, by the

time children are diagnosed with dyslexia around age 9, their brains may have compensated for

the auditory deficit, but early deficit may have laid the foundation for trouble with other subtle

phonological processing skills such as phonemic awareness.

Hearing loss, Phonology, and Literacy

It is now a well established consensus that phonological representation capacity plays a

crucial role in language development and language-based cognitive functions such as verbal

memory, speech perception, metalinguistic awareness, and literacy in normal hearing children.

If, for some reason, children's phonological representation is adversely affected, it is

hypothesized that skills mentioned above closely associated with appropriate phonological

foundation would be compromised. As far as hearing loss is concerned, the most important

question is to what extent children's phonological capacity will be damaged by limited auditory

input (Leybaert, 1998).









It should be mentioned that most of previous research have focused on children with

profound hearing loss and very few is known about the literacy or literacy-related cognitive skills

in children with mild-to-moderate hearing loss. Allen & Schoem (1997) reported that even hard-

of-hearing children, that is, children with only mild-to-moderate hearing loss, read at lower

median levels than do hearing children. Another limitation of the previous research on the

literacy development of heard-of-hearing children is that evaluations of reading skills were done

in very limited areas. Thus, understanding the developmental change of a wider and

comprehensive range of literacy and related cognitive skills is not yet seen in the current

literature.

Phonological Processing, Language Skills, and Hearing Loss

Children's poor word learning can be explained by either misrepresented phonological

features or insufficient memory resources (Ramus, 2001b). The quality of incoming auditory

input will be limited in children with hearing losses and this might cause the phonological

component to be affected especially during the so-called 'sensitive' or 'critical' period of

language acquisition. This time-specific delay in phonological component may be the cause of

persistent language impairment and later literacy weakness in children with HL. Under-specified

phonological input or lack of enough storage capacity can adversely affect normal lexical

growth. If a child has a permanent or temporary limitation of auditory input, his overall

phonological processing skills can be impacted by way of various routes. Therefore, different

from normal-hearing children, hard-of-hearing children's compromised phonological processing

skills can have a negative impact on later vocabulary, grammatical knowledge, and overall

language competence.









Otitis media with effusion (OME)

Fluctuating conductive hearing loss associated with repeated episodes of otitis media with

effusion (OME) occurs most often during the first three years of life, a time that is most

important for age-appropriate language development. OME is the presence of the fluid in the

middle ear that is not infected. OME typically causes a mild-to-moderate hearing that lasts as

long as the fluid persists and the variability in auditory input due to OME has been hypothesized

to disrupt children's ability to code phonological information accurately into their phonological

representation system (working memory), building up inaccurate sound representation in the

lexical storage (long-term memory).

Several studies have tried to relate an early history of OME to language development

during the first 3 years of life to examine the effect of OME, which usually occur during critical

period of spoken language acquisition. It is posited that early normal auditory functions would

facilitate the acquisition of language-specific speech perceptual strategy. In turn, normal access

to incoming sounds would also facilitate: (1) the phonological coding via high-quality resolution

of acoustic features; (2) stable formation of phonological representation of a word in lexical

memory; (3) efficient word retrieval, and (4) enhanced phonological short-term memory

(Nittrouer and Burton, 2005).

One correlation study reported that children with OME episodes during the first 2 years of

life scored slightly lower in expressive language, but caught up by second grade (Roberts,

Burchinal, & Zeisel, 2002). Early impact of OME (first 2 years) on expressive language skills

seems to shed light on the importance of "critical time window" for language acquisition, beyond

which children do not show fast developmental rate. But, this possible delay in expressive

language skills seems to be resolved as children enter 2nd grade level if children are provided









appropriate audiologic or speech-language intervention. This suggests that OME's effect on

language development is not substantial.

Similar observation can be found in a very recent study. Majerus, Amand, Boniver,

Demanez, Demanez, & Linden (2005) reported that 8-year-old children with OME had strictly

normal performance for (1) expressive/receptive vocabulary, (2) verbal STM (new word

learning/serial words recall), but (3) slight decrease of performance was found in phonological

processing/ awareness level (nonword repetition, a rhyme judgment task) when compared with

normal-hearing controls. In Roberts et al. (2002)'s study, students caught up with the hearing

peers' language skills by the second grade and this period approximates to the mean age of 8,

which was reported to be the chronological age when no continuing language delay was

documented in Majerus et al. (2005).

However, different observation was documented in Nittrouer et al. (2005). This study

analyzed 5-year-old children's phonological awareness, verbal short-term memory, and sentence

comprehension. Children with OME showed impaired phonological awareness and verbal

working memory. Different from other previous studies, which reported no significant risk for

language development (cf. Majerus et al., 2005, Briscoe et al., 2001), this research found

significant delay in language (sentence comprehension) and indicated that impaired phonological

processing was associated with later language delay (syntax). Nittrouer and her colleagues

suggested that this different result should be reinterpreted in the context of methods used. It is

indicated that most of previous studies reporting no significant language impaired in children

with OME used parental checklists (e.g., MCDI), or non-in-depth standardized assessment tools

(e.g., CELF), which they say may not assess in-depth language skill. With this in mind, it can be









expected that Majerus et al. (20015's documentation of normal language skills found in children

with impaired phonological processing skills might be changed if we use in-depth language tests.

Nittrouer and the colleagues also suggested that no group difference between normal-

hearing and hard-of-hearing children might be explained by the fact that the majority of children

in both OME and non-OME groups came from low-SES group. Because of high correlation

between SES and language skills, the possible language difference might have been masked by

similarly low language skills in both groups. For this reason, careful control of the SES factor is

required for exact evaluation of later language, or literacy capacity. Finally, the mean age of

participants also appears to be an important factor for this outcome discrepancy.

As mentioned in Majerus et al. (2005), overall evidence for residual STM impairment in

children with OME remains equivocal depending upon different tasks used and subject-related

factors such as family supportiveness, quality of later audiologic, speech-language intervention,

and socio-economic status (SES), which usually were not reflected in a measurable way.

According to Majerus et al. (2005), hearing loss can affect verbal STM or long-term lexical

memory in three different ways: (1) residual perceptual deficit might reduce the accuracy of

phonological coding (initial representation in the model) in STM area; (2) this poor phonological

coding can cause poor phonological memory because of insufficient acoustic resolution; or (3)

limited perceptual input during infancy may lead to less structured phonological form in mental

lexicon, which is related to later weak vocabulary level. Phonological representation of a lexical

item with weak and low acoustic resolution would hinder efficient both short-term memory tasks

and fast lexical retrieval for age-appropriate comprehension/production.

Majerus et al. (2005) reported that OME does not have adverse effect on verbal STM, and

phonological processing ability would be subtly impaired. They went on to say that this might









indicate the delay of phonological processing tasks in children with OME cannot be related to

their normal verbal STM capacity. In this study, nonword repetition skill was shown to be

slightly impaired, but it is not sure why this task was not classified as verbal STM. In Norbury,

Bishop, & Briscoe (2001) and Baddeley et al. (1998), it is clearly indicated that nonword

repetition is best understood as a measure of the capacity of the phonological short-term

memory. If we follow this standpoint, the data in this study confirm that part of verbal STM skill

(nonword repetition) is significantly decreased. In this respect, further in-depth analysis of the

memory skills tapped into by different tasks is strongly recommended, without which it is

meaningless to discuss the effect of hearing loss on phonological verbal memory (i.e., STM,

WM, or LTM).

Interestingly, verbal STM skill in children with a history of OME was shown significantly

delayed or quite normal according to the age of participants. In contrast to Majerus (2005),

Nittrouer et al. (2005) reported delayed verbal STM. The average age of participants in this study

was 5 (age range: 4; 11-5; 11 months) and the mean age of subjects in Majerus et al. (2005)'s

study, which reported 'normal' verbal STM capacity, was 8 years. Considering the fact that the

subjects in Majerus and colleagues' (2005) study showed strictly normal vocabulary skill, it is

conceivable that limitation in verbal STM in early ages reported does not seem to have negative

impact on later language functioning. That is, early history of OME can have adverse impact on

initial verbal STM capacity and this prompt loss of memory skill seems to be recovered. This

could be the reason why children's language does not show significant delay later on (Majerus et

al., 2005; Briscoe et al, 2001) at the age of 7 to 8 years (cf. second grade). It could be that

intensive audiologic or speech-language therapy combined with good family support and SES

helped children to catch up with the language skills of normal-hearing peers.









Sensorineural hearing loss

To date, substantial amount of research has been conducted to investigate the effect of

OME on later language. Thus, much less is known about the development of phonology and

language in children with permanent mild-to-moderate SNHL (Stelmachowicz, Pittman, Hoover,

& Lewis, 2004; Briscoe et al., 2001; Gilbertson & Kamhi, 1995; Plapinger and Sikora, 1995;

Davis et al., 1986).

Briscoe et al. (2001) assessed three phonological processing skills: phonological

discrimination, phonological awareness (onset-rime detection), and STM (nonword repetition) in

4 groups of children (SLI, SNHL, and 2 control groups). Children with SNHL showed scored

significantly weakened phonological processing and phonological STM skills than a control

group matched on chronological age. But, no difference was observed in SNH and CA control

groups on vocabulary and sentence comprehension (syntax). Impressively, considerable

individual variation within the SNH group was found; nearly 50% of the SNH group classified

'impaired group' showed phonological impairment associated with poorer vocabulary than

remaining children who had no impaired phonology and vocabulary, but both groups' language

skills were within normal range.

It is notable that the 'unimpaired' group showed relatively milder hearing loss. This might

suggest a causal link between the degree of hearing loss (impaired phonology) and the resulting

decrease in language skill in the 'impaired' subgroup. But, the overall result is that impaired

phonological processing due to SNHL can occur without clinically significant deficits in

language. As criticized by Nittrouer et al. (2005), this might be related to shallow assessment of

language skills, which may not find existing language difference in deep level.

Gilbertson and Kamhi (1995) were concerned about phonological skills and new word

learning in children with SNHL. Substantial variability was found on measures of vocabulary









and word learning, that is, half of the children with SNHL showed normal development and the

other half of the children (n=10) performed more poorly than the first group. Based on this

observation, they argued that children with SNHL can be divided into two distinct subgroups: (1)

one lower-functioning group with impaired word learning, vocabulary; (2) one higher

functioning group with normal language ability. They went on to say that hearing level was not

significantly related to word-learning or measures of language skills since the higher-functioning

group had poorer SRT scores than the lower-functioning group. It is argued that the higher-

functioning group's better language (lexical) skill is related to their better phonological memory

(new word learning, word repetition). This had led them to the conclusion that one out of every

two children with a hearing loss might be considered as language-impaired.

However, for the following reasons, this seemingly important conclusion needs to be

carefully interpreted. First, the data show that low-functioning children have better rapid naming

skill which is a measure of phonological long-term memory, but this observation was not

discussed at all. In contrast, the better STM function (word repetition) of the higher-function

group was regarded as a crucial factor for efficient lexical acquisition. Secondly, the scores in

grammatical understanding test did not yield a significant language gap between two groups.

Also, the small number of subjects in both groups (n=10 and 9) also needs to be considered in

terms of statistical power. Thirdly, it is argued that this intra-group difference is not related with

the severity of hearing loss, but no explanation regarding the possible link between hearing loss

and lower functioning group's impaired phonological memory (STM) was provided. Lastly, as

noted by the authors, substantial amount of variance was observed in the performance of

children. The delayed vocabulary in their lower-functioning group may have resulted from a set

of other extraneous factors such as poor family factor (responsiveness, supportiveness), low









quality of intervention, inappropriate educational program, higher age of identification, short

duration of hearing aid wearing, low SES, and so on. All of these factors might have contributed

to the group difference. Considering the inappropriate data interpretation and the variance

seemingly linked to other uncontrolled factors, vocabulary difference may not be a definitely

reliable criterion for suggesting that one out of every two hearing impaired children has

concomitant language disorder.

Hearing Loss and Literacy

Conductive hearing loss

Links between audition, language, and literacy have been widely studied in children with

conductive hearing loss caused by OME (Friel-Patti, 1990; Mody, Schwartz, Gravel, & Ruben,

1999). Despite a considerable number of studies conducted during the past three decades on

whether children with frequent OME score lower on measures of speech, language, and reading

than children without such a history, the literature is still controversial. Moreover, there are very

few comprehensive studies on the effect of OME-related hearing loss on literacy development

and the results of research, if any, are quite heterogeneous. One problem of most of studies is

that very limited areas of reading or underlying skills of reading are assessed. For example, one

longitudinal study with subjects from second graders measured only one area for reading

assessment: letter-naming (identification) skill and concluded that no correlation was found

between a history of OME occurred up to 4 years of age and literacy (Roberts et al., 2002). This

is surely a problem since the subjects in this study were at age range of 5 to 7 years of age, when

most of children would master letter naming.

Peters (1994) conducted a longitudinal study testing the effect of middle ear infection on

reading and spelling. Subjects were tested on non-word reading, word recognition, reading

comprehension, and sentence identification at ages of two, four, and seven. The results indicated









that ear infections had a significant effect on spelling, but not on reading. Some authors reported

difficulties with phonological skills with unimpaired general language skills in children with OM

(Peters, 1994), but others did not reveal any difference (Roberts, Burchinal, & Zeisel, 2002). A

recent meta-analysis by Roberts, Rosenfeld, and Zeisel (2004) reported no significant difference

a number of standardized language measures at preschool age and OME-related hearing loss.

Share and Chalmers (1986) is also a longitudinal study on the relationship between ear

infection and reading. They reported no significant effect of middle ear infection on reading

ability. Hemmer and Ratner (1994) studied the impact of middle ear infection on literacy using

six pairs of twins through a longitudinal observation. One of each pair had a repeated ear

infection and one did not. Middle ear infection was shown to have a negative effect on

vocabulary skill, but had no effect on speech, other linguistic skills, and reading. Abu-Rabia

(2002) sampled 49 first graders all from low SES families and 11 children with a history of at

least one episode of middle ear disease were assessed for a wide range of phonological

awareness skills and pseudo words reading. No significant differences were found on the

phonological awareness tasks, and non-word reading task. These findings accord with the above

results, suggesting that middle ear infection does not adversely impair children's phonological

sensitivity and reading development.

In contrast, different result was reported by Nittrouer (2005), in which four groups of

second graders participated. It is asserted that conductive hearing loss may have adverse effect

on language when it occurs in association with other social or health risk factors. In this study,

one group, the control, was from the middle class and had no ear infection; the second group,

also from the middle class, had a history of ear infection; the third and the fourth groups of low

SES, was with or without ear infection history, respectively. The rationale of the study was that









middle ear infection causes temporary hearing loss, which Nittrouer expected would affect the

amount of exposure to language. Nittrouer hypothesized that limited language experience would

account for the variance in phonological awareness skills.

As stated above, the underlying hypothesis of this study was that phonological awareness

can only be developed via explicit exposure to acoustical speech sounds. Significant difference

in the performance on phonological awareness tasks was found between the two groups from the

middle class, but no differences was reported between groups with low SES, i.e., both groups

from low SES families performed at a low level. Nittrouer explained that this lack of group

difference in samples from low SES families is due to the fact that slight exposure to print of

these children in their homes provided a negative environment against normal phonological and

reading development. So, regardless of the hearing status, the possible group difference might

have been masked by simultaneous impairment due to the low SES. Similarly, Yoshinaga-Itano

(1999) reported adverse overall language outcomes for children attending early intervention

services even when the loss was classified as mild, with outcomes more strongly related to age of

diagnosis than to severity of loss. Similarly, Davis, Elfenbein, Schum, and Bentler (1986)

reported receptive vocabulary, verbal ability, and reasoning to be more than one standard

deviation below the mean, even in children with 'mild' hearing loss.

As seen from the literature, the effects of fluctuating hearing loss related to middle ear

infection, more specifically the OME, on phonology, language, and literacy are still not

explicitly determined. However, the overall impression is that mild or moderate conductive

hearing loss does not severely impair children's phonological skills. Based on this observation, it

can be expected that skills of language and literacy associated with conductive hearing loss

would not be significantly compromised. However, reflecting upon possible delay though not









statistically significant, it is strongly suggested that further studies include the assessment of

wider and deeper range of phonological capacities, reading, and cognitive skills underlying

reading process in the children with mild or moderate conductive hearing loss for evidence-based

intervention design.

Sensorineural hearing loss and literacy

Reading development in children with permanent SNHL has received relatively much less

attention than conductive hearing loss related to middle ear disease. In contrast to children with

fluctuating hearing loss, very scant research efforts have been directed to literacy development in

children with permanent SNHL. Although there is a large body of research documenting the

language and literacy outcomes of children with SNHL, most of this was concentrated on

children with profound hearing impairment. So, much less is known about literacy development

for these children, even thought it can affect not only hearing level thresholds, but also frequency

discrimination (Moore, 1995) and further research strictly focusing on the reading and reading-

related cognitive skills in this population is warranted.

Briscoe et al. (2001) compared 5- to 10-year-old children with mild-to-moderate SNHL,

children with specific language impairments and children with no hearing or no language-

learning difficulties. In common with the SLI group, mean scores of children with hearing loss

were significantly poorer on tests of phonological STM (non-word repetition), phonological

discrimination, and phonological awareness than chronological age controls. But, no differences

were observed in SNH and age control group on language (vocabulary, sentence

comprehension), digit/sentence recall (STM), and literacy (reading comprehension, nonword

decoding, and word recognition). That is, while the data revealed little overall difference of the

reading abilities between the hearing impaired and normally hearing children, the hearing









impaired children's phonological skills were significantly inferior to children in the age-control

group.

Interestingly, Briscoe and colleagues' study also reported that there was considerable

individual variation within the SNH group. That is, nearly 50% of the SNH group showed

phonological impairment associated with poorer expressive and receptive vocabulary and

literacy. In addition, this subgroup showed higher hearing thresholds than remaining children

without phonological impairment. Thus, it was suggested that there was a link between

vocabulary (language) and literacy and phonological skills associated with hearing loss.

Strikingly, all three areas of reading (word/nonword reading, comprehension) were shown quite

normal (99.5, 91.6, and 95.3, respectively), revealing no significant between impaired and

unimpaired subgroups. Overall, this study suggested that major problems in nonword repetition

and depressed phonological component occurred ii ilthuit clinically significant deficits in wider

language and literacy abilities in children with mild-to-moderate SNHL.

Impaired phonology in children with hearing loss

In Briscoe and her colleagues' (2002) study, it was hypothesized that children with hearing

impairment might show phonological impairment similar to those seen in SLI and this was well

observed. But, this expected link between compromised damaged phonology and other language

and literacy measures was not found in children with SNHL.

The hearing impaired children's success with reading in spite of depressed phonological

processing skills is very notable. It was suggested that even though slight language deficiencies

(receptive/expressive vocabulary) were observed, phonological depression linked with auditory

limitation would not impede reading skills of children with mild-to-moderate SNHL and at least

some reading may be possible even without closely associated phonological skills (Gibbs, 2004).

This led the authors to conclude that auditory deficit can compromise phonological skills, but









impaired phonological skill does not necessarily lead to serious impairments in the reading

performance of hearing impaired children.

Theoretically, this argument poses a very important aspect of impaired phonological

component found in children with hearing loss since the argument itself is the converse of

numerous studies of children with dyslexia, which all agree in indicating that poor phonological

capacity will compromise the development of word reading.

A very similar phenomenon was reported in Gibbs (2004)'s study. Fifteen children with

mild-to-moderate bilateral SNHL were compared to normally hearing controls on their

performance on reading and underpinning skills. Gibbs showed that hearing impaired children's

reading abilities were indistinguishable from their hearing peers, while showing phonological

skills that were not as good as the controls. So, Gibbs (2004) supports the findings of Briscoe et

al. (2001) and offers a similar challenge to the necessity of phonological skills in the

development of early reading and the universality of phonological core deficit theory.

Summary

The following is a short summary of current literature about the effect of hearing loss on

phonology, language, and literacy.

Conductive hearing loss

* While there have been a substantial number of studies of profoundly deaf children's
reading, the literacy of children with mild-to-moderate fluctuating conductive and
permanent SNHL has relatively not received enough attention.

* Results from the studies conducted during the past three decades on the effect of OME-
related hearing loss on speech, language, and reading are still controversial.

* However, the overall impression is that mild or moderate conductive hearing loss does not
impair children's phonological and linguistic knowledge in a way that their reading level is
significantly delayed or depressed. Thus, we anticipate that skills of language and literacy
associated with conductive hearing loss would not be significantly compromised.









Permanent SNHL


* In contrast to children with fluctuating hearing loss, very scant research has been directed
to literacy development in children with permanent SNHL. So, much less is known about
literacy development in this population. Further research strictly focusing on the reading
and reading-related cognitive skills in this population is warranted.

* The data revealed little difference of language and reading abilities between the hearing
impaired and normally hearing children, but the hearing impaired children's phonological
skills were significantly inferior to children in the age-control group. The hearing impaired
children's success with reading in spite of depressed phonological processing skills was
noted.

* Thus, it is assumed that even though slight language deficiency (receptive/expressive
vocabulary) was observed, phonological depression linked with auditory limitation would
not impede reading skills of hearing impaired children. Furthermore, possible weakness in
parts of language areas (receptive vocabulary) does not seem to have significantly adverse
impact on later reading capacity in this population.

* Based on this, phonological strength would be regarded as a sufficient condition for normal
literacy, but not as a necessary condition for age-appropriate written language development
in hearing-impaired children.

* Children with less significant hearing loss or those with relatively more auditory access to
phonemes did not readily use phonological information for reading words in the same way
as hearing readers. This preferential use of the visual route was manifested by more
dependence upon sight word recognition strategy in children with mild or moderate
hearing loss. Further studies which can investigate hearing impaired children's use of sight
word use would reveal an interesting fact, which would be of clinical importance.

* To tap children's phonological memory skills (STM, WM, and LTM), a consistent, theory-
based theoretical framework should be used to categorize each memory task. Without this,
the interpretation of data could be misleading.

* A further recommendation is that the socio-economic status factor should be considered as
an influential factor for reading achievement.

* Finally, two issues can be mentioned regarding the research design. First, a longitudinal
study design with longer period of observation for reading development is quite warranted
to see the trend of literacy development in children with hearing loss. Second, children's
hearing levels (aided or unaided) should be representative of the population to be studied.
That is, the possible group difference associated with hearing level effect can be obscured
when the distribution of hearing loss is skewed.













Short-term coding (representation) of phonological information


Retrieval of phonological information stored in long-term memory


Phonological awareness (syllable, intra-syllable, phoneme)


Figure 2-1. Three different areas of phonological processing (Gillon, 2004).


Figure 2-2. Phonological loop (Baddeley, 1986)


Processing system
* phonological
awareness
* verbal rehearsal
* Pattern matching
(lexical retrieval)












Phonological form
(phonological long-term memory)


Sub-lexical phonological
representation
(words or non-words)

Short-term
retention
memory


2


Acoustic
representation 4




Speech [C-A-T] I


Working memory
4 (can process both words and nonwords)
PROCESSING CAPACITY
STORAGE (phonological awareness)
CAPACITY 1. meta-phonological
2. manipulation tasks


Processed as speech signal
*. . . .. . . .


: Perceived as Non-speech signal
.q..............................................


Figure 2-3. Word recognition through phonological route (revised from Ramus, 2001)


3
400O


Nsk,













Permanent lexical storage


Phonological form


t
Due to insufficient
phonological cue, access to
phonological information
will be hampered.


...............................................
Lexical retrieval I
.................... I....... .....


Weak representation in short-
term phonological buffer

Short term memory


Semantic content


Underspecified representation would
hinder long retention --
Phonemic blending needed for
nonword repetition or word decoding
would fail.

Working memory


(real word)
(nonsense word)
Underspecified representation


Figure 2-4. Model of (underspecified) phonological representation


Ultimate activation of semantic
content will be delayed (slow
word recognition and reading
comprehension)


Baddeley 's working memory


I


I











Phonological representation of words


Speech perception
(recognition)
> Phoneme perception
> Lexical decision
> Speech gating
Input representation deficit


Speech production
> Rapid Automatized Naming (slower activation of
lexical information in lexical retrieval
> Non-word repetition (STM)
> Less distinctness of vowel production
Post-input processing deficit


Figure 2-5. Two different aspects of phonological representations (Goswami, 2000).











Attention activation *


Visual processing
*PSR
(shape of letter 'B')


Corresponding lexical information *PSR
- phonological representation
- orthographical representation


Integration processes *PSR



Lexical process *PSR
[Access to and retrieval of]
phonological code
semantic code


Motor planning
*PSR



Articulation (naming)
*PSR


* PSR= processing
speed requirement


Figure 2-6. Simplified model of visual naming (Wolf et al, 2000:394)









CHAPTER 3
METHODS AND MATERIALS

Introduction

Despite the substantial number of empirical studies on phonological development and

phonology-based reading skills in children with deafness, little is known about hard-of-hearing

children's skills in similar areas. Most previous studies concur that hearing impaired

children's phonological processing skills could be adversely affected during the critical period

of language development possibly due to decreased acuity of peripheral audition related to

congenital hearing loss. However, impaired phonological processing skill has not been

reported to negatively impact reading and related cognitive skills (Briscoe et al., 2001; Gibbs,

2004). This study is designed to further explore relationship between phonological processes,

spoken language, and reading.

This cross-sectional study addressed the following primary research question: What are

the effects of mild to moderate sensorineural hearing loss on the phonological processing

skills, oral language, and reading ability? It was hypothesized that hearing loss would affect

phonological awareness, rapid naming, verbal memory, vocabulary and grammar, and reading

skills (reading fluency and comprehension). To investigate this question, the investigator used

a theoretical framework of 'phonological core deficit hypothesis,' established in the study of

developmental reading disorders.

A wide range of tests in the areas of reading and spelling, oral language, phonological

processing skills, auditory processing ability, and basic auditory skills was used to accomplish

this objective. Reading skills were measured for word reading under timed and untimed

conditions, spelling, oral reading fluency, and reading comprehension. Also, a range of

phonological processes were measured, including (1) phonological awareness (blending and









deletion), (2) phonological short-term memory (digit span, nonword repetition), and (3) rapid

naming skills (digits and letters). Further, auditory processing skills were investigated using a

subset of tests from a screening test of central auditory processing ability. Basic auditory skills

were checked by obtaining thresholds for stimuli consisting of pure-tones and speech signals.

The research methods presented in this section are addressed under the headings of

recruitment setting, participants and selection criteria, procedures, treatment of the data, and

research questions.

Setting and Participants

The purpose of this section is to describe the instructional settings where the study took

place and to provide a description of the participants. Demographic and audiologic

characteristics of the participants are provided for three groups of subjects.

Recruitment Setting

Three groups of children between 7 to 12 years of age, were recruited for the study: (1) 19

children with mild to moderate hearing loss measured by their better ear's pure-tone average

(PTA); (2) 29 children with normal hearing and reading skills; and (3) 30 dyslexic children with

normal hearing ability. The University of Florida Institutional Review Board (IRB-02) approval

(UFIRB 2006-U-0535) was received to recruit participants and conduct the research. Data were

collected from October 2006 through October, 2007. For the group of children with hearing loss,

2,500 flyers and recruiting letters containing information about the study and contact information

were sent out to public elementary schools located in the following ten counties in north central

Florida region: Alachua, Clay, Duval, Gilchrist, Lake, Marion, Orange, Putman, St. Johns, and

Volusia Counties. Around fifty parents or guardians of hearing impaired students expressed

interest in the study through reply letters, emails, or phone calls. All students were enrolled in

public elementary schools in the central Florida area. Participants with hearing loss were









recruited from several sources, including contracts with the Exceptional Student Education

department (ESE) of each county, advertisements posted at hospitals and several churches

located in Gainesville, FL., and personal acquaintances. For the hearing impaired subjects,

directors of ESE department were contacted to help distribute recruiting letters to schools which

had special program for hard-of-hearing children. Letters were sent to the families through

classroom teachers. Of the 69 (2.8%) children who returned permission forms or expressed their

interests, 50 (70%) were qualified to participate in the study. All participants with normal

hearing and reading ability were recruited from twelve elementary schools located in Gainesville,

FL through the department of research at the Alachua County School Board.

Selection Criteria

Potential participants were required to satisfy the following criteria.

* Each participant must be a native speaker of American English.

* Each participant's non-verbal intelligence screened by the Test of Nonverbal Intelligence-3
(TONI-3, Brown L., Sherbenou R. J., & Johnsen S. K., 1997) must be equal to or more
than 80.

* All participants were required to have normal tympanograms and no signs of middle ear
infection at the time of testing.

* All parents of participating children were required to read and sign the informed consent
letter.

* Children's hearing loss must be binaural sensorineural and not associated with any other
sensory impairments, neurological and/or neurodevelopmental disturbances. When
potential subjects were identified, children with monaural SNHL and/or cochlear implant
devices were excluded.

* Hearing loss must be congenital and the participants must wear hearing aids for both ears
at the time of testing. Exception was allowed only when the hearing loss was very mild (26
dB HL to 30 dB HL) and the participant's word recognition scores for both ears were
100% at the most comfortable level (MCL).

* All hearing-impaired participants were required to attend a mainstream school and use
speech mode as their primary communication mode.









Participants

Parental consent was obtained for each child participant at the onset of the investigation.

When a family agreed to have their child participate in the study, the child's parent or guardian

provided informed consent as required by the course of the study and a completed questionnaire

(Appendix A). The information in this form was used to check the subject's language and

reading ability, the status of hearing loss, age at identification of hearing loss, age at initial

audiologic intervention, duration and frequency of speech language services, developmental

change of patterns or types of hearing loss, family history of hearing loss, and any known

etiologic cues. Children were not compensated for their time. Instead, the participants were

provided a letter of test scores within three weeks.

The participants in the present study initially consisted of two groups of students between

7 and 12 years of age: (1) A sample of 29 normally developing children, and (2) a sample of 19

children with mild to moderate binaural SNHL. Ethnicity representation for the students in the

HI and the NH groups was predominantly Caucasian (n=38), followed by Asian American (n=5),

Hispanic (n=3), and African-American (n=2). After the data collection of these two groups was

over, an additional archival data set of thirty dyslexic children was incorporated for the purpose

of comparison and formed a third experimental group.

Participants with normal hearing and reading skills (NH group)

For the control group of children with normal hearing ability and reading skills, thirty eight

potential participants responded. After removing nine students who did not satisfied all inclusive

and exclusive criteria for the following reasons, the group was reduced to twenty nine (boys= 16,

girls=13). None of the children participating in the control group were known to have histories of

speech, language, or hearing problems, or any type of exceptional educational needs.









* Three students had bilingual background (one for Spanish and two for Korean language)

* According to the questionnaire form, four children were reported to have developmental
reading difficulties. One of them actually completed the protocol for the study, but due to
this problem, the data could not be included in the final data.

* Two students who also completed the protocol were excluded from the data since their
ages were either lower or higher than the suggested research plan.

The mean age (in months) and mean grade (in years and months) of this NH group were 111;9

(SD= 14.0) and 3.8 (SD=1.15), respectively.

Participants with hearing impairment (HI group)

Thirty-one children with prelinguistic, sensorineural hearing loss were contacted. The

children's parents or guardians were asked if their children had been or were currently receiving

speech or language therapy. Only four participants who enrolled in the study were not receiving

speech and language therapy. All subjects in HI group had binaural sensorineural hearing with

better ear's hearing thresholds in the mild to moderate range (26 to 70 dB HL) as assessed by the

pure-tone average (PTA). All children wore ear-level hearing aids, with only one subject aided

monaurally.

Two categories of hearing loss were identified in this sample, on the basis of the child's

PTA in the better ear. Mild hearing loss defined as having a PTA of 20 40dB HL and moderate

hearing loss was defined as a PTA of 41-70 dB HL. Audiometric assessments revealed that 6

children had mild hearing loss, whereas 13 children met the criteria for moderate hearing loss.

Information about hearing loss, audiologic services received, and clinical histories were

determined through interviews with at least one parent or guardian and/or a parental

questionnaire form provided in the recruiting letter packet. Among these thirty-one potential

participants, twelve children who did not satisfy the inclusive and exclusive criteria were

excluded from the study.









* Four children were reported to have only monaural hearing loss.

* Three children had bilingual background and they were all Spanish-speaking participants.

* Three subjects were reported to wear cochlear implants for at least one ear.

* Two children completed the protocol for the study; however, their data were excluded from
final data since their hearing losses were found to have developed after four and three
years of age, respectively.

Nineteen hearing impaired children included in the final selection (mean age [in months] =

110.8, SD = 19.3, mean grade [in years and months] = 3.4, SD = 1.61, boys = 11, girls = 8) were

designated as the experimental group (HI group). They were all enrolled in public elementary

schools that provided oral programming for hard-of-hearing students and emphasized use of their

residual audition for the development of speech and language skills.

Dyslexic group (RD group)

To enable a rigorous investigation of the strengths and the weaknesses of the HI group's

performance on phonological processes and literacy skills, the investigator decided to add a third

group of children with specific reading disorder who have no auditory perceptual limitations due

to cochlear damage. That is, the HI children performances on phonological processing and

literacy were compared to those of age and grade-level matched dyslexic controls of similar

nonverbal ability.

Many dyslexic children have a history of language difficulty (Rutter and Yule 1975) and

dyslexia is conceptualized either as a mild form of language impairment, affecting only the

phonological system, or as a residual problem that remains when oral language difficulties have

resolved (Aram, Ekelman, and Nation 1984; Scarborough and Dobrich, 1990).

Based on the fact that the auditory system is crucial for the development of language, many

researchers have suggested that for at least some of the children with phonologic dyslexia, there

may be a disorder within the auditory system that has disrupted normal acquisition of language.









However, unlike hearing impaired children, the disruption is not occurring at the periphery, but

at certain point in the ascending auditory pathway or the cortical level, through intrahemispheric,

interhemispheric or associative connections (Moncrieff, 2002). There is evidence to indicate that

dyslexic readers have abnormalities within some of the auditory structures necessary for

language development, including symmetry differences of the planum temporale (Hynd,

Semrud-Clikemand, Lorys, Novey, and Eliopulos Hynd, 1990; Kushch, Gross-Glenn, Jallad,

Lubs, Rabin, Feldman, and Duara, 1993; Larsen, Hoien, Lundberg, and Odegaard, 1990;

Leonard, et al. 1993), abnormal portions of the corpus callosum (Duara, Kushch, Gross-Gleen,

Barker, Jallad, Pascal, Loewenstein, Sheldon, Rabin, Levin, Lubs, 1991; Hynd, et al. 1995), and

Heschl's gyrus in the right hemisphere (Leonard, et al. 1998; Musiek & Reeves 1990; Penhune,

Zatorre, MacDonald, and Evens, 1996).

Therefore, these children with specific reading disorders would serve as an informative

comparison group when investigating the impact of congenital peripheral hearing loss on the

development of phonological processing and related literacy skills in comparison to the one

associated with cortical or central disruption of auditory processing. Currently, there is no in-

depth study that has looked at the different characteristics of hearing impaired and dyslexic

children's performances on phonology and reading measures.

Thirty dyslexic children's data were selected from an archival data set that has been

collected from 1999 to 2003 from the Dyslexia Clinic Program at the University of Florida

Speech and Hearing Clinic (mean age[in months]= 116.6, SD = 20.3, grade[in years and

months] = 3.96, SD = 1.62, boys = 18, girls = 12 ). All subjects were diagnosed with dyslexia, a

specific reading disability. Some parts of variables were not available for this RD group for the

following variables: (1) basic audiometry (puretone and speech), (2) auditory processing









measures, and (3) oral language measures. However, normal hearing abilities were confirmed for

all dyslexic participants through hearing screen on the day of testing by a certified audiologist in

UFSHC or by parent report of normal hearing based on previous testing.

Matching Variables

Normally developing (NH) and dyslexic subjects (RD) were selected to match age, grade,

and sex of the hearing impaired subjects (HI). Statistical comparisons revealed that these

variables were well matched on chronological age, grade, and gender for the groups (Age:

F[2,75] = .778, p = .463; Gender: F[2,75] = .011, p = .989; Grade: F[2, 75] = .818, p = .445).

A non-verbal IQ measure was selected as a covariate in data analyses to reduce any

existing effect of intelligence on reading skills. This test was only administered to children in the

NH and HI groups. A criterion nonverbal IQ of 80 as assessed by the TONI-3 was required.

Controlling for nonverbal IQ was deemed necessary because a univariate one-way ANOVA

confirmed a significant group effect (F[1,45] = 3.406, p = .042). Therefore, controlling for this

nonverbal IQ measure was justified (See Table 3-1).

Procedure

The experimental tasks were completed in two separate sessions in a quiet room in the

UFSHC and the speech hearing clinic in the University of Central Florida. In the first session,

which lasted for about 90 minutes, general information about the test and the test procedure was

explained to students and their parents or guardians. Upon successful confirmation of eligibility,

each participant signed a written consent for the data to be used for research purposes and

completed all audiologic testing. In a second session that followed within two weeks, tests were

administered in the areas of phonology, oral language, and reading for about 120 minutes.









Because of decreased speech intelligibility, hearing impaired children's responses on all of

phonological processing tasks were recorded digitally using Olympus digital voice recorder

(Model No. DS-40) and phonetically transcribed by the author for scoring.

The author and two other research assistants administered tests to all participants. Two

research assistants were senior undergraduate students in speech-language pathology. Both

assistants were trained by the investigator in test administration. For reading and language tests,

the author provided specialized instruction to the assistants and 50% of training sessions were

supervised by a certified speech pathologist. To ensure reliable test administration and scoring,

whenever any deviation from the protocol occurred, additional instruction was provided until

that assistant was able to demonstrate complete compliance with the testing protocols.

All audiologic tests were conducted by three doctoral students enrolled in the Doctor of

Audiology program (AuD) of UF and one certified audiologist from UCF. All tests were

administered in a soundproof suite using recorded material.

Materials

Each NH and HI participant was given tests individually in four separate domains,

including (1) auditory function, (2) phonological processes, (3) oral language, and (4) reading

and spelling from September 2006 through October 2007. However, RD subjects' data on

auditory function and oral language were not available from the archival database. Table 3-2 is a

list of all the tests that were used. Descriptions of the test instruments are presented in detail in

the next section.

Audiologic Measures

All subjects were examined otoscopically to rule out the presence of fluid in the middle-

ear-cleft. Each listener was tested in one 60-min session and received all tests in the above listed

sequence. Both left and right ears were tested to determine the better ear. The audiologic test









battery consisted of (a) a pure-tone threshold test at 250, 500, 1000, 2000, 3000, 4000, and 8000

Hz, (b) a speech recognition threshold test, (c) word recognition test, (d) two central auditory

processing tests, and (d) tympanometry.

Puretone and speech audiometry

The hearing acuity of the children was assessed in a conventional manner since all of

participants were 7 years or older. The air-conducted thresholds were examined in a sound proof

booth at 0.24 8 kHz with a GSI 61 clinical audiometer calibrated according to the ISO-389

standards (1985), employing insert earphones (E-A-R-TONE' 3A Insert Earphone) and TDH-39

headphones.

A standard audiometric staircase procedure (5-dB step size; down 10 dB, up 5 dB rule)

was used to obtain pure-tone air conduction thresholds. The intensity recorded as threshold was

the lowest intensity at which two correct responses were given (response in the presence of a

stimulus tone) out of four presentations. Children responded to test stimuli by hand-raising. Bone

conduction thresholds were obtained in a similar manner if air conduction thresholds were 20 dB

HL at any frequency. PTA (pure-tone average at the frequencies of 0.5, 1 and 2 kHz in the better

ear; right ear if equal hearing) defined each participant's hearing threshold.

Speech reception thresholds and word recognition scores in percentile were measured via

live voice using CID W-22 spondaic word lists (Audited Revised Auditory Tests CD). For the NI

group, stimuli were provided at the most comfortable level (MCL) as determined in the speech

reception thresholds (SRTs). For hearing impaired children, unaided thresholds were obtained.

Listeners wore various hearing aids ranging from basic analog to high performance digital

technology.

The subjects in the NH group had pure-tone thresholds of 15 dB HL or better at all octave

frequencies from 250 to 8000 Hz with an exception of one male child, whose right ear pure-tone









threshold for 2 KHz was 21 dB HL. All hearing subjects also had excellent word-recognition

scores ranging from 880% to 100% for the CID W-22 word lists (M = 99.0, SD = 2.95). Hearing

impaired subjects' pure-tone thresholds at 0.25, 1, and 2 kHz ranged from 0 to 13 dB HL and

their word recognition scores measured at the MCL level were moderately good (Right ear: M =

75.2, SD = 29.6, Left ear: M = 86.0, SD = 21.9).

Auditory processing tests

Based on recommendations by Musiek & Chermak (1994), four commonly used central

auditory tests were administered to all participants in the NH and HI groups: (1) One subtest, the

Dichotic Digit test, was selected from the Central Auditory Processing Tests developed by Frank

Musiek, and (2) three subtests were selected from the SCAN-C/A screening test for auditory

processing disorders (SCAN-C for children aged 3 to 11 years, SCAN-A for subjects aged 12

and more). Musiek's (1983) Dichotic Digits test is a test of binaural integration in which pairs of

digits are delivered simultaneously to each ear at the MCL, with each ear receiving a different

digit pair. This test has good validity data and is simple and quick to administer and to score.

Subjects were given three practice items to ensure their understanding of the task. All stimulus

items consisted of monosyllabic digits from 1 to 10 (excluding 7) spoken by a male speaker of

General American English. Stimuli were routed from a CD player through a two-channel

audiometer meeting ANSI (1996) standards and were delivered via TDH-49 headphones at an

intensity of 40 dB SRT. Twenty stimulus items consisting of four digits each were presented.

The test was administered following standard clinical recommendations (Musiek, 1983) using a

free-recall paradigm. Performance was scored as a function of percent correct for each ear. The

minimum subject age for this test is 7 years (Musiek, 1983). Administration time was

approximately 5 to 6 minutes.









The SCAN is a screening test for auditory processing disorders (Keith, 1986). It may be

used to identify potential factors related to poor social skills, language use, and academic

performance in children from 3 to 11 years of age. This test was developed to determine possible

central nervous system dysfunction by assessing auditory maturation, (b) identify children at risk

for auditory-processing or receptive language problems who may require additional audiological

or language testing, and (c) identify children who may benefit from specific management

strategies (Keith, Rudy, Donahue, & Katbamna, 1989).

For this study, first three SCAN subtests were given in the order stipulated by standard test

format (Filtered Words ([FW], Auditory Figure Ground [AFG], Competing Words [CW]). In the

Filtered Words subtest, the subject is asked to repeat words that sound muffled. The test stimuli

consist of one syllable words that have been low-pass filtered at 500 Hz. Two practice words and

20 test words are presented to each ear. In the Auditory Figure-Ground subtest, the subject's

ability to understand words in the presence of background noise is measured. One syllable words

were recorded in the presence of multi-talker speech babble noise at the 0 dB signal-to-noise

(S/N) ratio. Two practice words and 20 test words were presented to each ear. In the Competing

Words subtest, the subject hears two words simultaneously--one word presented to each ear. The

test stimuli consist of one syllable word pairs presented to the right and left ears. First, two

practice word pairs and 15 word pairs are presented. The subject is instructed to repeat the words

presented in each ear, repeating the word heard in the right ear first. Then, a second set of two

practice word pairs and 15 word pairs are presented, with the subject repeating the word heard in

the left ear first.

Middle ear function (tympanometry)

All participants were required to pass a tympanometry screening (i.e., type A with equal to

or more than 0.2 ml compliance) to ensure normal eardrum and middle ear functions. Acoustic









immittance measures used to determine middle ear function were obtained by using a

commercially available middle ear analyzer (Grason Stadler, Model GSI-33). Tympanograms

were obtained in both ears on all participants during the first session. Any tympanogram for

which tympanometric width could not be calculated (i.e., no measurable peak) resulted in a

rescheduling of the participant for testing at a later date. Children were considered to have

normal middle ear function when their tympanometric width was 250 daPa in both ears (criterion

based on Nozza, Bluestone, Kardatzke, & Bachman, 1994). Only one normally hearing

participant had middle ear infection at the time of testing. One month later, this subject was

retested after the infection had been treated.

Literacy (Reading and Spelling)

All subjects in the three groups (NH, HI, and RD) were administered exactly the same set

of tests in the areas of phonological processing and literacy (reading and spelling). To test

reading and spelling, eight standardized subtests from four published tests were used. To test

phonological processing skills, six standardized subtests were taken from one published test. A

description of each test is provided below.

Woodcock Reading Mastery Test Revised (WRMT-R; Woodcock, 1987): Three

subtests of the WRMT were administered to assess: (1) untimed phonemic decoding skills for

real words and nonwords, and (2) passage reading comprehension. The Word Identification and

Word Attack subtests comprise 106 and 45 pronounceable real and pseudo words of increasing

complexity, respectively. Especially, in the Word Attack subtest, five unfamiliar words are presented

at a time on one page, and the examinee is asked to read them aloud as accurately as possible.

Pronunciation rules are provided for the examiner to determine the accuracy of the child's responses. This

test is discontinued when six errors are made. The Word Attack subtest evaluates the child's ability to

use phonic skills to determine the correct pronunciation of unfamiliar words while reading aloud









letter combinations that form nonsense words. The Passage Comprehension subtest requires

participants to read a segment of prose with a missing word and say words) to fill in the blank(s)

in the printed paragraph.

The WRMT-R test record allows for raw scores to be converted to age equivalent scores,

grade equivalent scores, and standard scores (M= 100, SD = 15). The WRMT-R was selected for

two reasons: (a) to measure word and nonword reading normed on the same sample, and (b) to

use large numbers of test items in order to reduce the likelihood of idiosyncratic word knowledge

causing lack of reliability (Olson, Forsberg, Wise, & Rack, 1994). The WRMT-R has more items

than other tests with a similar format. Content and concurrent validity are well documented in the

test manual (Woodcock & Johnson, 1989). Internal reliability coefficients of the WRMT-R

calculated based on split half reliability for 1st through 3rd grade ranged from .91 to .98 (M

= .94; Woodcock, 1987).

Wide Range Achievement Test-3 (WRAT-3; Justas & Wilkensen, 1993): The WRAT-3

includes three subtests that measure reading, spelling, and arithmetic skills. The WRAT-3 is an

academic achievement test that has been shown to have good correlation with the Wechsler

Individual Achievement Test. For the purpose of this study, only the Spelling subtest was used to

assess the ability to spell single words. Children were asked to write single words on test form

after listening to the target word followed by a sample sentence. Children are asked to try as hard

as they can to spell every word. For each item, target word is spoken first in isolation and then in

a sentence in which the word is stressed. Finally, the word is spoken again. Standard scores

(M= 100, SD = 15) are provided for 32 age groups. Internal consistency reliability figures in the

range of r = .86 to .91 are reported for children ages 7 to 13 years. The inclusion of a spelling









measure is based on the strong association between early spelling ability, phonological

awareness, and beginning reading (Ehri & Wilce, 1987).

Test of Word Reading Efficiency (TOWRE; Wagner, Torgesen, & Rashotte, 1999): The

TOWRE was given to measure ability to pronounce both sight words (Sight Word Efficiency

subtest) and nonwords (Phonemic Decoding Efficiency subtest). In the Sight Word Efficiency

subtest, 104 context-free single words of increasing complexity in terms of phonological

structure are given. Participants were asked to read aloud as many words as possible in 45

seconds. Similarly, the Phonemic Decoding Efficiency subtest of the TOWRE was administered

to quantify rapid nonword decoding skill. The subtest presents 63 pronounceable pseudo-words

and participants were asked to read as many nonwords as possible in 45 seconds.

Gray Oral Reading Tests-4 (GORT-4; Wiederholt & Bryant, 2001): The GORT-4, a

measure of reading fluency (accuracy and rate) was administered individually to each participant

to test reading fluency only. Both the child and the examiner were audio-recorded with an

Olympus DS-40 Digital Voice Recorder for more accurate scoring procedure. Especially,

recording was necessitated especially because of the decreased speech intelligibility of hearing

impaired children in fast passage reading task.

The GORT-4 is designed for children aged 7;0 to 18;11 (in months). It consists of 13

passages. Each passage has one paragraph that is centered on a single topic. Across the test, there

is an increase of length and complexity of paragraph, sentence, grammatical structures, and

vocabulary content. The test yields raw scores, standard scores, percentiles, and grade-equivalent

scores. The fluency assessment is a composite score of two components: a Rate (i.e., the time

taken to read each passage) and an Accuracy (number of deviations from print). The mean for the









two subtest components is 10, with a standard deviation of 3. They were instructed to read each

story aloud as quickly and accurately as possible.

Phonological Processing Skills

Comprehensive Test of Phonological Processing (CTOPP; Wagner, Torgesen, &

Rashotte, 1999). Phonological coding is an oral language skill and consists of the analysis and

synthesis of phonemes converted from visual stimuli of letters. Beginning readers who have

deficits in phonological coding seem to have difficulty naming letters of the alphabet, identifying

sounds for alphabet letters, segmenting words into phonemes and syllables, and applying

knowledge of letter-sound correspondence to decode words (Vellutino, et al., 1996). For

example, phonological coding involves analysis such as recognizing that the first sound of the

word ball (/b/) can be replaced with /t/ to produce the word tall. Phonological coding abilities

associated with this process of changing 'ball' to 'tall' include: (1) letter-sound correspondence,

(2) phonemic awareness and segmentation, and (3) working with information in phonological

working memory. For this study, six subtests of the CTOPP were used, including the Elision,

Blending Words, Memory for Digits, Nonword Repetition, Rapid Letter Naming, and Rapid Digit

Naming subtests (Form A).

The CTOPP is an individually administered, norm-referenced measure that is used to

evaluate a wide range of aspects of an individual's phonological processing. A three-part model,

based on earlier studies in this area has been presented by the test developers (Torgesen &

Wagner, 1998; Wagner & Torgesen, 1987). That is, three pairs of scores were combined to

produce composite scores: Elision and Blending Words for Phonological Awareness composite,

Memory for Digits and Nonword Repetition for Phonological memory composite, and Rapid

Letter Naming and Rapid Digit Naming for Rapid Naming composite scores, respectively. The

following is a short description of each of these components.









* Phonological awareness: analysis and synthesis of the sound structure of oral language.
The order of progression of phonological awareness starts with syllables and moves toward
smaller units of speech sounds (Adams, 1990). Phonological awareness provides
individuals with the ability to break words into syllables and component phonemes, to
synthesize words from discrete sounds, and to learn about the distinctive features of words
(Torgesen & Wagner, 1998).

* Phonological memory: coding information phonologically for temporary storage in
working or short-term memory. Phonological short-term memory involves storing distinct
phonological features for short periods of time to be "read off" in the process of applying
the alphabetic principle to word identification.

* Rapid naming: efficient retrieval of a series of names of objects, colors, digits, or letters
from long-term memory. Rapid naming of verbal material is a measure of the fluid access
to verbal names, in isolation or as part of a series, and related efficiency in activating name
codes from memory (Wagner, Torgesen, & Rashotte, 1999).

The Elision subtest required participants to delete a phoneme in either the initial, final, or

medial position from a real word and then produce a new word (e.g., "Say tiger without saying

/g/." [tire]). The Blending subtest required isolated syllables and phonemes to be combined into a

word (e.g., "What word do these sounds make: haemm/ /er/?" [hammer]). In the Rapid Letter and


Digit Naming tests, children are shown a visual display of randomly presented items and asked to

name them in sequence as quickly as they can. The Rapid Digit Naming test uses numbers 2, 3, 4,

5, 7, and 8. Seventy-two numerals are presented on two pages in four rows, with a space after

each. The time in seconds to name the 72 items in the display is recorded. Standardized scores

are provided for individuals ages 5 to 24 years on a scale ranging from 0 to 20, with an average

of 10.

The authors were able to establish criterion predictive validity with a sample that

represented ethnic, gender, and age variations. Reliability of the CTOPP was investigated using

estimates of content sampling, time sampling, and scorer differences. Most of the average

internal consistency or alternate forms reliability coefficients exceed .80 according to the test

manual. The test/retest (time sampling) coefficients range from .70 to .92. It is known that the









magnitude of the coefficients reported from all the reliability studies suggests that there is limited

error in the CTOPP and that examiners can have confidence in the results. Reliability

coefficients for the rapid-naming subtests are r = .82 for letter and r = .87 for digits.

Standardized Oral Language Tests

The Peabody Picture Vocabulary Test-III (L. M. Dunn & Dunn, 1997) and Expressive

Vocabulary Test (Williams, 1997) were administered to assess lexical knowledge and word

retrieval. Both tests were administered according to the guidelines provided in the testing

manual. The EVT and the PPVT-III were standardized on the same population of 2,725

examinees ranging in age from 2-6 to 90. All hard-of-hearing participants were wearing hearing

aids for both ears and given the same oral instruction as hearing subjects. Throughout the data

collection, no subjects had difficulty understanding the task.

Peabody Picture Vocabulary Test-III (PPVT-III, L. M. Dunn & Dunn, 1997): The

PPVT-III is a standardized test of receptive lexical knowledge. Each easel page contains four

pictures. Participants are required to choose the picture drawing from four choices on a page that

best depicts a word orally presented by the test administrator. The normative sample included

2725 persons. And while the original PPVT was standardized only on white children from

Tennessee, the normative sample of the PPVT-III was selected to match the data of the 1994 US

Census. The sample was stratified with each age group by gender, race/ethnicity, geographic

region, and SES. Only individuals who were determined to speak and understand English were

included in the testing. The alpha reliabilities for the 25 standardized age groups were between

.92 and .98 with a median reliability of .95 for both forms A and B. The split-half reliabilities for

the 25 age groups ranged from .86 to .97, with a median of .94 for both forms. The alternate

forms reliabilities range from .88 to .96 with a median correlation of .94.









The Expressive Vocabulary Test (EVT; Williams, 1997): The EVT is a test of expressive

vocabulary (lexical knowledge and word retrieval) requiring picture labeling and synonym tasks.

For the 38 items, the tester points to a picture and asks a question. On the 152 synonym items,

the examiner presents a picture and stimulus words) within a carrier phrase. The examinee

responds to each item with a one-word answer. All stimulus pictures are in full color, carefully

balanced for gender and ethnic representation. The EVT reliability analyses indicate a high

degree of internal consistency. Split-half reliabilities range from .83 to .97 with a median of .91.

Alphas range from .90 to .98 with a median of .95. And test-retest studies with four separate age

samples resulted in reliability coefficients ranging from .77 to .90, indicating a strong degree of

test score stability.

The Comprehensive Assessment of Spoken Language (CASL, Carrrow-Woolfolk,

1999) was used to measure the knowledge of grammatical knowledge. The CASL is an

individually and orally administered oral language assessment battery for ages 3 through 21. This

test measures four main areas of oral language such as (a) lexical/ semantic, (b) syntactic, (c)

supralinguistic, and (d) pragmatic skills. In the Syntactic Tests, five subtests are included: Syntax

Construction, Paragraph Comprehension, Grammatical Morphemes, Sentence Comprehension,

and Grammaticality Judgment. For this study, the Grammatical Morphemes subtest was selected

to investigate hearing impaired subjects' sensitivity to morphological markers and syntactic

constituents for major grammatical information, including tense, voice, number, modality,

person, pronoun, comparative, and lexico-conceptual knowledge. The examiner read one pair of

words or phrases that demonstrated an analogy, then read the first word or phrase of a second

pair. The examinee must complete the analogy of the second pair (e.g., Skate is to skated, as talk

is to .).









The rationale for selecting this subtest is based on some previous theoretical suggestion

regarding phonetic salience of grammatical morphemes in English language. For example,

Leonard (1998) suggested that many inflections in English have low phonetic salience and that

this factor in combination with reduced speed of processing underlies the problems with

inflectional morphology described in SLI (Surface hypothesis). Even though the surface feature

of phonetic salience alone cannot account for the difference in difficulty (Hayiou-Thomas,

Bishop, & Plunkett, 2004; Rice & Wexler, 1996), it is well assumed that peripherally depressed

hearing acuity would be a negative factor for efficient and normal perception of inflectional

markers which have low phonetic salience. Thus, we hypothesized that hearing-impaired

children's performance on this specific test would show significant difference from that of

normally hearing subjects.

Interrater Reliability

To ensure test reliability, whenever any deviation from the protocol of test administration

occurred, additional instruction was provided until that assistant was able to demonstrate

stabilized skill. The test results and the background data were fed into SPSS for Windows

package 15.0. The first author and two research assistants, who assisted in the test administration,

score conversion and coding, met regularly to discuss any problematic matters. Interrater

reliability procedures were conducted to determine the reliability of scoring and coding

procedures.

To determine interrater reliability at the end of data collection period, a trained reliability

coder, a doctoral student of speech pathology, checked on all scores (raw and standard scores)

and coded data of 30 randomly selected participants (38% of the whole subjects). This coder

independently obtained children's raw scores on all tests, converted them to standard and

percentile scores, and checked the coded numbers in the SPSS. The reliability coder and the test









administrator were blind to each other's scoring. The coder and the test administrator conducted

an item-by-item comparison of their respective responses to each item administered in the

battery. A reliability score was calculated for all variables by dividing the number of agreements

by the number of disagreements plus agreements and multiplying this score by 100. Interrater

analysis showed 99% agreement for data coding (disagreement on 18 coding errors out of 2,100

numeric codes) and 97% agreement for score conversion (disagreement on 12 out of 420

standard scores).

Research Questions and Hypotheses

The following three main categories of research questions were investigated.

Category I (Group Effect)

Question: Are there any significant group effects on phonology, oral language, and reading

skills?

Hypotheses: It was hypothesized that children with mild to moderately severe SNHL would

demonstrate significantly lower performance than the comparison groups (NH and RD) on the

measures of (a) auditory processing skills, (b) phonological processing (phonological awareness,

verbal memory, and rapid naming), (c) oral language (receptive and expressive vocabulary and

grammar) and (d) reading (word/nonword reading, passage reading fluency, and

comprehension).

Category II (Relationships among Measures)

Question: What are the interrelationship among reading achievement, hearing ability, phonology,

oral language skills, and auditory processing skills? Which phonological processes are the most

strongly correlated with reading skills?

Hypotheses: It was hypothesized that there would be significant association among variables of

phonological processing, oral language, and reading skills of hearing impaired subjects. This









hypothesis was explored primarily using correlational analyses using a matrix of Pearson

product-moment correlation coefficients between the variables.

Category III (Regression Question)

Question: What are the contributions of phonological and auditory processing skills in

predicting hard-of-hearing children's reading achievement?

(1) Specifically, which measure of phonological processing skills is the best predictor of
reading skills?
(2) How much of unique variance of reading performance (reading comprehension, reading
fluency, word/nonword reading) is explained by phonological processing measures?
A full list of specific null hypotheses for the research questions to be tested in the study is

presented in Table 3-3.

Treatment of the Data

All variables were entered into Microsoft Excel spreadsheets and analyzed with the SPSS

16.0. Descriptive statistics were calculated for all variables: demographic, audiologic, auditory

processing, oral language, phonological processing, and reading/spelling skills.

Three statistical methods were used to analyze the data. First, to measure group effect, a

series of multivariate analysis of covariance (MANCOVA) tests was carried out on the scores of

auditory processing, phonological processing, and reading skills of the three groups (NH, HI, and

RD) to investigate any significant group effect (Category I). A set of MANCOVAs was also

conducted to compare oral language skills of children with normal and impaired hearing ability

(NH and HI only). The grade scores (in months) and/or the non-verbal intelligence score (TONI-

3) served as the covariates or control variables.

Secondly, to measure relationship among variables, a series of partial correlation analyses

was conducted to determine significant relationships among variables.









Finally, to examine the unique variance explained selected exploratory variables, block

hierarchical regression analyses were performed to determine which predictor variables could

most effectively account for variance in the following dependent variables (4 word/nonword

reading measures; 1 reading fluency measures; and 1 reading comprehension). The predictor

variables examined were: (1) background information such as age, grade, and non-verbal IQ; (2)

audibility, as measured by the pure-tone average; (3) auditory processing, as measured by the

dichotic digits, and (4) six phonological processing measures from the CTOPP. For all

multivariate statistical tests, correlation analyses and regressions were conducted with an alpha

level set to .05.










Table 3-1. Matching variables (Grade, Age, Gender, and non-verbal intelligence)
IQ Gradea Ageb Gender
Mean Mean Mean Boys Girls
Min Med Max Min Max Min Max ______Boys
(SD) (SD) (SD) Freq % Freq %
112.52 3.8 111.9
NH 88 110 150 2.1 5.9 89 134 16 55.2 13 44.8
(15.844) (1.15) (14.0)
100.89 3.4 110.8
HI 84 110 138 1.0 6.9 82 152 11 57.9 8 42.1
(12.918) (1.61) (19.3)
3.96 116.6
RD No data 1.2 6.8 86 152 18 60.0 12 40.0
(1.62) (20.3)
107.92 49.1 113.4
Total 84 107.9 150 12 92 82 152 45 57.7 33 42.3
(84-150) (18.9) (17.9)


Note: a: in year.months, b: in months, Freq: frequency, %: percentile









Table 3-2. List of tests used.

Area of measurement Tests Contents
Non-verbal intelligence
TONI-3 Non-verbal IQ
Oral language
Receptive vocabulary PPVT-III Receptive vocabulary
Expressive vocabulary EVT Expressive vocabulary
Syntactic structure CASL
Grammatic knowledge
(Grammatic Knowledge subtest)


Phonological processing CTOPP)
Phonological awareness Elision, Blending
Phonological memory Memory for Digits,
Short-term memory
Nonword Repetition,
Lexical access and
Rapid naming RAN-Digit, RAN-Letter
retrieval
Literacy
Reading (Untimed) WRMT-R (Word Identification) Word Reading
WRMT-R (Word Attack) Nonword decoding
Reading (Timed) TOWRE (Word Efficiency) Word reading
TOWRE (Phonemic decoding Nonword decoding
efficiency)
Spelling WRAT-3 (Spelling) Spelling words
Reading comprehension WRMT-R (Passage Comprehension) Silent passage
comprehension
Reading fluency GORT-4 Passage reading
(Reading accuracy and rate) fluency

Abbreviations: TONI (Test of Nonverbal Intelligence), PPVT-III (Peabody Picture Vocabulary
Test), EVT (Expressive Vocabulary Test), CASL (Comprehensive Assessment of Spoken
Language),CTOPP (Comprehensive Test of Phonological Processing), WRMT-R (Woodcock
Reading Mastery Test Revised), TOWRE (Test of Word Reading Efficiency), WRAT-3 (
Wide Range Achievement Test), GORT-4 (Gray Oral Reading Tests).









Table 3-3. List of research hypotheses.
Statistical
analyses


MANCOVA


Correlation



Regression


Hypotheses


*


1. There will be statistically significant differences on a measure of receptive vocabulary (PPVT-3) among
students in the NH and HI groups.
2. There will be statistically significant differences on a measure of expressive vocabulary (EVT) among
students in the NH and HI groups.
3. There will be statistically significant differences on a measure of grammatical knowledge (CASL) among
students in the NH and HI groups.
4. There will be statistically significant differences on the measures of phonemic elision, deletion, rapid naming
(letter and digit), and phonological short-term memory among students in the NH, HI, and RD groups
(CTOPP).
5. There will be statistically significant differences in untimed word and nonword reading skills (WRMT-R)
among students in the NH, HI, and RD groups.
6. There will be statistically significant differences in timed word and nonword reading skills (TOWRE) among
students in the NH, HI, and RD groups.
7. There will be statistically significant differences in spelling skill (WRAT) among students in the NH, HI, and
RD groups.
8. There will be statistically significant differences in passage reading comprehension (WRMT-R) among
students in the NH, HI, and RD groups.
9. There will be statistically significant differences in passage reading accuracy and rate (GORT-4) among
students in the NH, HI, and RD groups.
1. There will be significant correlations between hearing-impaired students' depressed phonology and reading.
2. There will be significant correlations between hearing-impaired students' depressed phonology and hearing
loss.
3. There will be significant correlations between hearing-impaired students' depressed phonology and oral
languages.
4. There will be significant correlations between hearing-impaired students' auditory processing and reading.
5. There will be significant correlations between hearing-impaired students' auditory processing and phonology.
1. Phonological awareness will significantly predict unique variance in word-level reading after controlling for
cognitive ability, hearing acuity, and grade in both untimed and timed reading tasks.









Table 3-3. Continued
2. Phonological short-term memory will significantly predict unique variance in word-level reading after
controlling for cognitive ability, hearing acuity, and grade in both untimed and timed reading tasks.

3. Rapid naming will significantly predict unique variance in word-level reading after controlling for cognitive
ability, hearing acuity, and grade in both untimed and timed reading tasks.

4. Auditory processing will significantly predict unique variance in word-level reading after controlling for
cognitive ability, hearing acuity, and grade in both untimed and timed reading tasks.

5. Phonological awareness will significantly predict unique variance in passage reading fluency after controlling
for cognitive ability, hearing acuity, and grade in both untimed and timed reading tasks.

6. Phonological short-term memory will significantly predict unique variance in passage reading fluency after
controlling for cognitive ability, hearing acuity, and grade in both untimed and timed reading tasks.

7. Rapid naming will significantly predict unique variance in passage reading fluency after controlling for
cognitive ability, hearing acuity, and grade in both untimed and timed reading tasks.

8. Auditory processing will significantly predict unique variance in passage reading fluency after controlling for
cognitive ability, hearing acuity, and grade in both untimed and timed reading tasks.

9. Phonological awareness will significantly predict unique variance in spelling after controlling for cognitive
ability, hearing acuity, and grade in both untimed and timed reading tasks.

10. Phonological short-term memory will significantly predict unique variance in spelling after controlling for
cognitive ability, hearing acuity, and grade in both untimed and timed reading tasks.

11. Rapid naming will significantly predict unique variance in word-level in spelling after controlling for
cognitive ability, hearing acuity, and grade in both untimed and timed reading tasks.

12. Auditory processing will significantly predict unique variance in word-level in spelling after controlling for
cognitive ability, hearing acuity, and grade in both untimed and timed reading tasks.









CHAPTER 4
RESULTS

The primary goal of this study was to examine the effect of congenital mild to

moderate SNHL on children's auditory processing, phonological processing, oral

language, and literacy skills. The main focus of this investigation was twofold: (1) to

examine differences among normally developing, hearing impaired students, and dyslexic

readers in oral language, phonology, and reading skills (group effect) and (2) to

investigate interrelationships between reading skills and phonology, auditory processing,

and oral language abilities. The following questions were formulated and tested as

follows.

* Are there any significant group effects for phonology, oral language, and literacy
skills (reading/spelling) between the three groups?

* What interrelationships exist between reading and phonological skills, auditory
processing and oral language ability in hard-of-hearing children?

* Do phonological processing skills make significant contributions to the literacy
skills of children with hearing loss?

This chapter has four subsections. The first section presents results of the

descriptive statistics for all of the variables investigated in this study. The second section

deals with the group effect on oral language, phonology, audiologic function, and reading

using a set of multivariate analyses of covariance. The third section provides the results

of correlation analyses among variables based on the Pearson-product correlations.

Descriptive statistics and first-order bivariate and partial correlation coefficients were

used to estimate the children's performance in the measures and interrelationships among

their performance in different measures. The final section investigates the results of

multiple hierarchical regression analyses. The implications of these findings are

discussed in Chapter 5.









Descriptive Statistics

Preliminary analyses were conducted to obtain descriptive information on the

samples' demographic characteristics, audiologic ability, phonological process skills, oral

language skills, and reading skills. Results of the measures on phonology, language, and

reading are displayed in the next section where the outcomes of the MANCOVAs are

reported.

Demographic Data

The demographic characteristics of the participants in the three groups, including

age, grade, non-verbal IQ, and gender are provided in Chapter 3 (See Table 3-1).

Audiologic Ability Measures

Information related to the hard-of-hearing children's audiologic characteristics is

displayed in Table 4-1. Demographic information was collated and included age, grade,

gender, information about the history of audiologic invention (age at identification of

hearing loss age at initial hearing aid fitting and length of hearing aid fitting), etiology,

and basic audiometric results (unaided PTA, SRT, and WRS). A subject number was

randomly assigned to each participant to ensure patient confidentiality.

Table 4-2 is a descriptive summary of the HI and NH groups' performance on the

basic audiologic measures. The HI group's unaided PTA on the frequencies of 0,5 Hz,

1KHz, and 2 KHz ranged from 27 to 69 dB HL (mean better ear's PTA = 46.9 dB HL).

Table 4-3 reports mean PTA of the HI and NH groups for every tested frequency. Figures

4-1 through 4-3 are graphical representations of this information. Figure 4-1 and 4-2 are

error bar charts for each ear based on the standard errors of each mean threshold across

tested frequencies, showing that, on average, the HI group had a symmetrical sloping

binaural SNHL, with substantial variance within the group. Figure 4-3 is a histogram of









better ear's pure-tone threshold for the HI group. Table 4-4 displays the means and

standard deviations by hearing group (NH, HI) for the three subtests of the SCAN-A/C

and the Dichotic Digits subtest of the CAPD test (Musiek, 1983).

Oral Language, Phonology, and Reading

Descriptive data of oral language, phonological processing, and reading measures

are provided in the next section dealing with the MANCOVA questions.

Inferential Statistics

All scores used in the statistical analyses were norm-referenced standard scores.

Multivariate analyses of covariance (MANCOVA), Pearson-product correlation analyses,

and hierarchical block regression analyses were used to analyze the data to answer the

research questions raised in the previous chapter. The results are organized to provide

answers to three questions below:

1. What is the effect of decreased auditory sensitivity due to congenital SNHL on
phonology, oral language, and reading skills? That is, is there any significant
group effect in each measured area?

2. What is the relationship among hard-of-hearing readers' impaired phonology,
reading, oral language, and hearing loss

Question Category I: Group Comparisons of Language, Phonology, and Reading

To compare the performance of the HI with the two control groups (NH and RD),

multivariate analyses of covariance (MANCOVA) with grade (in months) and/or non-

verbal intelligence as covariates were used for oral language, phonology, and reading

measures. MANCOVA was selected in order to guard against Type I errors due to

multiple univariate testing for each area.

Three separate MANCOVAs were conducted: (a) For oral language measures,

three dependent variables from the PPVT-III (receptive vocabulary), EVT (expressive









vocabulary), and a subtest of Grammatical Knowledge from the CASL were analyzed

(NH and HI groups only), (b) for the phonological tests, six subtests from the CTOPP

were analyzed (NH, HI, and RD groups), and (c) thirdly, for the reading measures,

standard scores of timed word/nonword reading (TOWRE), untimed word/nonword

reading (WRMT-R), reading comprehension (WRMT-R), passage reading fluency

(GORT-4), and spelling (WRAT) tests were included in the MANCOVA to investigate

between-group difference on the literacy skills.

MANCOVA is appropriate for use with multiple dependent variables that differ in

scales of measurement and fewer Type I errors occur with it than with a univariate

analysis of variance (Gabriel & Hopkins, 1974). Other than maintaining overall Type I

error at a constant level, MANCOVA also allows us to control the systematic group

difference caused by covariates. Covariates are variables that are correlated with the

dependent variables and are used to adjust for any differences in the scores affected by

extraneous confounding factors. By using MANCOVAs, the effects of confounding

variables such as non-verbal IQ, age, or grade, can be statistically removed, helping to

ensure that the findings reflect true differences in phonology, oral language, and reading

skills.

For the first MANCOVA, children's non-verbal intelligence and grade were used

as covariate variables. For the second and third MANCOVAs, only grade (in months)

was used as a covariate variable because the TONI scores were not available in dyslexic

participants.

Hair et al. (1992) recommends that Wilk's lambda is the best statistical measure to

assess whether an overall significant difference is found between groups. All multivariate









statistical analyses were conducted with an alpha level set to .05. Good approximations

for significance can be obtained from Wilk's lambda and can be transformed into an F-

statistic. If any significant differences are identified in MANCOVAs, univariate analyses

of covariance (ANCOVAs) are justified to investigate the direction and significance of

specific dependent measures. ANCOVAs are run on the residuals after predictions of the

dependent variable have been made from a set of covariates. In this type of analysis, a

predicted dependent variable is computed. This predicted score is subtracted from the

original dependent variable score. This difference or residual score is used in the

ANCOVA (Meyer, 1993). Effect sizes for significant differences between groups,

adjusting for covariates, are calculated as the difference between the estimated adjusted

means for the two groups, divided by the root-mean-square error and reported as Cohen's

d. Cohen (1988) designated an effect size of 0.2 as small, 0.5 as medium, and 0.8 as large.

For all ANCOVAs showing significant group effects, pairwise comparisons of

groups are conducted. As a precaution against making Type I errors, thep value is

adjusted with the Bonferroni method for all multiple pairwise comparisons.

MANCOVA for oral language measures

Performance on the oral language tests for the two groups of children (NH and HI

groups) was compared to determine if differences in group performances were significant

for overall measures of expressive/receptive vocabulary and grammatical knowledge. A

one-way multivariate analysis of covariance (MANCOVA) was used with group

membership (NH vs. HI) as the independent variable, standard scores from the PPVT-III,

EVT, and CASL tests as the dependent variables, and grade (in months) and non-verbal

intelligence as the covariates.









Two assumptions of the MANCOVA were tests: equal covariance and equal

variance. MANCOVA assumes that for each group the covariance matrix is similar.

Box's M statistic was used and the covariance assumption was not violated (F[6, 9896.2])

= 4.356, p = .673). MANCOVA also assumes that each dependent variable will have

similar variances for all groups. According to the Levene's test, all three measures had

equal variances in the two groups (PPVT-III: F[(1,46] = .325, p = .571; EVT: F[1,46] =

2.179,p = .147; and CASL: F[1,46] = 2.850,p =.098).

Results of the initial multivariate test using Wilks' Lambda (A) criterion indicated

that the groups differed significantly in overall oral language measures (F[3,42] = 10.34,

p = .000, partial l2= .425, and observed power = .997). Because significant differences

between groups emerged for overall measures, it was necessary to examine specifically

which of the tests were influenced across the two groups. That is, each of the oral

language variables was examined individually through the use of three one-way

univariate ANCOVAs. These follow-up univariate analyses indicated that differences

between groups were significant for all language measures: (a) receptive vocabulary:

F[1,44] = 16.123, p = .000, partial f = .268, observed power = .975; (b) expressive

vocabulary: F[1,44] = 8.694, p = .005, partial r = .165, observed power = .822; and (c)

grammatical knowledge: F[1,44] = 28.354, p = .000, partial q2= .392, observed power

= .999.

Descriptive statistics for oral language measures along with estimated adjusted

means and standard errors are reported in Table 4-5. Table 4-6 is the ANOVA table for

these ANCOVAs displaying the result of univariate F-tests for between-group differences.









Table 4-7 is a summary of pairwise post-hoc comparisons. Finally, Figure 4-4 displays

clustered box plots of oral language measures of the groups (NH, HI).

Pairwise comparison revealed that when controlling for non-verbal IQ and grade,

the hearing impaired group performed significantly below the normal group both on the

vocabulary and grammatical knowledge. Specifically, the differences of adjusted means

of the PPVT-III and Grammatic Knowledge subset of the CASL test was 16.651 and

18.769, respectively, which were larger than one standard deviation (15.0). The EVT also

showed substantial difference of 10.963. Notably, as seen in the adjustment mean

difference, the effect size was biggest on the measure of grammatical knowledge (partial

rf = .392).

MANCOVA for phonology

In order to determine whether there was a significant difference between the three

study groups (NH, HI, and RD) on phonological processing measures, six dependent

variables from the CTOPP representing components of phonological awareness,

phonological memory, and rapid naming skills were submitted to multivariate analysis of

covariance. MANCOVA was conducted with group membership as the independent

variable and grade in months as the covariate.

Box's M statistic was used to check the covariance assumption, which was not

violated: F[6, 11817.457] = 9896, p = .831), suggesting that the observed covariance

matrices of the dependent variables were equal across groups. The Levene's test was used

to test the equal variance assumption. All six measures had equal variances in the three

groups (dfi = 2, df2 = 75): (1) Elision: F = .272, p = .763; (2) Blending: F = 2.102, p

= .129; (3) RAN-Digit: F = .433, p = .650; (4) RAN-Letter: F = .005, p = .995; (5)

Memory for digit: F = .202, p =. 818; and (6) Nonword repetition: F = 1.344, p = .267.









The resulting MANCOVA was significant, yielding a main effect for group, F(12,

138) = 8.450, p <.001, showing a significant difference between the groups' overall

performance on the phonological processing measures. The observed means, the

observed standard deviations, and means adjusted after the statistical removal of the

grade effect, standard errors, and observed powers for three groups appear in Table 4-8.

Because of this significant difference in overall measures, a series of follow-up one-way

univariate ANCOVAs were conducted.

Differences between groups were significant for every phonological measure after

adjusting for grade in month (dfi = 2, df2 = 74): (1) Elision (F = 26.803, p = .000, partial

/= .420, observed power = 1.00); (b) Blending (F = 20.281, p = .000, partial r/2= .354,

observed power = 1.000); (c) RAN for Digit (F = 13.767, p = .000, partial r2= .271,

observed power = .998); (d) RANfor Letter (F = 13.584, p = .005, partial 2= .269,

observed power = .997); (e) Memory for Digits (F = 4.993, p = .009, partial rf = .119,

observed power = .798); and (f) NonwordRepetition (F = 25.78, p = .000, partial rf

= .411, observed power = 1.00).

The effect sizes for the Blending and Elision subtests were moderately large

(Elision = .420, Blending = .354). In contrast, the effect sizes for rapid naming skill were

smaller than those for phonological awareness skills (RAN-Digit = .271 and RAN-Letter

= .269). A summary of six univariate ANCOVAs is shown in Table 4-9.

For all phonological processing measure showing significant between-group

differences, a series of post hoc pairwise comparisons for the three groups (i.e., NH vs.

HI, HI vs. RD, and NH vs. RD) based on Bonferroni adjustment was conducted to

investigate how the groups are different across each of the dependent variables. Results









are displayed in Table 4-10. Consistent with the previous literature (Cornwall, 1992;

Cronin & Carver, 1998; Snyder & Downey, 1995; Wolf& Bowers, 1999; Wolf et al.,

2002), on average, dyslexic children in this study were severely impaired on all reading

measures when compared to the normal controls. Adjusted mean differences (NH minus

RD) and t-scores on each measure were as follows: (1) Elision: 4.593, t(57) = 6.917, p =

.000; (2) Blending: 2.645, t(57) = 4.401, p = .000; (3) RAN-Digit: 3.043, t(57) = 5.247 ,p

= .000; (4) RAN-Letter: 3.281, t(57) = 5.208, p = .000; (5) Memory for digits: 1.999,

t(57) = 2.750, p = .022; and (6) Nonwordrepetition:3.577, t(57) = 5.478, p = .000.

Similarly, HI subjects' phonological awareness and phonological memory skills

were significantly lower than in the NH group. Adjusted mean differences (NH minus

HI) and t-scores on each measure were as follows: (1) Elision: 3.500, t(46) = 4.888, p =

.000; (2) Blending: 3.849, t(46) = 5.949, p = .000; (3) Memory for digits: 1.953, t(46) =

2.494, p = .044; and (4) Nonword repetition: 4.496, t(46) = 6.395, p = .000.

Unexpectedly, however, no significant differences in the subtests of rapid naming were

reported between the NH and HI groups. Adjusted mean differences (NH minus HI) and

t-scores on rapid naming measures were as follows: (1) RAN-Digit: 1.302, t(46) = 2.083,

p = .122 and (2) RAN-Letter: 1.245, t(46) = 1.836, p = .211.

Hence, these data have shown that hearing impaired students' lexical access and

retrieval skills are well preserved. Additionally, our post hoc comparisons indicated that

the HI group performed significantly better than the RD group on the two RAN subtests.

While the HI group did not differ from the RD group on tasks of phonological awareness

and phonological memory, when adjusted means (HI minus RD) for the Nonword

Repetition and Blending subtests were used, RD subjects scored slightly higher with no









significant difference. Adjusted mean, t-scores, and associated p-values on each measure

were as follows: (1) RAN-Digit: 1.741, t(47) = 2.568, p = .033; (2) RAN-Letter: 2.035,

t(47) = 2.803, p = .019; (3) Elision: 1.093, t(47) = 1.427, p = .473 (ns); (4) Blending: -

1.204, t(47) = 1.737, p = .260 (ns); (5) Memory for digits: .047, t(47) = .056, p = 1.00

(ns); and (6) Nonword repetition: -.919, t(47) = 1.220, p =.679 (ns).

Overall, the HI group showed depressed phonological processing skills only in the

areas of phonological awareness and phonological memory. No differences existed

between the normal controls and the HI groups on the measures on rapid naming. In

contrast, the RD subjects in this study showed deficiencies in all three phonological

processing components (phonological awareness, phonological memory, and rapid

naming). Descriptive results are displayed graphically in Figures 4-5 through 4-7 as

clustered boxplots for each phonological component.

MANCOVA for literacy measures (reading and spelling)

For the reading and spelling measures, a one-way MANCOVA was used with

group membership (NH, HI, and RD) as the independent variable and grade in months as

the covariate.

Three out of eight dependent variables had statistically significant inequalities of

variance based on the Levene's test for equality of variance (p < 0.05): (1) spelling:

F[2,75] = 3.782, p = .027; (2) timed word reading (TOWRE): F[2,75] = 4.275, p = .017;

and (3) passage reading rate: F[2,75] = 3.834, p = .026. However, according to Hair et al.

(1992), a violation of this assumption has minimal impact if the groups (HI, NH, and RD)

are approximately of equal size or if the largest group size divided by the smallest group

size is less than 1.5. The ratio of dyslexic group size (n = 30) to the group of hearing

impaired students (n = 19) was near to 1.50 (ratio = 1.579), hence, any violation of this









assumption should have minimal impact. In light of the above findings, it is not

surprising that the Box's M-test for multivariate homoscedasticity indicated significant

differences as well (F = 1.544, df1 = 72, df2 = 11146.98, p = .002). This test is

"notoriously sensitive" (Tabachnick and Fidell, 1996, p. 382), particularly given the large

sample size involved. The group sizes in our data are not large and relatively equal to

each other, so the impact should be minimal; all analyses of variance have been

conducted assuming unique variance between groups.

The multivariate analysis results showed a significant difference between the three

groups, when MANCOVA was carried out for the entire measures of reading and spelling

(Wilks' Lambda = 0.228, F[16, 134] = 9.179, partial qr = .523, andp <.001). This

significant differences in the analyses for overall tests justified further investigation of the

significance of each dependent measure using univariate tests of ANCOVA. ANCOVAs

showed significant group effects for all seven measures: (1) untimed word reading

(F[2,74] = 58.610 p < .001); (2) untimed nonword reading (F[2,74] = 42.430 p < .001);

(3) passage comprehension (F[2,74] = 35.084, p < .001); (4) spelling (F[2,74] = 40.899,

p < .001); (5) timed word reading timed (F[2,74] = 68.634, p < .001); (6) timed nonword

reading (F[2,74] = 73.415, p < .001); (7) passage reading rate (F[2,74] = 77.143 p

< .001); and (8) passage reading accuracy (F[2,74] = 59.777, p < .001).

Descriptive statistics such as means, standard deviations, and estimated adjusted

means for all tasks are displayed in Table 4-11. Error bar charts and box plots based on

these descriptive data are also seen in Figures 4-7 through 4-10. Significant group effects

based on eight univariate ANCOVAs for all measures and the corresponding ANOVA

tables are seen in Table 4-12 and Table 4-13, respectively.









Next, pairwise comparisons using Bonferroni's adjustment were conducted.

Inconsistent with the previous literature (Briscoe et al., 2001; Gibbs, 2004), on average,

reading and spelling skills of hearing impaired children in this study were significantly

lower than normal controls on all reading measures (HI vs. NH) with one exception of the

Word Decoding subtest on the TOWRE. Adjusted mean differences (NH minus HI) and

t-scores on each measure were as follows: (1) untimed word reading: 14.386, t(46) = 3.75,

p = .001; (2) untimed nonword reading: 10.003, t(46) = 2.65, p = .029; (3) passage

comprehension: 12.058, t(46) = 3.40, p = .003; (4) spelling: 13.222, t(46) = 3.72, p

= .001; (5) timed word reading: 7.873 t(46) = 2.46,p = .048; (6) timed nonword reading:

6.155, t(46) = 1.90, p = .185; (7) passage reading rate: 2.664, t(46) = 3.14, p = .007; and

(8) passage reading accuracy: 2.883, t(46) = 3.10, p = .007. Descriptively, adjusted mean

differences in timed word/nonword reading were smaller than untimed word-level

reading, suggesting a contributive role of hearing-impaired readers' preserved lexical

access skills to fast reading.

In addition, as expected, every mean score on the tests of reading were low average

for the RD group and the HI group in our sample was significantly better than the RD

controls on all measures. Remarkably, in timed measures, mean differences were more

than 1.5 to 2 standard deviations. Adjusted mean differences (NH minus RD) and t-scores

on each measure were as follows: (1) untimed word reading: 21.851, t(47) = 5.71, p

= .000 ; (2) untimed nonword reading: 20.091, t(47) = 5.34, p = .000; (3) passage

comprehension: 14.018, t(47) = 3.96, p = .001; (4) spelling: 15.013, t(47) = 15.013, p=

4.23; p = .000; (5) timed word reading timed: 24.020, t(47) = 7.53, p = .000; (6) timed

nonword reading: 26.832, t(47) = 8.28, p = .000; (7) passage reading rate: 6.394, t(47) =









7.56, p = .000; and (8) passage reading accuracy: 5.782, t(47) = 6.3, p = .000. Table 4-13

is a summary of pairwise post-hoc comparisons. Lastly, RD group's reading and spelling

scores were significantly lower than normal group's for all measures.

Question Category II: Correlations among Measures

Our second category of research questions involved the interrelationship between

hearing impaired participants' literacy skills and other exploratory variables, including

oral language, phonological processing, and auditory processing ability: The first step in

this analysis was to explore the pattern of associations in correlation analyses.

Correlational analyses are presented for the research questions below:

1. Question 1: What is the relationship between phonology and reading?
2. Question 2: What is the relationship between phonology and hearing loss?
3. Question 3: What is the relationship between phonology and oral language?
4. Question 4: What is the relationship between phonology and auditory processing?
5. Question 5: What is the relationship between reading, auditory processing?

Question 1: Phonology and reading

To assess the independent associations of each of the measures on phonological

processing and reading skills, partial correlation coefficients were calculated with the

effects of age, non-verbal intelligence (TONI-3), and hearing loss removed. Hearing loss,

as measured by the better ear's pure-tone, was also removed from the correlations

because the degree of hearing loss exhibited strong negative relations with phonological

awareness when age and grade in months were partialed out (Elision: r = -.631, p = .007;

Blending: r = -.728, p = .001; See Table 4-), short-term memory span (Nonword

repetition: r = -.608, p = .010). Partial coefficients were computed for phonological

processing and reading measures. The partial correlation matrices for the HI and NH

separately are shown in Table 4-15 and Table 4-16. The results will be reported

separately for the HI and NH groups.









Hearing-impaired group: As can be seen in Table 4-15, the Elision subtest was

more highly correlated with reading measures than any other phonological processing

variables. It was the only measure of phonological awareness which was significantly

associated with all reading and spelling measures in both word and passage levels.

Correlation coefficients between Elision and reading measures ranged from .562 to .751

as seen below: (1) untimed word reading (r = .582, p < .01); (2) untimed nonword reading

(r = .751, p < .001); (3) timed word reading (r = .562, p < .05); (4) timed nonword

reading (r = .611, p < .05); (5) spelling (r = .618, p < .01); (6) passage comprehension (r

= .610, p < .01); (7) passage reading rate (r = .605, p < .01); and (8) passage reading

accuracy (r = .711, p <.001). Even though the Blending is a measure of phonological

awareness, it was not correlated with any word-level reading measures, spelling, and

comprehension. It was only significantly associated with oral passage reading rate (r =

.470, p < .05).

The pattern of correlations between rapid naming skills and reading measures was

remarkable. Both RAN measures did not show any significant correlations with untimed

tasks of reading at the word or passage level. Correlation coefficients among these

variables were as follows: (1) untimed word reading (RAN-D: r = .307, p = .124; RAN-

L: r= .259,p =. 166); (2) untimed nonword reading (RAN-D: r =.411, p = .057; RAN-L:

r = .301, p = .129); and (3) passage comprehension (RAN-D: r = .341, p = .098; RAN-L:

r = .314, p = .118). Notably, however, significant correlations were found for rapid

naming measures and timed reading tests at both word and passage levels: (1) Timed

word reading (RAN-D: r = .459, p = .037; RAN-L: r = .559, p = .012); (2) timed nonword

reading (RAN-D: r = .560, p = .012; RAN-L: r = .656, p = .003); (3) timed passage









reading rate (RAN-D: r = .723,p = .001; RAN-L: r = .581, p = .009); and (4) timed

passage reading accuracy. (RAN-D: r = .635,p = .004; RAN-L: r = .571, p = .003).

Correlation coefficients ranged from .459 to .656 at word level and from .571 to .723 at

the passage level. All correlations in the partial correlation matrix were higher than

medium (about .3) or large (above .5) correlations (Cohen, 1988). The phonological

memory tasks as measured by the Nonword Repetition task was correlated with untimed

word reading (r = .594, p = .008). With other literacy measures, phonological STM tasks

approached significant level (See Table 4-15).

Normally hearing controls: As can be seen in Table 4-16, after the effects of age

and non-verbal intelligence had been controlled, phonological awareness variables were

highly correlated with reading and spelling variables: (1) untimed word reading (elision: r

= .711, p < .001; blending: r = .378, p < .05), (2) untimed nonword reading (elision: r

= .692; p < .001; blending: r = .472, p < .05), (3) spelling (r = .825, p < .001), (4) passage

comprehension (elision: r = .520, p <.05), (5) timed word reading (r = .474, p < .001), (6)

passage reading rate (elision: r = .443, p < .05), and (7) passage reading accuracy (r

= .671, p < .001). Highest correlations were observed between the Elision and spelling

tests. Quite differently from the HI group, significant small to medium intercorrelations

were found between rapid naming and untimed word reading tests for the NH subjects

(word reading (RAN-D: r = .421, p = .016; RAN-L: r = .414, p = .018); (2) untimed

nonword reading (RAN-D: r = .382, p = .027; RAN-L: r = .437, p = .013). Also, similar

to the HI group, high correlation coefficients were found in the NH group between rapid

naming tasks and timed reading tests for both word and passage level: (1) timed word

reading (RAN-D: r = .559, p = .001; RAN-L: r = .542, p = .002); (2) timed nonword









reading (RAN-D: r = .596, p = .001; RAN-L: r = .629, p = .000); (3) timed passage

reading rate (RAN-D: ; r = .382,p = .027; RAN-L: r = .385,p = .026); and (4) timed

passage reading accuracy. (RAN-D: r = .465,p = .008; RAN-L: r = .540,p = .002).

Another notable correlation pattern was found for phonological short-term memory

measures. Different from the HI group, the two untimed word-level reading and spelling

measures were highly correlated in the NH group. Partial coefficients are as follows: (1)

Memory for Digits: untimedd word reading: r = .657, p = .000; untimed nonword reading:

r = .603, p = .001; and spelling: r = .519, p = .003) and (2) Nonword repetition: untimedd

word reading: r = .686, p = .000; untimed nonword reading: r = .650, p = .000; and

spelling: r = .534, p = .002).

Question 2: Phonology and hearing loss

The partial correlation matrix adjusting for the variance accounted for by age and

non-verbal IQ for the HI group (n=19) are shown in Table 4-17. Table 4-17 includes the

six measures of phonological measures and variables related to hearing loss such as better

ear's pure-tone, age at identification, age at initial hearing aid fitting, and length of

hearing aid use. As expected, the better ear's PTA was correlated with phonological

awareness and short-term memory span as seen in the following partial coefficients: (1)

Elision (r = -.473, p = .024); (2) Blending (r = -.597, p= .004); (3) Nonword repetition (r

= -.424, p = .040), underscoring the effect of decreased auditory input on meta-

phonological skills.

On the other hand, the RAN tasks (letters and digits) were not correlated with the

severity of hearing loss. This finding was also supported by the MANCOVA results,

which revealed no effect of hearing loss on rapid naming skills.









Question 3: Phonology and oral language

It was also questioned whether lexical and grammatical knowledge would be

related to phonological processing skills which measures children's ability to encode,

store, and retrieve phonological information. The correlation and partial correlation

matrices for the HI group are shown in Table 4-18. The results revealed that phonological

short-term memory (nonword repetition, memory for digits) were strong correlates of

both receptive and expressive vocabulary, and grammatical knowledge. Notably, the

Memory for digit subtest was highly correlated with all language measures: (1) PPVT-III

(r = .533, p = .014), (2) EVT (r =.592,p =.006), and (3) CASL (r = 558,p = .010). The

Nonword repetition was also moderately correlated with the receptive and expressive

vocabulary, but only approached the significance level: (1) PPVT-III (r = .406, p = .053)

and (2) EVT (r = .359,p = .078).

In contrast, no significant association was found between the RAN tasks and any of

oral language measures. Similarly, phonological awareness tasks did not show any

correlations with vocabulary tests. However, a pattern of significant correlations was seen

between phonological awareness and grammatical knowledge test, revealing coefficients

higher than medium level: (1) Elision (r = .511, p = .018) and (2) Blending (r = .443, p

= .038).

Question 4: Auditory processing, phonology, and reading measures

Table 4-19 shows correlations among the auditory and phonological measures and

reading for the HI group, with partial correlations controlling for age, grade, duration of

hearing aid use, and nonverbal intelligence.

Auditory processing skills measured by the three SCAN subtests (Filtered Words,

Auditory Figure-Ground, and Competing Words) had no significant correlations with any









of phonological awareness and rapid naming tasks. However, a high degree of significant

associations were found between the Competing Words subtest and short-term memory

span tasks (i.e., Memory for Digits: r= .653, p =.008; Nonword Repetition: r= .579, p =

.024). The Competing Words subtest is a test of dichotic listening, where the subject

hears two words simultaneously, one word presented to each ear. The subject is

instructed to repeat the words presented in each ear. For a better performance, the two

stimuli presented simultaneously should be efficiently stored in phonological buffer for

later repetition. Thus, the high level of association between this type of dichotic listening

test and short-term memory span was not surprising. Similarly, high correlations were

found between the Dichotic Digit test and short-term memory span, an en expected

results based on the previous explanation (i.e., Memory for Digits: r = .617, p = .014;

Nonword Repetition: r = .548, p = .034).

None of the SCAN subtests showed significant correlations with any of reading

measures, including word-level reading and passage reading and spelling. However, the

Dichotic Digit task correlated with the two nonword reading tasks untimedd nonword: r

= .546, p= .035 and timed nonword: r = .525, p= .044).

Question 5: Reading and oral language

Partial correlation coefficients were calculated between all of reading measures and

oral language scores with the effects of age, grade, nonverbal intelligence, and the degree

of hearing loss partialed out. Table 4-20 displays the correlations matrix. Passage

comprehension was the only reading tests that correlated with the oral language

measures: (1) PPVT-III (r = .474, p = .037), (2) EVT (r = .536, p = .020), and (3)

grammatical knowledge (r = .442, p .049). Interestingly, even though the GORT test









measured reading fluency in passage level, no correlations were found between

vocabulary and the passage reading fluency.

Question Category III: Exploratory Hierarchical Regressions

Although MANCOVA procedures can test for the statistical significance of group

differences in one or more domains, these procedures do not necessarily provide a clear

indication of how hard-of-hearing readers' depressed phonological skills are predictive of

reading skills across wide range of domains.

In an effort to explore the relations between the reading measures and the

exploratory variables, including phonological and auditory processing abilities, selected

based on the previous partial correlation analyses, a series of multivariate analyses (i.e.,

hierarchically blocked regressions) was performed separately for each reading measure.

According to Whitley (1996), there are three different forms of multiple regression

analysis: simultaneous (enter), stepwise, and hierarchical and each of them has a specific

purpose. Hierarchical regression should be used for hypothesis testing while

simultaneous and stepwise regressions should be used only for simple prediction.

In this study, specific hypotheses were tested regarding the significance of selected

exploratory variables, so the use of hierarchical regression model was justified. Findings

from multiple regression analyses allowed for examination of whether the children's

performance in reading was significantly associated with the proposed set of exploratory

variables, including phonological awareness, short-term memory span, rapid naming, and

auditory processing ability

For regression analyses reported in this section, six phonology measures from the

CTOPP test and one auditory processing test (Dichotic Digit) served as the exploratory

variables. All of seven reading measures and a spelling measure were used as dependent









variables. Each of these exploratory predictors was entered into the model individually in

Step 2 after the effects of children's background characteristics were statistically

eliminated. Specifically, grade, nonverbal IQ, and degree of hearing loss (better ear's

PTA) were entered to control for the effects of these variables on reading performance

controlled for. These three factors were entered at Steps 1 in all analyses. These

hierarchical regression results are presented in Table 4-22 through Table 4-28. The

research questions are as follows:

Regressions on word-level reading

Separate fixed-order hierarchical regression analyses with were carried out

separately on four word-level reading achievement measures (i.e., untimed/timed,

words/nonwords) to determine the variance contributed to the word-level reading

accuracy by phonological awareness, rapid naming, short-term memory, and auditory

processing skills (32 regressions = 4 reading measures 8 predictors). As reported in the

previous section, the zero-ordered correlations had shown significant correlations

between reading and phonological processing tests. Especially, the Dichotic Digit test

was included in the regression analyses since it was the only central auditory processing

measures which was significantly correlated with phonological measures (i.e., elision,

blending, nonword repetition, and memory for digits) and reading measures (i.e., timed

and untimed nonword reading). The results are displayed in Table 4-21 through Table 4-

24.

Regression on untimed word reading accuracy: The Word Identification subtest

on the WRMT-R was the dependent variable. As shown in Table 4-21, when grade, better

ear's PTA, and nonverbal IQ were entered in Steps 1, they explained 31.9% of the

variance in the timed single-word reading scores. The results demonstrate that only the









Elision task significantly explained substantial amount of variance (30.3%) in the Word

Identification test (p = .005).

Taken as a whole, the three control variables (grade, better ear's PTA, nonverbal

IQ) and the Blending test accounted for 64.1% of variance for the Word Identification,

showing the importance of phonological awareness in word recognition. The Nonword

Repetition and Dichotic Digits subtests only approached significance with the Word

Identification test(p = .054 andp = .093, respectively). Both the RAN-Letters and RAN-

Digit tasks did not explain variance in this untimed word reading test (p =. 131 and .081,

respectively).

Regression on untimed nonword reading accuracy: The Word Attack subtest on

the WRMT-R was the dependent variable. In Step 1, grade, better ear's PTA, and

nonverbal IQ explained 11.0% of variance in the score. Similar to the WordIdentification,

when the Elision scores were entered in Step 2, it explained almost more than half of

additional variance in the scores of the Word Attack subtest in a significant way (53.1 %,

p < .0001). When the Nonword Repetition was entered instead of the Elision, it also

explained 37% of additional variance. So, both the Elision and Nonword Repetition

represented a substantial effect on the untimed single nonword reading scores (Table 4-

22).

However, phonological short-term memory (Memory for Digits) only approached

significance level and it explained about 19% of variance (p = .072). Unlike the Blending,

when the RAN (Digits and Letters) scores were entered in Step 2 individually, none of

them was significantly associated with the Word Attack scores. With the Dichotic Digit









entered in Step 2, the overall regression model did not significantly explained the

variance in the Word Attack scores (i.e., F[4,14] = 2,300, p = .110).

Regression on timed word reading accuracy: The Word Efficiency subtest on the

TOWRE was the dependent variable. As Table 4-23 shows, when entered first in the

regression function, grade, PTA, and nonverbal IQ accounted for 19.3% of the variance

in timed nonword reading. In Step 2, following these controlled variables, each

exploratory variable was entered separately into the regression equations. Similar to the

results of regressions on the untimed reading condition, phonological awareness skill as

measured by the Elision subtest continued to significantly account for unique variance in

the dependent variable (28.7%, p = .015). It was noted that the Memory for Digits (p

= .178), Nonword Repetition (p = .115), and Dichotic Digit subtests (p = .306) did not

explain the variance in a significant way.

It was of further interest to determine if the rapid naming tasks accounted for

significant variance for timed reading activities. Recall that our correlation analyses

indicated that rapid naming skills were strongly associated with timed reading tasks only

and no significant partial correlations were observed for untimed reading measures:

Correlation coefficients for timed measures were high, ranging from .459 to .656 at word

level and from .571 to .723 at passage level. This strong correlation between rapid

naming and timed reading was confirmed by the regression results. That is, in the second

order of entry, the alphanumeric rapid naming measures accounted for significant

variance in the Word Efficiency scores. Specifically, the RAN-Letter accounted for an

additional 24.5% of the variance of scores in the timed word reading (p = .027) and the

RAN-Digit test also explained almost 20% of the variance (19.4%, p = .049).









Regression on timed nonword reading accuracy: The Word Decoding subtest on

the TOWRE was the dependent variable. When entered first in the regression function,

non-verbal IQ, grade, and PTA accounted only for 9.7% of the variance in the Word

Decoding. Similar to the previous results, phonological awareness measure (Elision)

explained a highly significant 37% of unique variance as shown in Table 4-24.

No significance was seen for the contribution of phonological memory. Children's

performance on the Dichotic Digit test only approached significance with the Word

Decoding (p = .054). There was also a significant additional contribution from the

alphanumeric naming measures at Step 2, explaining a highly significant 39.2% (RAN-

Letter, p = .005) and 31.3% (RAN-Digit, p = .016) of unique variance, respectively. This

confirmed the fact that there is a significant additional contribution of alphanumeric RAN

to timed reading tasks.

Regression on passage reading rate and accuracy

Another set of analyses were carried out to test for the effects of our exploratory

variables on reading skills in passage level. Recall that the RAN tasks did uniquely and

significantly account for the variances of all timed, but not for untimed reading tests. So,

we wanted to see if a similar trend would be found as well in passage level reading.

As Table 4-25 shows, each of our two RAN variables, when separately entered at

the Step 2, accounted for substantial proportions of variance in both passage rate and

accuracy scores below the significant level: The RAN-Digit accounted for 37.6% (p

= .002) for rate scores and 31.7% for accuracy scores (p = .010), respectively. In a similar

pattern, it was also observed that the RAN-Letter accounted for 19.7% (p = .044) for rate

and 22.1% (p = .039) for accuracy.









Like our previous observation, the Elision consistently explained unique variances

for both rate and accuracy scores below the significant level: 29.1% for rate (p = .011)

and 42.8% for accuracy (p = .002), respectively. Children's performance on the Memory

for Digit subtest also significantly explained 18.9% of variance in the passage rate scores

(p = .049), but not for the accuracy score. The Nonword Repetition did not significant

account for any of passage-level tests.

Finally, the Dichotic Digit test did not show significant association with the

performance on the rate measure, but it uniquely and significantly accounted for 26.2%

of variance in the rate scores (p = .017). This rate test was the only literacy measure for

which the variance was significantly accounted for by the Dichotic Digit test.

It was noteworthy that when the RAN-Digit was entered in Step 2, the overall

regression models explained significantly and uniquely 61.3% and 50.0% of variance,

respectively, for passage reading rate and accuracy.

Regression on spelling

One final set of multiple regression analyses were carried out to test for the effects

of our exploratory variables on spelling. The Spelling subtest on the WRAT was the

dependent variable. In Step 2, chronological grade, better ear's PTA, and nonverbal IQ

accounted for 18.3% of the variance. The results are reported in Table 4-27. Rapid

naming measures did not significant predict children's spelling ability.

Again, significant contribution of phonological awareness was consistent: The

Blending subtest explained statistically significant 33.8% of variance in spelling (p

= .007). With the controlled variables entered in Step 1 taken together, the overall model

accounted for more than half of variance in spelling scores (52.1%). Similarly,









phonological short-term memory span as measured by the Memory for Digit explained

almost 30% unique variance (p = .015).

When the Dichotic Digit subtest was entered in Step 2, which had significant

standardized coefficient (P = .663, p = .016), the overall regression model itself did not

significantly reduce the error term (F[4,14] = 3.071, p = .052).

Role of rapid naming in further regression analyses

In our previous regression analyses, rapid naming tests did not account for unique

variance in untimed reading tasks in both word and passage level. Instead, it was shown

that the RAN tasks are significantly associated only with timed reading tests, which

demands high efficiency in extraphonological processing such as rapid lexical access and

retrieval. Because this finding is relevant to the role of speed of processing (RAN) as a

non-phonological factor in reading disabilities (Compton, DeFries, & Olson, 2001;

Cornwall, 1992; Kirby, Parrila, & Pfeiffer, 2003; Savgrade et al., 2005; Strattman &

Hodson, 2005; Wolf& Bowers, 1999; Wolf& O'Brien, 2001), it was necessary to ensure

that the RAN would still accounted for unique variance of each timed reading task after

the effects of phonology-based cognitive skills were also controlled for.

As a further test, a set of additional hierarchical regressions were performed by

entering two phonological awareness (Elision Blending) and two phonological short-term

memory (Memory for Digit Nonword Repetition) in the same block (Step 2) as a group.

Each of the RAN tests (digit/letter) was entered separately in Step 3 for all timed reading

measures. Similar to our previous regression, grade, non-verbal IQ, and better ear's PTA

were entered in Step 1. The results of these regressions are displayed in Table 4-28.

In accordance to our hypotheses, the results of further analyses demonstrated that

once IQ, grade, and better ear's PTA were considered first, a significant additional









contribution from the alphanumeric naming measures was preserved: In the model where

the RAN-Digit was entered in Step 3, it continued to account for unique variances in

passage reading rate significantly (23.6%, p = .005), passage reading accuracy (16.0%, p

= .022), and timed nonword reading (20.1%, p = .029), respectively. Similarly, the

second set of analyses also confirmed that there was significant additional contribution of

the RAN-Letter measure at Step 3, explaining 21.2% (p = .023) and 35.8% (p = .001)

unique variances of timed word and nonword reading accuracy. The RAN-Letter also

explained smaller but significant amount of variances in passage reading rate and

accuracy (14.9% and 15%, respectively).

Summary:

1. The Elision from the CTOPP test, a measure of phonological awareness, was the only
measure, which consistently and significantly explained variance in all of literacy
measures ranging from word and nonword reading, spelling, to passage reading
fluency. Unexpectedly, the Blending subtest was not significant for any of tests. The
results also showed that alphanumeric rapid naming skills only accounted for unique
variance in timed reading achievement in both word and passage level. In contrast, no
significant contribution to timed reading was revealed.

2. When the effects of grade, non-verbal IQ, and better ear's PTA were eliminated,
auditory processing skill as measure by the Dichotic Digit was not significant for any
word-level reading and spelling tests. It explained statistically significant 26.2% of
variance in passage reading rate (26.2%).

3. In addition, the significant contribution of phonological short-term memory skill was
seen only for two literacy measures. The Memory for Digit subtest explained 28.7%
unique variance in spelling scores and the Nonword Repetition accounted for 37.0%
unique variance in untimed nonword reading (p = .015 and p = .007, respectively).









Table 4-1. Background information and basic auditory skills of individuals in the HI group.

Subj Gende Agea Gradeb PTAc SRTo WRSd HA- Age Ida HA Etiology Hearing aid type
r R L Fita Use
1 F 115 4.3 63 45 88 100 24 24 88 Genetic PhonakPico Forte
2 M 111 4.3 58 55 100 100 48 48 57 Unknown Oticon 380P
3 F 114 4.4 38 40 100 100 30 30 84 Genetic Phonak Pico Forte
4 M 100 2.5 70 65 46 67 20 20 75 Unknown Oticon 380P
5 F 112 3.5 68 65 76 89 27 27 81 Genetic Oticon Multi-Focus
6 F 106 2.5 38 35 93 100 42 42 66 Genetic Phonak Sono Forte
7 M 86 1.5 31 30 100 100 24 24 62 Unknown Unitron UM 60
8 F 125 3.5 27 20 100 100 36 36 75 Unknown Oticon 39 PL
9 F 121 5.6 56 45 90 100 72 72 48 Genetic Telex 366
10 M 91 2.7 65 35 100 88 7 7 80 Unknown Phonak Pico Forte
11 M 99 1.9 38 35 96 88 48 48 25 Unknown Oticon 380P
12 M 122 2.9 41 35 100 100 18 18 98 Genetic Oticon 380P
13 M 84 1.0 43 35 88 96 29 29 33 Unknown Phonak Sono Forte
14 M 152 6.9 41 40 100 100 36 36 115 Unknown Unitron Icon
15 F 84 1.1 41 35 100 100 61 61 20 Genetic Phonak Pico Forte
16 F 152 6.0 43 30 52 68 72 72 80 Unknown Unitron Icon
17 M 106 3.0 28 10 56 72 Unknown Unitron UM 60
18 M 111 3.2 45 45 72 80 24 24 51 Unknown Oticon 38 P
19 M 116 3.2 58 50 52 76 24 24 82 Unknown Rion HB 75AL
11
M B 110.8 3.4 46.9 39.5 75.2 86.0 35.6 35.6 67.8
(SD) Boys (19.3) (1.61) (13.5) (13.5) (29.7) (21.9) (17.8) (17.8) (25.1)

Note. a:in months, b:in years.months, c :unaided, d: measured in the Most Comfortable Level, PTA: Pure-tone threshold, SRT:
Speech recognition thresholds, WRS: Word recognition scores, HA_Fit: Age at initial hearing aids fitting, Age_Id: Age at
identification, HA Use: Duration of hearing aids use









Table 4-2. Descriptive statistics for the NH and HI groups on PTA, SRT, and WRS scores.
NHa,b HIa,c

Mean SD Min Max Mean SD Min Max


PTA(L)d

PTA(R)d

PTA
(Better ear)d
SRT(L)e

SRT(R)e

WRS(L)t

WRS(R)f


5.1

3.8

3.31

4.1

5.2

99.7

99.0


4.7

1.15

3.9

3.8

3.4

1.48

2.96


13.0

21.0

13.0

15.0

10.0

100

100


51.0

54.9

46.9

40.8

47.8

86.0

75.2


15.4

21.1

13.5

14.5

22.4

21.9

29.6


Note. a: in dB HL, b: Normally hearing children, c: Hearing-impaired children, d: unaided pure-tone threshold average,
e: unaided speech recognition thresholds, f: word recognition scores in the most comfortable level (MCL)









Table 4-3. Mean pure-tone thresholds for all tested frequencies (NH and HI groups only).
Ear 250 Hz 500 Hz 1000 Hz 2000 Hz 3000 Hz 4000 Hz 6000 Hz 8000 Hz

M 37.63 47.11 59.21 63.68 60.79 62.37 68.16 70.53

*HI SD 20.640 21.558 20.225 18.845 69.450 23.943 22.249 20.541

Range 5 -75 5 -90 30 100 35 105 25 90 20 95 30- 105 30 105
Right
M 4.31 5.86 4.66 5.69 4.14 6.03 3.10 5.17

NH SD 5.782 5.012 4.805 6.778 5.680 6.322 5.414 5.745

Range 0 20 -5 15 0 20 0-35 -5 20 -5 25 -5 15 -5 25

M 32.63 39.21 48.42 56.32 56.32 57.89 61.58 65.63

*HI SD 16.361 16.352 19.794 15.532 16.570 21.168 22.977 21.660

Range 5-65 10 70 15 95 35 -95 25 90 20-95 30 105 30 105
Left
M 3.79 4.66 4.14 3.79 5.34 6.03 2.07 4.41

NH SD 5.454 5.659 4.446 5.287 4.616 6.461 6.053 6.033

Range -5 15 -5 15 -5 15 -5 15 0 20 -5 25 -10 20 -5-20

Note. *: unaided thresholds, HI (n=19), NH (n=29)
















Hearing Impaired Group (Left Ear)


+ -50- .



-60i



-70-



-80"

250HZ 500HZ 1000HZ 2000HZ 4000HZ 6000HZ


Figure 4-1. Mean PTA thresholds for the left ear (HI Group, N=19)


Hearing Impaired Group (Right Ear)


LU

cn
+ -50-



-Eo-



-70-



-80-


250HZ 500HZ 1 000HZ 2000HZ 4000HZ 6000HZ



Figure 4-2. Mean PTA thresholds for the Right ear (HI Group, N=19)


8000HZ


8000HZ















arouo: imDaired


Normal

Mean =46.95
Std Dev. =13.493
N =19


20 30 40 50 60 70 80

Better ear PTA



Figure 4-3. Histogram of 19 hearing impaired children by better ear pure-tone average.









Table 4-4. Descriptive statistics of auditory processing variables.
Group SCAN (FW) SCAN (AFG) SCAN (CW) DD

Mean 12.10 9.41 10.14 8.54

SD 1.970 2.383 2.812 8.533
NH Min 6 5 4 67.50


Max 15 15 14 98.75


Mean

SD

Min

Max


5.89

4.653

1

13


3.11

2.331

1

7


4.11

2.492

1

8


66.97

18.13

38.75

90


Note. SD: standard scores, FW: Filtered Words, AFG: Auditory Figure Ground, CW: Competing Words,
DD: dichotic digit in %









Table 4-5. Descriptive statistics and estimated adjusted means of oral language measures (MANCOVA)
Dependent 95% CI
Group M SD Min Max Median EAM SE
variables lower Upper

NH 112.00 12.45 81 132 113.0 110.45 2.496 105.42 115.47
PPVT-III
HI 91.42 15.31 62 115 93.0 93.80 3.133 87.48 100.10

NH 107.38 10.46 92 134 105.0 106.55 2.238 102.04 111.05
EVT
HI 94.32 13.46 63 112 95.0 95.59 2.809 89.92 101.24

NH 117.86 12.68 89 149 118.0 116.51 2.122 112.24 120.78
CASL
HI 95.68 10.40 75 115 96.0 97.74 2.663 92.38 103.12

Note. PPVT-III (receptive vocabulary), EVT (expressive vocabulary), CASL (grammatical knowledge), SE (standard error),
NH: normally hearing group, HI: hearing-impaired group, M: mean, SD: standard deviation, Min: minimum, Max: maximum,
EAM: estimated adjusted mean, SE: standard error









Table 4-6. ANOVA tables for univariate ANCOVAs for each oral language measure.
MS MSE F(1,44) p-value partial rl2 Observed power

PPVT-III 2736.54 169.733 16.123 .000** .268 .975

EVT 186.241 136.444 8.694 .005* .165 .822

CASL 3477.002 122.627 28.354 .000** .392 .999

Note. MS: mean square, MSE: mean q2= effect size. *: p < .01, **: p < .001


Table 4-7. Pairwise comparison results based on adjusted group means
Dependent Adjust means 95% CI for Difference
Standard ____________________
Variable A B difference p-value
errors Upper Bound Lower Bound
(A-B)

PPVT NH HI 16.651 4.147 .000** 8.294 25.009

EVT NH HI 10.963 3.718 .005* 3.470 18.457

CASL NH HI 18.769 3.525 .000** 11.666 25.873

Note. NH = normally hearing group, HI = hearing impaired group. *: p < .01, **: p < .001
All pairwise comparisons are based on estimated marginal means and all scores are adjusted for multiple
comparisons using Bonferroni's method. The mean difference is significant at the .05 level.









14 *PPVT
SEVT
140- O CASL
18
0

II
120- 6
6
8 41


SI 1
0 m Ma I f


NH HI


Figure 4-4. Clustered box plot of oral language measure for the NH and HI groups. Note: Dotted line indicates the mean score.









Table 4-8. Descriptive statistics and estimated adjusted groups means for phonological processing measures

group M SD Min Max Adjusted SE 95% CI
means Lower Upper
NH 12.10 2.410 7 16 12.080 .452 11.179 12.981
Elision HI 8.63 2.140 4 13 8.581 .571 7.443 9.719
RD 7.43 2.528 3 13 7.488 .466 6.559 8.416
NH 11.52 2.400 7 16 11.416 .409 10.601 12.232
Blend HI 7.79 1.584 4 11 7.567 .517 6.538 8.597
RD 8.53 2.345 3 13 8.771 .422 7.931 9.611
NH 11.17 2.406 7 18 11.064 .395 10.278 11.851
RAN-Digit HI 10.0 1.700 7 12 9.762 .498 8.769 10.755
RD 7.77 2.112 4 13 8.022 .407 7.211 8.832
NH 11.28 2.389 7 18 11.257 .429 10.403 12.111
RAN-Letter HI 10.05 2.094 6 14 10.012 .541 8.933 11.091
RD 7.93 2.258 4 14 7.977 .442 7.097 8.857
Memory for NH 11.07 2.685 6 17 11.078 .495 10.092 12.064
Digits# HI 9.11 2.622 5 14 9.125 .625 7.881 10.370
RD 9.10 2.551 4 15 9.079 .510 8.063 10.094
Nonword NH 12.41 2.292 8 17 12.433 .444 11.547 13.318
Repetition # HI 7.89 2.787 3 12 7.936 .561 6.818 9.055
RD 8.90 2.107 5 14 8.855 .458 7.943 9.768
Note. NH: normally hearing group, HI: hearing-impaired group, M: mean, SD: standard deviation, Min: minimum,


Max: maximum, SE: standard error, #: NH>HI=RD,


": NH=HI>RD.









Table 4-9. Summary of three univariate ANCOVAs on the phonological measures from the CTOPP.
MS MSE F(2,74) p-value partial f2 Observed power Group comparisons

Elision 155.657 5.807 26.803 .000 .420 1.000 NH > HI = RD

Blending 96.433 4.755 20.281 .000 .354 1.000 NH > HI= RD

RAN Digits 60.917 4.425 13.767 .000 .271 .998 NH = HI > RD

RAN Letters 70.881 5.218 13.584 .000 .269 .997 NH = HI > RD

Memory for Digits 34.694 6.949 4.993 .009 .119 .798 NH > HI = RD

Nonword Repetition 144.587 5.609 25.780 .000 .411 1.000 NH > HI= RD

Note. MS: mean square, MSE: mean square for error, partial 2 =effect size, df1 = 1, df2= 44










able 4-10. summary ot post noc pairwise comparisons ot pnonological measures (CT PP subtests).
95% CI for Mean
Dependent Variables A B Mean SE t Sig Differencea
Difference Lower
(A-B) Upper Bound Bound


Elision


Blend


RAN-Digit


RAN-Letter


Memory for Digits



NWR


NH
NH
HI
NH
NH
HI
NH
NH
HI
NH
NH
HI
NH
NH
HI
NH
NH


HI
RD
RD
HI
RD
RD
HI
RD
RD
HI
RD
RD
HI
RD
RD
HI
RD
RD


3.500
4.593
1.093
3.849
2.645
-1.204
1.302
3.043
1.741
1.245
3.281
2.035
1.953
1.999
.047
4.496
3.577
-.919


.716
.664
.766
.647
.601
.693
.625
.580
.669
.678
.630
.726
.783
.727
.838
.703
.653


4.888 .000**
6.917 .000**
1.427 .473
5.949 .000**
4.400 .000**
1.737 .260
2.083 .122
5.247 .000**
2.568 .033*
1.836 .211
5.208 .000**
2.803 .019*
2.494 .044*
2.750 .022*
0.056 1.00
6.395 .000**
5.478 .000**
1.220 .679


1
2

2.
1
-2
-.
1.
.1

1
.2
.2

-2
2
1
-2


.747 5.252
.965 6.220
784 2.970
.263 5.435
.172 4.118
.902 .4940
2280 2.832
.622 4.464
020 3.379
416 2.907
.738 4.823
:560 3.814
?350 3.870
:190 3.780
.006 2.100
.774 6.219
.978 5.177
.763 .9260


Note. *: p < .05, **: p < .001. NH = normally hearing group, HI = hearing impaired group, RD = Dyslexic group,
All comparisons are based on estimated marginal means using Bonferroni's method for multiple comparisons.











MElision
MBlend
15.0-



S 2.5-

0I
SI








46


2.5-


NH HI RD


Figure 4-5. Clustered boxplot for phonological awareness measures for the NH, HI and RD
groups. Note: Dotted line indicates the mean score.











Memory for digit
Nonword Repetition


SB
15- 0
58

5- -" 0+
0


















NH HI RD

Figure 4-6. Clustered boxplot for phonological short-term memory for the NH, HI, and RD
groups. Note: Dotted line indicates the mean score.











6 ORAN-Digit
o M RAN-Letter
17.5-

20
0

5.0-
S64
II

2.5"













5.0-


I I I
NH HI RD

Figure 4-7. Clustered boxplot for rapid naming for the NH, HI, and RD groups. Note: Dotted line
indicates the mean score.












Table 4-11. Descriptive statistics and estimated adjusted group means for literacy measures.

M SD Adjusted SE 95 % CI
means lower upper


Word
NH
HI
RD
Nonword
NH
HI
RD
Word
NH
HI
RD
Nonword
NH
HI
RD
NH
HI
RD


115.24
108.11
83.10

115.62
110.16
82.40

119.34
106.37
82.63

115.38
106.42
84.93
112.93
99.89
84.63


Comprehension
NH 110.62


HI
RD
Reading rate
NH
HI
RD


8.947
9.451
13.469

12.982
11.388
8.830

13.036
15.475
13.405

13.270
14.698
12.031
13.564
13.552
8.720

9.951


98.89 12.391
84.43 13.318


14.00 2.104


11.26
4.97


3.754
2.822


Reading accuracy
NH 14.55 2.759
HI 11.63 3.804
RD 5.90 2.820


Timed
reading
(TOWRE)


2.398
2.984
2.364

2.360
2.937
2.327

2.220
2.763
2.189

2.226
2.770
2.195
2.000
2.489
1.972


114.728
99.174
78.558

110.796
99.643
80.767

106.235
93.096
80.221

108.517
94.210
80.344
111.341
102.494
79.504


124.283
111.066
87.980

120.201
111.348
90.041

115.082
104.106
88.945

117.387
105.250
89.091
119.310
112.411
87.361


119.5
105.1
83.3

115.5
105.5
85.4

110.7
98.6
84.6

113.0
99.7
84.7
115.3
107.5
83.4

115.7
109.5
82.7


13.9
11.3
4.9


14.6
11.7
5.9


2.527
2.002


.530
.660
.523


.572
.712
.564


104.509
78.723


12.935
10.013
3.892


13.407
10.245
4.758


114.580
86.702


15.048
12.643
5.976


15.688
13.084
7.008


2.030 111.654 119.745


Untimed
reading
(WRMT)




Spelling


Passage
reading






14
14 0
0


*WRMT(Word)
*WRMT(Nonword)


I*****'


I.


I~


Il


100 -- -- -- ----- --


I
I TI


i ....I


Figure 4-8. Clustered boxplot for untimed word-level reading (Word Identification and Word
Attack subtests on the WRMT-R). Note: Dotted lines indicate + 1 SD.


62


....."...I





*TOWRE(Word)
*TOWRE(Nonword)


I."1


I.


.-


100*--I t


1 i I
...................mmlmmllmmm ml mn ml ml ml mmmlmmmmmm a1


I.......


Figure 4-9. Clustered boxplot for untimed word-level reading (Word Efficiency and Word
Decoding subtests on the WRMT-R). Note: Dotted lines indicate + 1 SD.


.I


I ..i.


;....................


I ....... I









I559
T...........


10 i -I MP- -


".......I1


EGORT-Rate
*GORT-Accuracy


Figure 4-10. Clustered boxplot for connected-text reading fluency (GORT-4 Rate and Accuracy)
(Note: Dotted lines indicate + 1 SD.


.......I


I.......I


I..........


I.I


SS
0 0


I ...... *

















120.




100- -




BO-
39



60-


62
o


*Passage Comp
*WRAT(Spell)


Figure 4-11. Clustered boxplot for passage comprehension (WRMT-R) and spelling (WRAT)
(Note: Dotted lines indicate + 1 SD).









Table 4-12. Summary of univariate ANCOVAs on the reading and spelling measures.

partial Observed Group
Dependent Variable MSB MSE F(2,74) p 2 Power comparison

Word 9767.786 166.657 58.610 .000 .613 1.000 NH > HI > RD


Untimed


Nonword

Word


Timed


Nonword

Spelling

Comp*
ge
Rate


ng


Accuracy


WRMT-R
6850.592

7955.022
TOWRE
8774.433

WRAT 5874.333

WRMT-R 5012.224

GORT 629.020

GORT 567.806


161.457

115.904

119.518

143.631

142.865

8.154

9.499


42.430

68.634

73.415

40.899

35.084

77.143

59.777


.000

.000

.000

.000

.000

.000

.000


.534

.650

.665

.525

.487

.676

.618


1.000

1.000

1.000

1.000

1.000

1.000

1.000


NH > HI > RD

NH > HI > RD

NH = HI > RD

NH > HI > RD

NH > HI > RD

NH > HI > RD

NH > HI > RD


Note. The F tests the effect of group in each ANCOVA. Covariate was grade in months.
MSB: mean squares for between group, MSE: mean square error


*: passage comprehension.


Passa;

readir










Summary of post hoc pairwise comparisons of eight reading and spelling measures.


Dependent
Variables

WRMTa

untimedd

word)

WRMTa

untimedd

nonword)

WRMT

(compre-

hension)

WRAT

(spelling)

TOWREb

(timed

word)

TOWREb

(timed

nonword)

GORT

(reading

rate)

GORT

(reading

accuracy)


Mean
A B Difference
(A-B)

NH HI 14.386

NH RD 36.237

HI RD 21.851

NH HI 10.003

NH RD 30.094

HI RD 20.091

NH HI 12.058

NH RD 26.076

HI RD 14.018

NH HI 13.222

NH RD 28.235

HI RD 15.013

NH HI 7.873

NH RD 31.893

HI RD 24.020

NH HI 6.155

NH RD 32.987

HI RD 26.832

NH HI 2.664

NH RD 9.058

HI RD 6.394

NH HI 2.883

NH RD 8.664

HI RD 5.782


t Siga


SE


3.833

3.365

3.825

3.772

3.312

3.765

3.548

3.115

3.541

3.558

3.124

3.551

3.196

2.806

3.190

3.246

2.849

3.239

.848

.744

.846

.915

.803

.913


3.75

10.77

5.71

2.65

9.09

5.34

3.40

8.37

3.96

3.72

9.04

4.23

2.46

11.37

7.53

1.90

11.58

8.28

3.14

12.17

7.56

3.1

10.9

6.3


.001**

.000***

.000***

.029*

.000***

.000***

.003***

.000**

.001*

.001*

.000**

.000**

.048*

.000***

.000***

.185

.000***

.000***

.007**

.000***

.000***

.007***

.000***

.000***


Note. a: timed test,
and years)


b: untimed test, *: p < .05,


** p <.01, *** p <.001, Covariate: grade(in months


95% CI for Mean
Difference
Upper Lower
Bound Bound
4.997 23.774

27.994 44.479

12.481 31.221

.762 19.243

21.981 38.207

10.869 29.314

3.366 20.750

18.444 33.707

5.343 22.693

4.507 21.938

20.583 35.887

6.314 23.711

.044 15.703

25.019 38.767

16.206 31.834

-1.795 14.106

26.007 39.967

18.897 34.767

.587 4.740

7.235 10.881

4.322 8.467

.641 5.124

6.697 10.632

3.545 8.019


Table 4-13.









Table 4-14. Correlations between the degree of hearing loss and phonological measures,
partialing out for age and grade.
... RAN- RAN- Memory Nonword
elision blending Digit Letter for Digit repetition
r -.644 -.721 -.367 -.396 -.287 -.597
PTA HI
sig. .007** .002** .162 .129 .281 .015*
(better
r -.203 .049 -.174 -.108 .322 .031
ear) NH
sig. .320 .814 .394 .598 .109 .880

Note. *p <.05, **p < .01



Table 4-15. Partial correlation matrix between phonology and reading (HI group).
untimed timed Passage

word nonword word nonword spell compare rate accuracy
hension
elision r .582 .751 .562 .611 .618 .610 .605 .711
sig .009 .000 .012 .006 .005 .006 .007 .001
blend r .189 .304 .193 .159 .379 .289 .470 .378
sig .241 .126 .237 .278 .074 .139 .033 .074
RAN- r .307 .411 .459 .560 .418 .341 .723 .635
Digits sig .124 .057 .037 .012 .054 .098 .001 .004
RAN- r .259 .301 .559 .656 .479 .314 .581 .571
Letter sig .166 .129 .012 .003 .030 .118 .009 .010
Memory r .160 .394 .287 .285 .558 .284 .525 .393
for Digits sig .277 .066 .141 .143 .012 .143 .018 .066
Nonword r .362 .594 .323 .324 .291 .324 .375 .322
Repetition sig .084 .008 .111 .110 .137 .111 .076 .112









Table 4-16. Partial correlation matrix between phonology and reading (NH group).
untimed timed Passage
Compre-
word nonword word nonword spell rate accuracy
hension
elision r .711 .692 .142 .474 .825 .520 .443 .671
sig .000 .000 .244 .007 .000 .003 .012 .000
blend r .378 .472 -.023 .326 .511 .185 .116 .392
sig .028 .007 .456 .052 .004 .183 .286 .024
RAN- r .421 .382 .559 .596 .311 -.018 .382 .465
Digits sig .016 .027 .001 .001 .061 .465 .027 .008
RAN- r .414 .437 .542 .629 .399 .006 .385 .540
Letter sig .018 .013 .002 .000 .022 .489 .026 .002
Memory r .657 .603 -.106 .305 .519 .502 .185 .377
for Digits sig .000 .001 .302 .065 .003 .005 .182 .029
Nonword r .686 .650 .137 .366 .534 .484 .433 .558
Repetition sig .000 .000 .253 .033 .002 .006 .014 .002









Table 4-17. Correlation matrix between better ear's PTA and phonology, partialing out age (in months).
Control Memoryfor Nonword
Variables Elision Blend RAN-Digit RAN-Letter Memory for
Vanables Digits Repetition
age_month Better ear r -.473 -.597 .120 .085 .116 -.424
PTA p .024 .004 .318 .368 .323 .040
Better ear r -.565 -.528 -.012 .032 .166 -.438
SRT p .007 .012 .481 .449 .255 .034
r 1.000 .608 .320 .215 .326 .579
Elision
p .004 .097 .196 .094 .006
r .608 1.000 .225 .160 .415 .461
Blend
p .004 .185 .263 .043 .027
r .320 .225 1.000 .868 .197 .149
RAN-Digit
p .097 .185 .000 .216 .278
r .215 .160 .868 1.000 .126 -.004
RAN-Letter
p .196 .263 .000 .309 .494
Memory for r .326 .415 .197 .126 1.000 .502
Digits p .094 .043 .216 .309 .017
Nonword r .579 .461 .149 -.004 .502 1.000
Repetition p .006 .027 .278 .494 .017









Table 4-18. First-order correlation matrix between oral language and phonology measures.
Memory for Nonword
Elision Blend RAN-Digit RAN-Letter Digit Repetition
PPVT r .324 .362 -.063 -.130 .533 .406

p .102 .077 .405 .309 .014 .053
EVT r .252 .224 .112 -.044 .592 .359

p .164 .194 .335 .434 .006 .078
CASL r .511 .443 .210 .162 .558 .494

p .018 .038 .209 .267 .010 .022









Table 4-19. Correlation matrices among auditory processing measures, and phonology and reading scores, partialing out for age,
grade, duration of hearing aid use, and nonverbal intelligence.
Phonological measures Reading measures

PA RAN STM Untimed timed Passage reading

elision Blend Digit Letter MD NWR Word NW Word NW Spell Comp Rate Accu

SCAN1 .354 .410 -.022 -.012 .143 .409 -.135 -.028 .294 -.100 -.191 -.110 -.270 -.225

(FW) .196 .129 .939 .965 .610 .130 .633 .921 .287 .722 .496 .696 .330 .420

SCAN2 .087 .432 -.131 -.333 .430 .563 -.193 -.155 .075 -.361 -.323 -.118 -.158 -.404

(AFG) .757 .108 .641 .225 .110 .029 .491 .580 .790 .187 .240 .676 .575 .135

SCAN3 .222 .451 -.105 -.250 .653 .579 -.215 -.120 -.059 -.152 -.192 -.272 -.264 -.514

(CW) .426 .091 .711 .368 .008 .024 .442 .671 .835 .589 .493 .327 .341 .050

DD_perc .714 .659 .391 .069 .617 .548 .306 .546 .234 .525 .479 .322 .439 .261

.003 .008 .149 .808 .014 .034 .267 .035 .400 .044 .071 .242 .102 .348

Note. MD: Memory for Digit, NWR: Nonword Repetition, Comp: Comprehension, FW: Filtered Words,
AFG: Auditory Figure-Ground, CW: Competing Words, DD_perc: Dichotic Digits in percentile, Accu: Accuracy











Table 4-20. Correlation matrix between oral language and reading measures.
PPVT EVT CASL

Word Identification .286 .283 .203
(WRMT-R) .151 .153 .234

Word Attack .247 .361 .291
(WRMT-R) .187 .093 .147

Passage .474 .536 .442
Comprehension
(WRMT-R) .037 .020 .049
.187 .319 .238
Spell (WRAT-3)
.252 .123 .197
Word Efficiency .368 .127 .363
(TOWRE) .088 .326 .092
Word Decoding -.025 .128 .129
(TOWRE) .465 .325 .324
.271 .442 .194
GORT-4 Rate
.164 .050 .245
.092 .189 .129
GORT-4 Accuracy
.372 .250 .323

Note. Age, grade, duration of hearing aid use, and nonverbal
intelligence were partialed out.









Table 4-21. Hierarchical regressions of variables separately predicting untimed word reading (the Word Identification subtest on the
WRMT-R).
B SE B p t sig R2 LR2 AF


Step]
Age
Better ear's PTA
Nonverbal IQ
Step 2


.319


Elision 4.673 1.395 .646

Blending 3.260 2.517 .334

RAN-Digits 4.061 2.156 .446

RAN-Letter 3.055 1.903 .413

Memory for Digits 1.662 1.263 .282

Nonword Repetition 2.459 1.169 .443

Dichotic Digits .380 .211 .445

Note. B: estimated beta, SE B: standard error of estimated beta,
AR2: amount of increased multiple correlation coefficient.


3.351 .005 .622

1.295 .216 .392

1.883 .081 .457

1.606 .131 .425

1.316 .209 .394

2.103 .054 .482

1.803 .093 .447

3: estimated standardized beta, R2:


.303

.073

.138

.106

.075

.164

.128

multiple correlation,


11.227

.216

3.547

2.578

1.732

4.423

3.249


I









Table 4-22. Hierarchical regressions of variables separately predicting untimed nonword reading (the Word Attack subtest on the
WRMT-R).

B SE B 3 t sig R2 AR2 AF

Stepi

Age
Better ear's PTA .110
Nonverbal IQ
Step 2


Elision 5.873 1.291 .855 4.550 .000 .641 .531 20.698

0 Blending 4.008 2.686 .432 1.492 .158 .233 .122 2.227

RAN-Digits 4.495 2.329 .520 1.930 .074 .297 .187 3.726

RAN-Letter 2.636 2.134 .396 1.235 .237 .198 .087 1.525

Memory for Digits 2.505 1.290 .447 1.943 .072 .299 .189 3.774

Nonword Repetition 3.513 1.113 .666 3.156 .007 .480 .370 9.957

Dichotic Digits .539 .209 .665 2.576 .022 .397 .286 6.637

Note. B: estimated beta, SE B: standard error of estimated beta, 0: estimated standardized beta, R2: multiple correlation,
AR2: amount of increased multiple correlation coefficient.











Table 4-23. Hierarchical regressions of variables separately predicting timed word reading (the Word Efficiency on the TOWRE).

B SE B t sig R2 AR2 AF

Step]


Age
Better ear's PTA

Nonverbal IQ

Step 2


.193


Elision 2.780 .999 .629

Blending 1.794 1.705 .301

RAN-Digits 2.953 1.397 .531

RAN-Letter 2.840 1.148 .629

Memory for Digits 1.180 .832 .327

Nonword Repetition 1.366 .814 .403

Dichotic Digits .159 .150 .305

Note. B: estimated beta, SE B: standard error of estimated beta,
AR2: amount of increased multiple correlation coefficient.


2.784 .015

1.053 .310

2.113 .050

2.473 .027

1.418 .178

1.678 .115

1.064 .306

3: estimated standardized beta, R2:


.481 .287

.252 .059

.388 .195

.438 .245

.294 .101

.328 .135

.253 .060

multiple correlation,


*


7.748

1.108

4.465

6.117

2.010

2.817

1.131


I












Table 4-24. Hierarchical regressions of variables separately predicting timed nonword reading (the Word Decoding on the TOWRE).


B SE B 3 t sig R2 AR2 AF

Stepi


Age
Better ear's PTA
Nonverbal IQ
Step 2


.097


Elision 3.809 1.217 .716

Blending 1.990 2.194 .277

RAN-Digits 4.510 1.653 .673

RAN-Letter 4.328 1.319 .796

Memory for Digits 1.484 1.063 .342

Nonword Repetition 1.723 1.039 .422

Dichotic Digits .364 .173 .579

Note. B: estimated beta, SE B: standard error of estimated beta,
AR2: amount of increased multiple correlation coefficient.


3.130 .007

.907 .147

2.728 .016

3.280 .005

1.397 .184

1.658 .120

2.105 .054

3: estimated standardized beta, R2:


.469 .372

.147 .050

.411 .313

.489 .392

.207 .110

.245 .148

.414 .217

multiple correlation,


9.795

.823

7.445

10.761

1.951

2.748

4.431












Table 4-25. Hierarchical regressions of variables separately predicting timed passage reading rate on the GORT-4.

B SE B 3 t sig R2 AR2 AF

Stepi


Age
Better ear's PTA

Nonverbal IQ

Step 2


.238


Elision

Blending

RAN-Digits

RAN-Letter

Memory for Digits

Nonword Repetition

Dichotic Digits


1.110

1.198

1.628

1.011

.640

.506

.132


.378

.604

.441

.458

.298

.317

.049


.633

.505

.737

.564

.447

.375

.636


Note. B: estimated beta, SE B: standard error of estimated beta, 0:
AR2: amount of increased multiple correlation coefficient.


2.938 .011

1.982 .607

3.690 .002

2.210 .044

2.149 .050

1.596 .133

2.705 .017

estimated standardized beta


.528 .291

.405 .167

.613 .376

.435 .197

.427 .189

.355 .117

.499 .262

R2: multiple correlation.


8.631

3.929

13.614

4.884

4.616

2.547

7.317


I












Table 4-26. Hierarchical regressions of variables separately predicting timed passage reading accuracy on the GORT-4.

B SE B 3 t sig R2 AR2 AF

Stepi


Age
Better ear's PTA

Nonverbal IQ
Step 2


.183


Elision

Blending

RAN-Digits

RAN-Letter

Memory for Digits

Nonword Repetition

Dichotic Digits


1.366

1.047

1.514

1.085

.520

.482

.180


.348

.660

.509

.476

.332

.337

.056


.768

.436

.676

.596

.359

.353

.515


Note. B: estimated beta, SE B: standard error of estimated beta, 0:
AR2: amount of increased multiple correlation coefficient.


3.929 .002

1.586 .135

2.977 .010

2.279 .039

1.565 .140

1.427 .175

1.931 .074

estimated standardized beta


.611 .428

.307 .124

.500 .317

.404 .221

.305 .122

.287 .104

.355 .172

R2: multiple correlation.


15.435

2.514

8.860

5.195

2.449

2.038

3.727


I












Table 4-27. Hierarchical regressions of variables separately predicting scores on the Spelling on the WRAT.

B SE B 3 t sig R2 AR2 AF

Stepi


Age
Better ear's PTA

Nonverbal IQ
Step 2


.183


Elision

Blending

RAN-Digits

RAN-Letter

Memory for Digits

Nonword Repetition

Dichotic Digits


4.322

4.091

3.816

3.377

2.849

1.790

.495


1.375

2.310

2.079

1.770

1.034

1.195

.181


.683

.478

.479

.522

.551

.368

.663


Note. B: estimated beta, SE B: standard error of estimated beta, 0:
AR2: amount of increased multiple correlation coefficient.


3.143 .007

1.771 .098

1.836 .088

1.908 .077

2.755 .015

1.498 .156

2.735 .016

estimated standardized beta


.521 .338

.332 .150

.341 .159

.351 .169

.470 .287

.296 .113

.467 .285

R2: multiple correlation.


9.877

3.137

3.369

3.642

7.592

2.245

7.480


I











Table 4-28. Summary of hierarchical regressions on timed reading measures, controlled for phonological awareness and short-term
memory scores (showing the results from Step 3 only).

Criterion variables Entered exploratory variable (RAN-Digit) in Step 3
B SE B 3p t Sig. R2 AR2

Reading rate (GORT-4) 1.522 .425 .689 3.583 .005 .816 .236

Reading accuracy (GORT-4) 1.270 .471 568 2.695 .022 .779 .160

Words (TOWRE) 2.449 1.592 .440 1.538 .155 .592 .096

Non-words (TOWRE) 4.259 1.666 .636 2.556 .029 .692 .201



Criterion variables Entered exploratory variable (RAN-Digit) in Step 3
B SE B 3p t Sig. R2 A R2

Reading rate (GORT-4) .927 .396 .517 2.340 .041 .728 .149

Reading accuracy (GORT-4) .945 .370 .520 2.553 .029 .770 .150

Words (TOWRE) 2.786 1.035 .617 2.691 .023 .708 .212

Non-words (TOWRE) 4.368 .895 .803 4.882 .001 .850 .358

Note. B: estimated beta, SE B: standard error of estimated beta, 0: estimated standardized beta, R2: multiple correlation,
AR2: amount of increased multiple correlation coefficients.









CHAPTER 5
DISCUSSION

Introduction

This present study sought to explore four major questions concerning the specific link

between mild to moderate SNHL and reading and spelling abilities.

(1) Are there significant effects of SNHL on hard-of-hearing children's skills of spoken

language, phonological processing, and reading/spelling?

(2) What are the patterns of correlations among reading, phonological ability, and spoken

language? Which phonological variables are associated with which types of reading tasks

in the HI group?

(3) Are phonological processing skills important in reading achievement of hearing impaired

readers? Specifically, do they predict unique variance in reading and spelling?

Oral Language Skills Findings

The first purpose of the present study was to replicate the results of previous studies on the

spoken language ability of the HI children compared to children with normal hearing. There are

only a few scientific studies of the language skills of children with mild to moderate SNHL

(Davis, Shepard, Stelmachowicz, & Gorga, 1981; Gilbertson & Kamhi, 1995; Davis, Elfenbein,

Schum, & Bentler, 1986; Blair, Peterson, & Viehweg, 1985; Blamey, Sarant, Paatsch, Barry,

Bow, Wales, 2001; Gilbertson and Kamhi, 1995; Hansson et al., 2004; Stelmachowicz et al.,

2004). Most previous studies, conducted on children with a range of linguistic skills, revealed

that a significant proportion of children with mild to moderate SNHL have depressed vocabulary

and morphosyntactic performance.









Group Comparison

The investigator's goal was to determine whether mild to moderate SNHL has significant

effects on hard-of-hearing children's vocabulary and grammatical knowledge and, if so, to what

degree, and in what areas when controlling for non-verbal IQ and grade.

As indicated in Delage and Tuller (2007), SNHL is associated with not only lowered hearing

thresholds, but also distortion of speech signals and degraded language input, factors that are

assumed to affect age-appropriate development of linguistic knowledge. Considering the fact that

normal speech recognition development is contingent upon successful auditory input during a

constrained critical period in the first few years of life (e.g. McConkey Robbins, Burton Koch,

Osberger, Zimmerman-Phillips & Kishon-Rabin, 2004; Sharma, Dorman & Spahr, 2002), one

would expect that many children with mild to moderate HL are at risk for delayed language

development.

The average age for initial hearing aids fitting of the HI subjects included in this study was

close to 4 years (mean: 44.4 months). Researchers have indicated that although language input

can be noticeably improved through hearing aids, mild to moderate level of SNHL in children is

detected and remediated at relatively older ages than profound hearing loss, with the average age

for remediation between 4 and 5 (Hansson, Forsberg, Lofqvist, Maki-Torkko, & Sahlen, 2004;

Stelmachowicz, Pittman, Hoover, & Lewis, 2004; Tuller & Jakubowicz, 2004). Thus, due to

perceptual distortion of phonological input during the critical language acquisition period, if not

aided early, hard-of-hearing children are likely to experience inaccurate speech perception and

develop unclear phonological representations of words, and delayed of phonological sensitivity

(awareness). Since phonological awareness is required for efficient phonemic segmentation,

which is necessary for distinct phonological representations of newly learned words, hearing loss

may ultimately cause language impairments (Chiat, 2001).









Consistent with previous studies, the findings described in this study also indicate that the

group of aided, school-aged children with mild to moderate SNHL have significantly lower

spoken language skill in all tests areas than would be expected for children with normal hearing.

Adjusted means for the HI groups were more than 1 SD below the mean of the NH children for

both receptive vocabulary and grammatical knowledge tests: 16.651 for the PPVT-III and 18.769

for the CASL (See Table 4-5). The largest group difference was noted for morphosyntactic

knowledge task, as evidenced by a strongest effect size of .392. Although the HI group's

performance on the PPVT-III, EVT, and CASL tests was not impaired below the mean score of

100 (See Figure 4-4), further inspections of raw scores in each test showed that 31% of the HI

children had standard scores of 1 SD below the mean and only 26% of them had raw scores

above the mean.

Findings of oral language skills in this study are different from those reported by Briscoe et

al. (2001), who suggested that the vocabulary and receptive grammar of HI children were not

significantly lower than their control group matched on age. One possible reason for this

discrepant result is associated with the statistical treatment of the data. In Briscoe et al. (2001),

multiple univariate ANOVAs were used to investigate the group effect instead of multivariate

analyses which can control the effects of confounding covarying factors (i.e., age, grade, or IQ).

Secondly, Briscoe and colleagues' (2001) sample was comprised primarily of children with mild

hearing loss; only 3 children had moderate HL. Further, Briscoe et al. reported that their HI

subjects can be divided into two subgroups based on the severity of phonological impairments:

unimpaired and impaired groups. The latter subgroup showed significantly poorer vocabulary

skills than the group with unimpaired phonological skills. Close scrutiny of their data showed

that the impaired group's scores on vocabulary were at or below 1 SD below the normative mean









scores: 86.6 (receptive) and 84.6 (expressive), respectively. Interestingly, these two subgroups

did differ significantly in the severity of hearing loss, that is, the impaired group's hearing loss

was more severe across all frequencies. Based on the graphic display of the hearing thresholds

(Briscoe et al., 2002:337), the unimpaired group's PTA based on 0.5, 1, and 2 KHz was around

25 dB HL. This strongly suggests that the unimpaired group's oral language skills were not

affected by this mild level of hearing loss. Based on these observations, it can be inferred that the

phonologically unimpaired group's normal level of language skills masked the negative effect of

hearing loss on spoken language ability causing overall mean scores of the HI group to be

exaggerated. That is, significant differences might have been captured if the sample of the HI

group was more representative of the population with mild to moderate SNHL. Lastly, existing

significant difference could have been observed when confounding variables such as age, grade,

or non-verbal IQ were statistically controlled.

Interrelationships between Phonological Processing and Language Skills

The investigator also tried to determine whether linguistic knowledge was related to HI

subjects' phonological processing skills that include the ability to encode, store, and retrieve

phonological information.

Phonological short-term memory and vocabulary

Previous research has consistently indicated a close association between children's

phonological short-term memory (STM) and their vocabulary knowledge (Baddeley, Gathercole,

and Papagno, 1998; Gathercole & Baddeley, 1990; Gathercole, Hitch, Service, & Martin, 1997;

Jarrold, Baddeley, Hewes, Leeke, & Phillips, 2004). Baddeley et al. (1998) argued that the

central purpose of verbal STM to facilitate the efficiency of the long-term learning of vocabulary.

For example, learning an unfamiliar word requires ordering the constituent phonemes of the

word along with its accurate phonological representation for storage in memory. They suggested









that phonological material is first represented in phonological STM, which makes it available for

long-term learning. As reviewed in chapter 2, the authors described the phonological loop as a

language learning device and argued that the ability to retain small amounts of phonological

information in STM evolved as a human characteristic that has a selection advantage for

language acquisition.

Gupta (2003) provided a similar explanation when he suggested that during learning of

unfamiliar phonological sequences for new word acquisition, incoming phonemes are stored in a

specific short-term memory system, called 'sequence memory'. This stored information permits

the replay or verbal rehearsal of the new phonological sequences and increases the probability

that phonological representation in the long-term memory will be stable.

Findings from the current study support this word learning hypothesis. In fact, the subjects'

STM tested by a digit span and a nonword repetition test, was strongly correlated with

vocabulary (See Table 4-18). Especially, the Memory for digits was significantly associated with

all vocabulary and morphosyntactic with partial coefficients ranging from .533 to .592. The

Nonword repetition was highly correlated with morphosyntax measure (partial coefficient = .494,

p = .022) and partial coefficients for receptive and expressive vocabulary measures approached

the significant level (r = .406, p = .053 with the PPVT-III and r = .359, p = .078 with the EVT,

respectively). In Baddeley and colleagues'(1998) study of memory in 6-year old children, first-

order correlation coefficients between two measures of verbal STM (auditory digit span and

nonword repetition) and vocabulary knowledge were .44 and .56. Similarly, in Gathercole and

colleagues' (1997) study, digit span scores correlated significantly with all tested vocabulary

measures, with significant correlations ranging from .38 to .44 (p < .01).









Grammatical knowledge and phonological processing

In a study of short-term memory in children with specific language impairment (SLI),

Gathercole and Baddeley (1990) developed a theoretical framework linking deficits in

phonological STM (i.e., nonword repetition) to impaired language acquisition. The authors

pointed out that poor performance on nonword repetition is a marker of specific language

impairment. However, as pointed out by Leonard (1998), vocabulary is typically less impaired

than syntax in children with SLI (Leonard, 1998). This raises the question of whether

phonological STM is implicated in acquisition of syntactic knowledge as well as vocabulary.

Considerably less work has addressed the role of phonological STM in the acquisition of

syntax, even though there is increasing support for the notion that the same psychological

mechanisms underpin vocabulary and syntax acquisition (Marchman & Bates, 1994). In their

1998 study, Baddeley et al. indicated that in typically developing children, nonword repetition, a

measure of phonological loop capacity, was related not only to vocabulary learning but also to

acquisition of syntax. In this respect, the HI group's performance on the Grammatic Knowledge

subtest in this study is of particular interest since it was strongly associated all of the

phonological awareness and phonological memory tests (range of partial coefficients: .443

-.558).

How would we explain this strong association between phonological STM, phonological

awareness, and morphosyntactic knowledge in the HI group? Gathercole (2006) offers three

hypotheses to explain the link between weak phonological STM and poor morphosyntax. The

first hypothesis relates to a storage capacity, proposing that incoming speech stream needs to be

stored in a temporary buffer while the morphosyntactic analysis is carried out. To process the

grammatical morpheme, each phoneme should be perceived well and the inflection marker must

be separated from the content part. Next, its grammatical function must be realized based on the









utterance situation. All of this must be completed before the next item in an utterance enters the

system. So, perceived material may be lost owing to restrictions in working memory or because

previous material is still being processed. That is, an inflected verb will be fully processed some

times and at other times it will not. Therefore, if stored phonological material is rapidly decayed,

subsequent operations will become inefficient.

The second hypothesis is related to the contribution of phonological awareness to

morphosyntactic awareness. It proposes that poor morphosyntactic skills will follow if a child

persists in segmenting or identifying incoming speech at the supra-phonemic level such as

syllable. For instance, past tense markers in monosyllabic words such as "walked," "hopped,"

and "laughed" may not be recognized if the child is only processing incoming words at syllable

or onset-rime borders, not being able to separately recognize the word-final grammatical

morphemes. Thus, underdeveloped phonological awareness can cause the skills for extracting a

morphosyntactic rule, such as past tense formation to be delayed. That is, children's

phonological sensitivity (awareness) will help to successfully divide a word into content and

bound morphemes (phonemic segmentation), enhancing their ability to perceive and manipulate

morphological units, too. The high association found in this study between phonological

awareness and knowledge of grammatical morphemes (r = .511 for the Elision and r = .443 for

the Blending) well supports this theoretical standpoint.

The third and final hypothesis addresses the possible challenges faced by some children to

perceive English grammatical morphemes that have low-perceptual salience (Rom & Leonard,

1990), which might result in weaker or low-quality representation of grammatical markers such

as '-ed' (past tense) and '-s' (plural or third person singular marker). Recent data from children

with cochlear implants suggests that the pattern of language development is strongly influenced









by the perceptual prominence of relevant morphological markers (Svirsky, Stallings, Lento, Ying,

and Leonard, 2000). As indicated above, lower-quality phonological representations resulting

from inaccurate speech perception may cause slowed morphosyntactic computation. Although

the speech of children with mild to moderate hearing loss is intelligible, research indicated that

aided speech perception would be still problematic. For instance, Elfenbein et al's (1994) study

reported that misarticulation of fricatives and affricates was common, particularly for children

with 3-frequency pure-tone average thresholds greater than 45 dB HL.

Thus, as hypothesized in Nobury et al. (2001), even mild hearing impairment might result

in morphemes of low perceptual salience being missed and therefore delaying grammatical

development.

Phonological Processing Findings

We know that an intact phonological system provides an important foundation for learning

to read in normally hearing children (Nathan, Stackhouse, Goulandris, & Snowling, 2004;

Rayner, Foorman, Perfetti, Pesetsky, & Seidenberg, 2001). This present research was aimed at

investigating what happens to the hard-of-hearing children's phonological system when auditory

stimulus signals are distorted. Specifically, this study was designed to investigate which part of

the phonological system is affected and which part is preserved (i.e., the whole phonological

system, or sub-processes such as some aspects of the metaphonological skills, phonological

memory, or lexical access/retrieval skills?)

Previous studies have reported that, overall, phonological processing skills of children with

mild to moderate SNHL are significantly lower than those children without hearing loss.

Surprisingly, however, little information is known regarding (1) the degree to which these skills

are impaired and (2) the range of phonological skills affected.









Furthermore, no previous studies have compared HI children's phonological skills with

those of dyslexic children at the varying range of phonological processing skills. Since dyslexic

children are known to have persistent deficits in phonological processing with no auditory

perceptual limitations related to cochlear damage, comparing these two groups should help to

elucidate the nature and degree of deficit in HI children's phonological processing skills.

The fundamental hypothesis underlying this study was that hearing loss would affect HI

children's auditory perceptual accuracy, leading to phonological processing abilities. In general,

phonemic manipulations become very difficult if the relevant segments are not distinctly

specified in the representation of the words (e.g. Elbro, 1996, 1998; Elbro, Borstrom & Petersen,

1998; Foy & Mann, 2001; Fowler, 1991; Griffith & Snowling, 2002; Swan & Goswami, 1997a;

Wesseling & Reitsma, 2001). Group data in this study revealed statistically significant

differences between the HI and NH groups on phonological awareness and phonological STM

skills. HI children had lower scores on tasks of: Elision (p < .0001), Blending (p < .001),

Memory for digits (p < .05), and Nonword repetition (p < .0001). Adjusted mean differences (NH

minus HL) were more than 1 SD for the Elision (3.50), Blending (3.849), and Nonword

repetition (4.496) measures. Associated with this depression, the HI group's skills on

phonological awareness and STM were not better than those of the RD group.

The results of this present investigation are only partially in line with previous studies.

Most of research indicated that mild to moderate hearing SNHL would adversely affect

phonological processing ability. For instance, in Briscoe et al. (2001), children with SNHL were

significantly lower than normal age controls on phonological awareness (as measured by onset

and rhyme matching), phonological discrimination, and nonword repetition. Interestingly, they

found that hearing-impaired children's digit recall was not lower than that of normal controls. In









this study, however, the HI group's phonological STM skill was significantly lower than the NH

group on both tasks. But, the Nonword repetition task showed bigger group mean difference than

the Memory for Digit task. One way of explaining this result is that the of degree of familiarity of

the phonological representation for digit names (e.g., 'one', 'two', ...) should place fewer

demands on long-term memory than the nonword repetition task (Bruck, 1992; Bruck & Treiman,

1990). Since immediate nonword repetition is highly dependent on verbal memory function, it

should be more sensitive than digit span in detecting any existing memory deficit.

In contrast, this present study found no significant differences between the HI and NH

groups on the alphanumeric RAN tasks. At the same time, the HI group showed significantly

better rapid naming skills than the RD children. This remarkable finding suggests that hearing-

impaired students' lexical access and retrieval skills are preserved in spite of their impaired

phonological awareness and short-term memory skills.

This finding was supported further by correlations which revealed that degree of hearing

loss was only associated with phonological awareness and short-term memory skills, not with the

RAN measures. Specifically, when chronological age (in months) was controlled, better ear's

PTA and SRT showed moderate to high negative correlations with the Elision, Blending, and

Nonword repetition. This tight association suggests that hearing and speech perception would

impact the HI children's phonologically-based skills.

Gibbs (2004) also reported similar negative correlation levels between hearing thresholds

and phonological awareness for rhyme awareness and initial phoneme awareness. As noted

above, no significant relation was noted between the RAN tasks and the scores of better ear's

PTA or SRT. This lack of significant association indicates that peripheral hearing loss does not

affect the efficiency of fast lexical access and retrieval or speed of processing.









Hearing Loss and Phonological Representation

It is now well established that early reading problems are associated with difficulties in

phonological domain of language (Mann, 1998; Stanovich, 1993; Wagner & Torgesen, 1987)

and most researchers view weak phonological awareness skills as contributing directly to

dyslexics' deficiency in word recognition skill (Blachman, 1989; Bowey, Cain, & Ryan, 1992).

However, despite the strong relationship between phonological skills and reading, the nature of

the phonological deficits that underlie poor reading remains to be elucidated.

Elbro and Jensen (2005) proposed that dyslexic readers' problems with phonemic

awareness are due to can be explained by their poorly specified phonological representations. In

line with Elbro and Jensen (2005)'s proposal, other researchers have suggested that the

relationship between impaired phonological awareness and reading difficulty be considered from

the perspective of the quality (or 'distinctness' as in Elbro, 1998) of phonological representations

of words in the mental lexicon. This hypothesis, referred to as the 'phonological representation

hypothesis,' states that a lack of distinctness or segmental specificity in phonological

representations cause phonological processing difficulties (e.g. Elbro, 1996; Fowler, 1991;

Hansen and Bowey, 1994; Metsala, 1997; Snowling etal., 1986; Swan and Goswami, 1997).

Fowler (1991) has proposed that words in each individual's mental lexicon vary according

to the precision of the phonological specification of the underlying representations. So, words

with less detailed representations are those that lack full segmental organization into a sequence

of discrete phonemic elements. As a consequence, segmental manipulation of the phonetic form

of the word may be hampered as a "crystallized" phonetic code is not available. Snowling and

Hulme (1989), Hulme and Snowling (1992), and McDougall, Hulme, Monk, and Ellis (1994)

have also suggested that an awareness of the phonological segments of words may not be as

critical to success on phonological awareness tasks as the accuracy of the underlying









phonological representations of the words. In accordance with Fowler, they argue that if the

integrity of a person's phonological representations is compromised, it follows that the person

should have difficulty performing segmental operations on those representations. Similarly,

Manis et al. (1997) noted that if children cannot perceive clear distinctions between phonemes, it

will be hard for them to develop representations that can be easily accessed to further processing.

Several lines of evidence also have suggested that poor phonological awareness relates to

poor speech perception. For example, consonant perception in newborns is associated with

language competence at 3 to 5 years, is predictive of reading achievement at 8 years

(Scarborough, 1990). Children and adults with reading difficulties have difficulties perceiving

speech in noise (Brady, Poggie, & Rapala, 1989; Brady, Shankweiler, & Mann, 1983), and

children who are poor readers require a longer segment of a gated word in order to perceive it

correctly (Metsala, 1997).

In sum, problems with accessing phonological representations, in turn, lead to difficulty in

segmenting and manipulating phonemes (phonological awareness). Thus, knowing that children

with congenital SNHL may experience degraded auditory input that interferes with their speech

perception (Abraham, 1993) and even small speech perception problems may be very important

for the ability to form fully specified phonological representation (Manis et al., 1997), it is

expected that hypothesized that hard-of-children with prelinguistic auditory disadvantage should

thus have lexical representations that are less distinct or accurate, leading to difficulties in

phonological awareness. The HI group's weak phonological awareness skills found in this study

support this theoretical position.

Phonological Representation and Short-term Memory Skill

The results of this study replicated previous studies showing that hearing loss can also lead

to inefficient storage capacity. However, while the measures on nonword repetition and digit









span are good markers of depressed phonological STM, it is not clear why hearing loss should

affect the capacity of verbal memory.

Research has documented that phonological STM is also constrained by the quality of

stored material. Service, Maury, and Luotoniemi (2007) explored the idea that good verbal STM

depends on the quality of representations of phonological sequences that are encoded into the

phonological store. Generally speaking, this idea is justified because the efficiency of

phonological segments in temporary storage should be maximized if the material to be stored is

as distinct or clear as possible. As Nairne (1990) and Neath & Nairne (1995) suggested,

phonological representations of good quality are less prone to being overwritten and should not

quickly decayed. Of course, such better-quality traces will serve as a good resource for more

efficient and accurate phonological awareness.

The present findings of reduced STM in children with SNHL appear to support the theories

that predict a relationship between the quality of phonological representations and STM. That is,

individuals with hearing loss whose auditory-phonetic inputs are not clearly represented in STM

have smaller spans. As they can hold less material of good quality in working memory, they

would also not be as good at manipulating stored phonemes and passing information from the

phonological store to the articulatory control process as would children with normal hearing do.

Articulatory control process is an active process component in Baddeley's memory model that

can keep the contents of the phonological store active and counteract time-based decay of

phonological traces by a kind of inner speech (Baddeley, 2003).

Intact Rapid Naming

In the present study, the rapid naming skills of HI children with mild to moderate SNHL

were preserved in spite of their impaired phonological awareness and phonological memory

abilities. While rapid naming predicts reading skills, the nature of the RAN as a measure









phonological skill is still unclear. Wagner and Torgesen (1987) argued that rapid naming is a

phonologically-based skill. Wagner, Torgesen, and Rashotte (1999) proposed that young readers

first retrieve sounds associated with letters or letter pairs, then retrieve the pronunciations of

word segments, and finally retrieve the whole word.

From this perspective, rapid naming should predict reading skill because this task measures

the efficiency of retrieving phonological codes associated with phonemes, word parts, or entire

words (Shankweiler & Crain, 1986; Share, 1995; Torgesen & Burgess, 1998). To them, RAN

tasks are an index of the speed with which phonological information can be accessed from

memory and, thus, are best described as tapping into an aspect of phonological processing.

However, other researchers have argued that rapid naming is not exclusive a phonological

skill and that it requires several other skills, including executive functioning (Denckla & Cutting,

1999), the ability to detect and represent orthographic redundancy (Bowers & Wolf, 1993; Wolf

& Bowers, 1999), global processing efficiency (Kail, Hall, & Caskey, 1999), and attention skill

(Neuhaus, Foorman, Francis, & Carlson, 2001).

In a similar context, Wolf and Bowers (1999) have proposed that processes related to serial

naming speed represent a second core-deficit in children with RD. This alternative model, known

as the double-deficit hypothesis, presumes that phonological processing and rapid automatized

naming (RAN) deficits are separable sources of reading dysfunction. As discussed in chapter 2,

three separate subtypes of RD individuals (i.e. phonological-only, rate-only, and double deficit)

can be predicted based on the varying effects of deficits in phonological processing and RAN.

Among these, the phonological-deficit-only subtype is characterized as having significant

deficits in phonological processing (PA and STM) with otherwise intact naming speed processes.









This present study revealed that the HI group's performance on the rapid naming tasks was

well preserved and hearing loss differentially affected their phonologically-based skills only (PA

and STM). Hence, their profile is, at least to some degree, characteristic of the single

(phonological-only) deficit group described by Wolf and Bower (1999). The RD group who

showed depression in both rapid naming and phonologically based skills (PA and STM) is

characteristic of double-deficit group.

Literacy Findings

Based on the findings discussed in the previous section, it was hypothesized that if poorer

hearing in the HI children influenced their phonological processing performance, then their

performance on a range of reading and spelling tasks will also be affected. This position is

consistent with the well-recognized phonological core deficit hypothesis. It was also

hypothesized that the correlations between phonological processing ability and reading

performances would be significant.

Group Comparisons

Although the HI children's performances on literacy measures were within normal range

on all tasks using standardized test scores as the criterion, statistically significant quantitative

differences between the NH and HI groups were found. That is, the reading and spelling skills of

the hearing-impaired children studied lagged behind those of their age-matched normal-hearing

peers and were significantly better than those of their RD peers.

Specifically, the HI group's reading and spelling skills were significantly lower than

normally developing group on all word-level and passage-level measures with the exception of

the nonword reading subtest (Decoding Efficiency on the TOWRE). When compared to the NI

controls, both the HI and NH subjects had significantly better scores that their RD peers on all

tested measures. The RD group overall showed marked deficits in both phonological awareness









and rapid naming (i.e., double-deficits) while the HI group showed deficits in phonological

awareness (i.e., phonological deficit only). That is, the data clearly showed that two core deficits

(i.e., phonological deficit and rapid naming deficit) were selectively depressed in the HI and RD

groups' literacy skills. These results lend support to the double-deficit hypothesis, which

proposes that reading disability can be caused either by phonological processing deficits, by

RAN deficits, or (in the most severe cases) by a combination of both deficits (Wolf and Bower,

1999).

However, these findings were moderately different from the previous studies (Briscoe et al.,

2001, Gibbs, 2004, Halliday & Bishop, 2006). Halliday & Bishop (2006) used two timed word

reading tests from the TOWRE and they reported no significant difference between the HI and

NH groups on both tests. Briscoe et al. (2001) used three reading tests, including two untimed

word and nonword reading tests and one passage comprehension test and reported that their HI

children's reading was not different from the age-matched controls and was slightly better than

that of an SLI group in spite of their depress phonological processing skills. Based on this

observation, the authors concluded that hard-of-hearing children's reading development would

not necessarily require the support from phonological processes.

If deficient PA skills in HI kids do not impact their reading, this finding would pose a

serious challenge to the universality of phonological core deficit theory. To date, many

experimental studies with special populations have provided findings that support the case for

the universal necessary role of phonological processing capacity in reading acquisition. For

example, a series of studies with deaf readers provides strong evidence for the necessity of

phonological coding in memory for skilled reading. Hanson, Goodell, & Perfetti (1991)

documented that congenitally deaf individuals who had become fluent readers had somehow









acquired a phonological coding strategy despite never having heard speech. Fowler, Doherty, &

Boynton (1995) also documented the contribution of phonological memory to reading

acquisition in young adults with Down syndrome. These subjects, performance in phonological

memory accounted for a substantial portion of the within-group variability in reading

achievement.

Furthermore, neuroimaging studies support a universal phonologically-based

neurocognitive differences in all dyslexics. Researchers posited that if dyslexia is caused by

phonological deficit, this deficit should be observed in neurobiological measures of the brain of

dyslexics across cultures and language. Recently, Paulesu, Demonet, Fazio, McCrory, Chanoine,

Brunswick, Cappa, Cossu, Habib, Frith, and Frith (2008) analyzed three dyslexic groups who

spoke either Italian, French, or English on tasks of single word reading and phonological short-

term memory. Positron emission tomography (PET) scans during reading showed the same

reduced activity in a region of the left hemisphere in all dyslexics groups, with the maximum

peak in the middle temporal gyrus and additional peaks in the inferior and superior temporal gyri

and middle occipital gyrus. They concluded that dyslexia is a result of a universal phonological

processing deficit that is evident across different orthographies.

In a similar vein, this study showed that the HI group's partly impaired phonological

processing skill has a strong association with literacy, confirming its contributive and necessary

role in developing skilled reading. Thus, as suggested by previous studies, the central role of

phonological processing skills should not be diminished in the population of children with mild

to moderate SNHL. In the following section, the results from the correlational analyses and

hierarchical regression are discussed.









Correlation and Regression Findings

No previous studies focused on the relationship between reading and phonological

processing ability in children with mild to moderate SNHL, when the effect of hearing loss was

statistically partialed out. Four key findings emerged from the correlational and regression data.

First, as similar before, the correlations between the HI children's phonological awareness

(Elision) and all reading and spelling measures was remarkably consistent once covariations with

age, nonverbal IQ, and degree of hearing loss were controlled (partial coefficients' ranged

from .582 to .751). These correlations occurred regardless of the test conditions (i.e., timed,

untimed) and level of tasks (i.e., word/passage) (Bowers, 1993; Bradley & Bryant, 1985;

Cornwall, 1992; Cronin & Carver, 1998; Muter, Hulme, Snowling,& Taylor, 1997).

This strong relationship between phonological processing skill and reading skill was

supported with hierarchical regression analyses. Regression analyses on scores of untimed tasks

indicated that the Elision subtest accounted for around 33% of the variance of word reading and

50% of variance of nonword reading measure. More than 60% of variance was reduced for each

skill by including the Elision subtest and background control variables (age, nonverbal, and

better ear's PTA). Similarly, on timed word-level reading tasks, an average of 33% of the

variance were associated with the Elision subtest. The Elision test also significantly accounted

for the variance in both passage reading rate (53%) and accuracy (43%).

Interestingly, consistent with the findings reported in the CTOPP manual (Wagner et al.,

1999), the Blending subtest did not correlated with any literacy measures. Findings from this

study support that although Elision and Blending tests are phonological awareness measures,

they tap into different cognitive abilities. Blending may be less challenging than elision because

it requires only one operation, the synthesis of separate sounds into a word while holding these

sounds in STM. On the other hand, the task of elision requires segmenting sound in words,









omitting the target sound 'g', and creating a new word while holding all sounds in STM. Hence,

elision appears to be a more rigorous test of phonological awareness. In their study of children

with dyslexia, Katzir, Kim, Wolf, O'Brien, Kennedy, Lovett, & Morris (2006) found that the

dyslexic group's blending ability was correlated with only one reading test (Word Attack subtest

on the WRMT-R) out of five phonological awareness and reading measures. In contrast, elision

was significantly correlated with all five measures. These cumulative findings imply that not all

phonological tasks are created equal and that greater precision in our conceptualization of

phonological tasks is required.

Secondly, the relationships between rapid naming skill and reading were examined. Data

of the HI group in this study underscored the independent characteristic of the RAN task. That is,

it was remarkable that the RAN measures did not show any significant correlations with untimed

tasks of reading at both the word or passage level, but large correlations were found with timed

reading tests at both word and passage levels. These data were supported in the regression

analyses.

These cumulative results support previous findings suggesting that the rapid letter naming

task assesses different underlying constructs than those assessed by phonological measures

(Misra, Katzir, Wolf, & Poldrack, 2004; Wolf& Bowers, 1999; Wolf et al., 2002). Thus, in a

task that primarily places no demands on speeded reading, phonologically-based skills are

important. For a task requiring fast lexical access (timed measures), the rapid naming skills play

more of a role. In other words, RAN primarily affects performance on reading tasks that require

speeded fluent response, and phonological awareness and short-term memory primarily affect

performance on reading tasks that emphasize phonological processing skill.









In a study of 476 children with reading disorder, Compton, DeFries, & Olson (2001)

similarly reported that the double-deficit group most resembled the rate-deficit group on

measures that require fluent/speeded word reading skill and reading comprehension, whereas the

double-deficit group tended to perform similarly to the phonological-deficit group on measures

emphasizing phonological processing skills. Katzir et al. (2006) also reported that phonological

awareness (elision and blending) contributed more variance to phonologically-based reading

measure and RAN contributed more to high frequency word reading and word-reading efficiency,

that are more related to speed of processing. In support of this observation, no significant

correlations were found between the two alphanumeric rapid naming tasks and all other

phonologically-based measures in the HI group (PA and STM) (See Table 4-17).

For further analyses, the investigator tried to evaluate explicitly the additive nature of the

RAN task's prediction of timed reading measures. A final set of regression analyses was

conducted, where phonological awareness and short-term memory variables were entered in Step

2 and each of the two RAN measures (letter and digit) was separately entered in Step 3 showed

that RAN contributed independent variance to timed reading measures even after controlling for

children's background variables, phonological awareness, and phonological short-term memory

variables. This additive prediction by the RAN seems to confirm the findings from Wolf and

Bowers (1999), who viewed rapid naming as an extraphonological construct and view rapid

naming deficits as an additive source of reading difficulty over phonological awareness.

As yet, it is not clear which cognitive processes underlie RAN and account for its

relationship with reading. Even though Wolf and Bower's theoretical framework can neatly

explain the heterogeneity of clinical data, complex interrelationships between RAN,

phonological awareness, and reading itself make causal inferences very difficult.









In their large-sized study which assessed 1,010 children's reading performance, Powell,

Stainthorp, Stuart, Garwood, & Quinlan (2007) also reported that RAN deficits occurred in the

absence of phonological awareness deficits. In their structural equation modeling, solutions

where RAN was subsumed within a phonological processing factor did not provide a good fit to

the data, suggesting that processes outside phonology may drive RAN performance and its

association with reading. Convincingly, children with single RAN deficits performed more

slowly in speed of processing than did closely matched controls with normal RAN skill.

Kail, Hall, & Caskey (1999) also proposed that speed of processing underlies the RAN

skills. The reason for the lack of clarity about the causal nature of the link between RAN and

reading is that most of previous studies have been correlational. Using this approach, they could

not rule out the possibility that, rather than being causal in nature, the relationship between RAN

and reading may be driven by a third unknown factor. This view was proposed by Kail and

colleagues (Kail, 1991; Kail & Hall, 1994; Kail et al., 1999), who argued that, rather than being

constricted to the reading system or even to linguistic processes, RAN performance reflects more

generalized processing speed. According to this approach, early childhood is characterized by a

general and gradual increase in global processing speed. Thus, the relationship between RAN

and reading should be understood by the fact that both are influenced by the same underlying

factor, namely, processing speed. The present study's results from the correlation and regression

analyses which revealed important role of rapid naming in timed reading seem to fit into this

approach.

Thirdly, relationships between reading and oral language skills were investigated. As

expected, only grammatical knowledge and vocabulary skills were correlated with passage

comprehension, suggesting that linguistic capacity dealing with vocabulary or grammatical









knowledge does not significantly contribute to word-level reading tasks. This is likely to be the

case because it appears that word-level reading does not necessarily require syntactic or semantic

computation. That is, at the word level, poor phonological difficulties are likely to hinder

children's word recognition ability, but at the sentence level, children may have more difficulties

integrating the information and extract meaning due to grammatical and syntactical difficulties

and/or their restricted semantic knowledge (Hartas, 2005; Perfetti, 1985; Perfetti, 1992).

Lastly, the predictive role of auditory processing skills for reading ability was investigated.

Most of auditory processing tasks scores failed to show significant correlations with reading and

spelling measure and did not predict reading in the HI group. Specifically, no SCAN subtests

were correlated with reading measures. Only the Dichotic Digit test showed significant

correlations with two nonword reading tests. Based on the lack of association, it appears that

auditory tests of this nature provide little predictive value for assessing reading skills, at least in

children who have normal levels of intellectual abilities.

Summary

Contemporary research has revealed a great deal about the factors that interfere with the

process of learning to read (Snow, Burns, Griffin, 1998). Perhaps the greatest contribution over

the past two decades is accumulating evidence of a phonological processing deficit as the core

problem leading to poor reading (Stanovich & Siegel, 1994). Marschark (1997) explained hard-

of-hearing children's reading difficulty as resulting from their impaired phonological channel.

Such findings support the theory about the role of impaired auditory perception in reading

disability and provided the impetus for the comparison between two populations with impaired

phonological processing skills and depressed reading ability: (1) dyslexics with a good peripheral

hearing system, but whose phonological processing is impaired, and (2) a non-dyslexic hearing-









impaired population, whose peripheral hearing system is impaired. This dilemma was the basis

of this present research.

This study was designed with the premise that that literacy development requires the same

acquisition of phonological processing skills, whether a child's hearing is impaired or normal.

First, the principle aim of the study was to examine the effect of mild to moderate SNHL on

children's performances on a range of spoken language, phonological processing skills, and

literacy (reading and spelling). Performances of two controls groups of children (NH and RD),

matched for age, grade, and non-verbal intelligence were used to investigate the HI group's

weaknesses and strengths in phonological processing ability and its relationship to reading skills.

Consistent with previous findings, the HI group's receptive and expressive vocabulary and

morphosyntactic skills were significantly delayed when compared to those of the NH group.

Similarly, the HI group's performance on the phonological processing tasks was significantly

lower than their normal controls. However, their phonological processing skills were selectively

depressed such that only phonologically-based tasks (i.e., phonological awareness and

phonological short-term memory) were seen significantly lower than the normal controls and the

RAN skills, a measure of general speed of processing or efficiency of lexical access, was seen

intact.

In line with Katzir and colleagues' (2006) explanation, this finding suggests that auditory

perceptual distortion due to SNHL only affects phonologically-based processing abilities and not

extraphonological processing abilities (RAN).

Secondly, turning to performance on the literacy tasks, this study sustains the view that

reading is the 'most difficult academic challenge for the hearing impaired' (Marschark & Harris,

1996). It was found that the HI group's literacy skills were significantly lower than their









normally hearing controls in most of reading and spelling measures. Therefore, it was clearly

noted that phonological processing deficit related to early perceptual inaccuracy is very likely to

affect later literacy and related tasks of children with mild to moderate SNHL. This critical role

of phonological processing component was also confirmed by the correlational and regression

analyses which showed that phonological processing skills are important correlates and

predictors of hearing-impaired children's reading and spelling as well. This finding contradicts

previous studies which argued that reading skills of hearing-impaired would not necessarily

require the support from phonological processing ability. As noted previously, RAN measures

were not associated with any of untimed reading tasks, but had significant correlations only with

timed reading tasks. This extraphonological property of the RAN was clearly demonstrated in

the regression analyses, where RAN measures additively contributed independent variance to

timed reading measures after phonologically-based variables were controlled for.

One final issue for comment is the impact of degree of hearing loss on language,

phonological processing and literacy. Many previous studies of reading in children with mild to

moderate SNHL have not reported a significant effect of hearing loss on reading (Blair et al.,

1985; Davis et al., 1995; Elfenbein et al., 1986; Gilbertson et al., 1995). In a similar vein, degree

of hearing loss did not show any significant correlations with reading and spelling measures after

age and non-verbal IQ were controlled for. One possible explanation for this insignificance is

that, with the advent of better hearing aid technology, the impact of degree of hearing loss

perceptual might have become latent in predicting reading achievement, only affecting

phonological processing component. It is possible that the latent relationship between hearing

loss and reading could be mediated by other confounding factors such as age at identification









and/or initial hearing aid fitting, clinical efficacy of hearing aids, educational environment, and

so on.

Clinical Implications

First, consistent with prior research, the current study revealed that hearing-impaired

children's depressed phonological processing abilities are significantly related to reading and

spelling skills. The clinical implications are evident. Language, phonological processing skills,

and reading ability should be thoroughly examined even in children with mild to moderate HL.

Secondly, in clinical settings, a common view is that some children with mild or moderate

impaired hearing loss who are trained orally do show systematic phonological deficits similar to

those of normal hearing children with phonological disorders (Abraham, 1993). It is generally

accepted that deaf and hard-of-hearing children learn to read following the same sequence of

skill development that hearing children do (Chall, 1996; Hanson, 1989; King & Quigley, 1985;

Paul, 1998, 2001). Therefore, it can be assumed that hard-of-hearing individuals, like hearing

individuals, will also benefit from the development of phonological processing skills as part of

their beginning reading instruction. Nielsen and Luetke-Stahlman (2002) maintained that if

hearing-impaired children are to read the sound-based printed words and phrases of English, they

must develop and improve phonological sensitivity (awareness). Thus, previous results which

suggested that phonological ability is not necessarily required in children with mild to moderate

HL might lead many educators to assume that HI children's weak phonological awareness would

not be a barrier to successful reading development and ignore the importance of phonological

awareness training.

There is research evidence to suggest that the ability to use phonological information while

reading is a distinguishing variable when comparing accomplished deaf and hard-of-hearing

readers to average deaf and hard-of-hearing readers (Conrad, 1964; Engle, Cantor, & Turner,









1989; Hanson, 1982; Hanson & Fowler, 1987; Hanson & Lichtenstein, 1990; Hanson, Goodell,

& Perfetti, 1991; Leybaert, 1993; Musselman, 2000; Perfetti & Sandak, 2000). These findings

imply that effective programs and strategies for teaching hearing-impaired children these skills

may be the key to obtaining higher levels of reading achievement for this population.

Unfortunately, survey studies of instructional methods being used by teachers of hard-of-hearing

children reported that the overwhelming majority of teachers do not incorporate direct teaching

of phonological awareness in their reading instruction (Coley & Bockmiller, 1980; Hasenstab &

McKenzie, 1981; LaSasso, 1978, 1987; LaSasso & Mobley, 1997). Thus, currently, it is not

surprising that hearing impaired students are at risk for acquiring sufficient decoding skills due to

their limited access to and instruction focused on the phonological aspects of the English

language (e.g., phonemic awareness and phonics). In this vein, the findings of this research

would have implications on the direction of intervention planning and the strategies of teaching

reading to members of the studied populations. Furthermore, speech pathologists who work with

hard-of-children need to be aware of clinical methods of explicit phonological instructions

beyond traditional intervention techniques used for children with articulation or phonological

disorder.

Limitations

While this study was the first of its kind to provide a comprehensive profile of the

phonological and literacy profiles HI children's with mild to moderate SN hearing loss, a few

limitations in the study's methodology must be acknowledged.

First, the relatively small sample size of the HI group (N=19) makes it difficult to draw

conclusions regarding the effect of degree of hearing loss beyond this study. In most

psychometric studies of hearing impaired children, it is well assumed that there will be a problem

associated with participation. Children with mild hearing impairments are less likely to









participate in extensive testing than children with more extensive hearing losses because their

problems are not perceived as being as serious and parents of these children may not feel the

immediate need to see results of their children's performance in the areas of language and

literacy. Secondly, the administration of more than ten psychometric tests combined with hearing

tests might have been too demanding for some of the children, reducing the validity of tests.

Lastly, in a recent comprehension review of reading research for students who are deaf or hard of

hearing, Schirmer and McGough (2005) indicated that intervention research evaluating the

effectiveness of phonological instruction for students is extremely limited. Unfortunately, no

clinical studies exist regarding the efficacy of direct phonological instruction for children with

moderate levels of HL. This limited number of studies certainly warrants attention. As a cross-

sectional study, this research could not address the developmental trajectory of reading skills in

children with mild to moderate SNHL. Future longitudinal studies are needed to unravel the

long-term effects of hearing loss on academic achievement.









APPENDIX A
INFORMED CONSENT LETTER FOR PARENTS AND GUARDIANS

Informed Consent Letter for Parents or Guardians

Jungjun Park
Doctoral Student, Dept. of
Communication Sciences and Disorders,
Univ. of Florida, Gainesville,
FL 32611-7420
Tel.: (352) 392-2113 ext. 293
E-mail: pajj_gsc@hotmail.com


Jan, 2007

Dear Parent/Guardian,
I am a doctoral student of Communication Sciences and Disorders at the University of
Florida, conducting research under the supervision of Dr. Linda Lombardino. We are interested
in examining the effect of hearing loss on oral language, phonological processing, and reading
skills of 7- to 12-year-old children. In spite of increasing number of children with hearing loss,
little is known about their reading performance. Especially, very few research efforts have been
directed to literacy development of children with less-than-profound hearing loss, so studies
strictly focusing on the reading ability in this population are needed.
The main purpose of this study is to assess the oral language, phonology, and reading
skills of children with or without sensorineural hearing loss and compare them. For the
study, we need; 1) 30 children with normal hearing/reading ability, 2) 30 children with mild or
moderate sensorineural hearing loss. Children with conductive hearing loss would NOT be
able to participate in the study and only children whose primary communication mode is spoken
language are qualified. Students will be mostly engaged in linguistic activities that have them
listen to the sounds, read and write words and sentences, or give appropriate answers to the
questions of the researcher. Assessment is expected to take 90 to 100 minutes and will be done
using a set of standardized oral language and reading tests. Only when we could not finish the
assessment, your child will be asked to work with us once more on the other day. Either a trained
student or me will be present during all sessions to help your child to do the tests.

There is no expected risk of physical, psychological, or economic harm to your child.
This assessment project will directly benefit your child's literacy skills regardless of their
hearing level. That is, through the whole assessment procedures, your child's profile of oral
language and reading development will be investigated. Based on this, you and the classroom
teachers are also expected to benefit from this study by being informed of the child's current
level of oral language and literacy performance, which can be used to decide the steps needed to
improve reading and oral language instruction. A copy of the results of the assessment will be
available upon your request.
With your permission, we will audio record your child's responses during assessment and
obtain your child's hearing loss level from school files if your child has a hearing loss. Tapes









will be used only to validate our assessment results and the contents will not be transcribed. It
will be available only to the assistant student, my supervisor, and me.

I also have attached a questionnaire asking for some information regarding the
developmental characteristics of your child and family background. Including the
information from this questionnaire, all the data will be numerically coded and will not be
marked with your child's name. All individual records including the audiotapes will be destroyed
once the study has ended. Your child's identity will strictly be kept confidential to the extent
provided by law. No real names, initials, or other identifying information will be used during
spoken or written presentation of study results. Participation or non-participation in this study
will not affect your child's grades or placement in programs at all.

As mentioned, children with normal hearing ability are also needed for this study. If
your child would participate in the study as a subject with normal hearing regardless of his/her
reading proficiency, his/her hearing will be screened at the University of Florida Speech Hearing
Clinic (UFSHC) by a graduate clinician of Doctoral of Audiology program of the University of
Florida. Hearing will be screened based on puretone audiometry and tympanometry and there is
no cost for this evaluation. For this audiologic assessment, I will make an appointment with
you beforehand. You and your child have the right to withdraw consent for your child's
participation at any time without consequence. No compensation will be offered. If you have any
questions about this research protocol, please contact me, Jungjun Park at (352) 392-2113, ext
293 or (352) 328-7671(pajjgsc@gmail.com) or Dr. Lombardino at (352) 392-2113, ext 285.
Questions or concerns about your child's rights as a research participant may be directed to the
UFIRB Office, University of Florida, Box 112250 Gainesville, FL., 32611, (352) 392-0433. If
you should decide to grant permission, please complete the attached questionnaire form and
return it with this consent letter to the classroom teacher. You can also keep a copy of this
informed consent letter for your information. Thank you for the consideration.

Sincerely,
Park, Jungjun










Please sign below and return this letter to the
classroom teacher

I have read the procedure described above and I have received a copy of this form.
I understand that my decision is entirely voluntary and agree to allow my child
to participate in the study to investigate oral and written language
skills.


I also authorize the principal investigator, Jungjun Park, to secure and use information in
the questionnaire form for his research.



Signatures:

Parent/Guardian: Date:

Principal Investigator: Date:

Supervisor/Investigator: Date:









Child Assent Script


The following is a script that will be used prior to each session to ensure that the student knows
of his/her involvement and that he/she may choose not to participate if he/she does not want to.

Investigator: We are going to do some activities with sounds and words. During these
activities you will be looking at listening to, or reading different sounds and words. If you
don't want to do this, that is okay. Do you want to do the activities with me?

If the student indicates yes, the investigator will begin the assessment session.

If the student indicates no, the investigator will say:
That's okay. Maybe I will come back later to see if you want to do these activities.

When the investigator returns, the same script will be used.




Date:





Signature:








APPENDIX B
QUESTIONNAIRE FORM


I IDENTIFYING INFORMATION I


1. Child's Legal Name:
2. Date of Birth:
3. Address:


Age: ____ years and


Street
4. Mailing Address (if different):


5. Person Completing This Form:


Relationship


6. Phone (Home):( )
(Work): ( )
(Cell): ( _____)
7. E-mail address (if available)


months


City/State


Zip Code


Name









I EDUCATIONAL HISTORY & GENERAL HEALTH I


1. EDUCATION
Name of Current School:
Grade:
Teacher's Name:
Describe the child's progress in school:


Does your child have any problems in reading books or classroom materials? If yes, please
describe in your own words your child's difficulty. Give examples, if possible.


2. SPEECH/LANGUAGE/READING
Does your child have any speech and/or language problems?
If yes, please describe in more detail.


w Yes w No


Does your child have any reading problem? w Yes w No
Please describe in your own words your child's difficulty. Give examples, if possible.





Because of this, has your child received speech-language therapy or any remedial reading
program at clinic or school? wYes w No


If yes, please indicate the following:
When/where:
Dates:









w Length and frequency of sessions:
w Primary focus of therapy:


How do you feel about your child's reading competence?
w Excellent w Good w Fair w Slightly Poor IiVery Poor


3. GENERAL HEALTH
Do you feel your child has normal vision? w Yes w No
If not, has the child's vision been examined? w Yes w No
Describe any illness, accidents, injuries, operations, and/or hospitalizations and include the
age of the child:










FAMILY BACKGROUND I


1. Mother's Name: Age:


2. Father's Name: Age:


3. Guardian's Name: Age:


4. Education level of the primary income earner in the home (Please mark in the box.)


Completed elementary school
Completed junior high school
Received general education degree
Completed high school
Completed 1 or more years of technical/vocational school
Completed technical/vocational school
Completed 1 or more years of university/college
Bachelor's degree
Completed 1 or more years of graduate school
Master's degree
Course work completed for Ph.D., but no dissertation; law degree without bar;
medical degree without internship completed
Ph.D.; law degree with bar; medical degree with internship completed


4. Annual Income: $
(This information is especially very important for successful comparison of the
children's performance and will be kept completely confidential in any circumstances.)










I AUDIOLOGIC HISTORY (For Hearing impaired Child) I

Please check your child's hearing status and use an appropriate section for this
audiologic history information.
O If your child has a hearing loss, please answer only to the following questions.
o For children with normal hearing, different questions are provided at page 6 of this
questionnaire form.


Hearing loss
1. When and how did you first notice your child's hearing problem?



2. When did you first consult an Ear, Nose, and Throat physician or an audiologist
since your initial notice?


3. If you remember, please indicate the clinic name.


4. What was the initial treatment after the physician or the audiologist identified your
child's hearing loss?



5. Do you know the results) of initial audiologic diagnosis? This information may be
available from a copy of initial audiologic evaluation. If you know any one of the
followings, please indicate based on exact information. If not sure, just leave it
blank.
o Cause(s) of hearing loss (e.g., congenital, genetic, developmental)


" Degree of hearing loss ( in dB units; if known) :
Right ear (dB HL) Left ear (dB HL)
" Type of hearing loss (if known) :
w sensorineural w conductive w mixed









6. Does your child currently wear hearing aid(s)?
If yes, please describe:
Make Model
Both ears Right ear only Left ear only
Does the child wear it all time ? w Yes w No
If not, when does your child not wear it?


7. When did your child first start to wear hearing aid(s)?
Based on this information, for how many years has your child been wearing a
hearing aid? years and _months
8. Is there a history of hearing loss in the family? w Yes w No w Not sure
If yes, which family members) have a hearing loss?
What caused their hearing loss?




* Speech and Language
1. Do you have any trouble understanding your child's speech? w Yes w No
2. If yes, briefly provide the difficulties.




3. Does your child use both oral and sign language at home? w Yes w No
4. Which one of the followings is your child's major communication mode at home?
w oral language w Sign language
5. Does any parent of your child use sign language fluently? w Yes w No
6. Which language mode does your child's school use mostly?
w oral language w Sign language










I AUDIOLOGIC HISTORY (normal hearing) I


A. History of Ear Infections
1. My child has suffered from ear infections. w Yes w No
2. My child had more than 3 ear infections between birth and 12 months of age.
w Yes w No
3. My child has had at least one ear infection which lasted more than 3 months.
w Yes w No
4. What was the treatment for the ear infections at that time?


B. Hearing
1. My child has never failed hearing screenings) in school. w Yes w No
2. My child has at least once failed a hearing screening in school. w Yes w No
Date of screening (if available):
3. Has your child ever been referred to professionals for hearing evaluation?
w Yes w No
4. If yes, what was the result?










APPENDIX C
SCORING SHEETS


Date:


I/ / 2007 Subject #


Language and Literacy


Name DOB & age(year/month) / /19 ( years / months)
Gender School
Grade Testing Location
Tester Others
Hearing level Normal mild moderate severe RD
PTA Left: Ri2ht: Tympanometry

Oral language
area Raw Standard %-ile Age-equi(month) Grade-equi
PPVT-III ( )
EVT ( )
CASL ( )
Phonology
Grade Sum Composite Comp
area raw Stand %-ile Age equi equi standard standard %-ile

Elision ( )
Blending- ( )
Rapid Digit Naming ( )
Rapid Letter Naming ( )
Memory for Digit ( )
Nonword Repetition ( )
Rapid Object Naming ( )
Digit Ordering Raw:
WM
Digit Backward Raw:
WMRaw:












Literacy

Age-Norm .Grade-Norm
area raw Age-Norm AE GE Grade-Norm
stan % stan %
Word ID (WRMT) ( )
Word attck(WRMT) ( )
Pass Comp(WRMT) ( )

Spelling (WRAT)

Age-Norm Grade-Norm
raw stan % Su R-rofic sta % um R-profici AE GE
stan % stan %
Word effi(TOWRE)
word deco(TOWRE)

raw stand % Age-equi Grade-equi
Passage Reading
Rate(GORT-4)
Passage Reading
Accuracy(GORT-4)
GORT fluency

TONI ( )









Audiologic Tests


Name I DOB & age(year/month) / /19 ( years / months)
Gender Testing Date

Puretone Audiometry (Unaided)
250 500 1 K 2K PTA 3K 4K 6K 8K SRT WRS(%)
Hz Hz
Right
Left
*SRT: Threshold level for speech (word) recognition (dB HL)
*WRS: Word recognition score based the MCL (most comfortable hearing level) % score

Type of Hearing loss:
Normal : E Only Right Ear E Only Left Ear a Both Ear Normal
If not: Left Ear : Sensorineural hearing loss
Right Ear : Sensorineural hearing loss

Tympanometry (Middle ear status):
- Normal
If not, Right Ear: Left Ear:

SCAN-C/A TEST (Unaided; Using 50 dB HL for normal group and the MCL for HI group)

Raw Score
Subtests Riaht Left Sum(L/R) Standard %-ile
Right Left
Filtered Words /20 /20 /40
Figure-ground /40 /40 /80
Competing Words /30 /30 /60
(Dichotic words)
Sum of Standard scores
Mean of 3 Standard scores

Dichotic Digits (Two Digits only) :
(Unaided using 50 dB above the PTA for the normal group and the MCL for HI group)


Raw Score
Left Right Sum (L/R)

Double Digits /40 /40 /80
%









LIST OF REFERENCES


Abraham, S. (1993). Differential Treatment of Phonological Disability in Children With
Impaired Hearing Who Were Trained Orally. American Journal of Speech-
Language Pathology, 2, 23-30.

Abu-Rabia, S. (2002). Phonemic awareness and middle-ear disease among Bedouin
Arabs in Israel. Reading Psychology, 23, 289-296(8)

Adams, M. (1990). Beginning to read. Thinking and learning about print. Cambridge,
MA: MIT Press.

Allen, T. E. (1986). Patterns of academic achievement among hearing impaired students:
1974 and 1983. In A. N. Schildroth & M. A. Karchmer (Eds.), Deaf Children in
America (pp. 161-206). San Diego, CA: College Hill Press.

Allen, T. E. (1994). What are the deaf and hard-of-hearing students leaving high school
and entering postsecondary education? Unblished manuscript. Washington, DC:
Gaullaudet University.

Allen, T. E., & Schoem, S. R. (1997). Educating deaf and hard-of-hearing youth: What
works best? Paper presented at the Combined Otolaryngological Spring Meetings
of the American Academy of Otolaryngology. Scottsdale, AZ.

Anthony, J. L. & Francis, D. (2005). Development of phonological awareness. Current
Directions in Psychological Science, 14, 255-259.

Archer, A. L., Gleason, M. M., & Vachon, V. L. (2003). Decoding and fluency:
Foundation skills for struggling older readers. Learning Disability Quarterly, 26,
89-101.

Baddeley, A. D. & Hitch, G. J., (1974). Working memory. In Bower, G.H. (ed.), Recent
Advances in Learning and Motivation (Volume 8), New York: Academic Press.

Baddeley, A. D. (1986). Working Memory. London: Oxford University Press.

Baddeley, A. D., Gathercole, S. E., & Papagno, C. (1998). The phonological loop as a
language learning device. Psychological Review, 105, 158-173.

Baddeley, A.D. (2003) Double dissociation: Not magic but still useful. Cortex, 39, 129-
131.

Beitchman, J. H. & Young, A. R. (1997). Learning disorders with a special emphasis on
reading disorders: A review of the past 10 years. Journal of the American
Academy of Child and Adolescent Psychiatry, 36, 1020 1032.









Bess, F. H., Dodd-Murphy, J., & Parker, R. A. (1998). Children with minimal
sensorineural hearing loss: Prevalence, educational performance, and functional
status. Ear andHearing. 19, 339-354.

Biddle, K. R. (1996). Timing deficits in impaired readers: An investigation of visual
naming speed and verbal fluency. Unpublished doctoral dissertation, Tufts
University, Boston, MA.

Blachman, B. (1989). Phonological awareness and word recognition: Assessment and
intervention. In A. Kamhi & H. Catts (Eds.), Reading disabilities: A
developmental language perspective (pp. 133-158). Boston, MA: College-Hill
Press.

Blachman, B. A. (1991). Early intervention for children's reading problems: Clinical
applications of the research in phonological awareness. Topics in Language
Disorders, 12, 51-65.

Blair, J., Peterson, M., & Viehweg, S. (1985). The effects of mild hearing loss on
academic performance of young school-age children. Volta Review, 87, 87-93.

Blair, J., Peterson, M., Viehweg, S. (1985). The effects of mild sensorineural hearing loss
on academic performance of young school-age children. Volta Review, 87, 87-93.

Blarney, P. J., Sarant, J., Paatsch, L. E., Barry, J. G., Bow, C. P., Wales, R. J. (2001).
Relationships among speech perception, production, language, hearing loss and
age in children with impaired hearing. Journal of Speech, Language and Hearing
Research, 44, 264-285.

Bowers, P. G. & Wolf, M. (1993). Theoretical links among naming speed, precise timing
mechanism, and orthographic skill in dyslexia. Reading and Writing: An
Interdisciplinary Journal, 5, 69-85.

Bowers, P. G. (1989). Naming speed and phonological awareness: Independent
contributors to reading disabilities. In S. McCormick & J. Zutell (Eds.), Cognitive
and social perspectives for literacy research and instruction: 38th Yearbook of
the National Reading Conference (pp. 165-173). Chicago: National Reading
Conference.

Bowers, P. G. (1993). Text reading and rereading: Determinants of fluency beyond word
recognition. Journal of Reading Behavior, 25, 133-53.

Bowers, P. G. (1995). Re-examining selected reading research from the viewpoint of the
"double-deficit hypothesis." Paper presented at the Annual Meeting of the Society
for Research in Child Development, Indianapolis, IN.

Bowers, P. G., & Swanson, L. B. (1991). Naming speed deficits in reading disability:
Multiple measures of a singular process. Journal of Experimental Child
Psychology, 51, 195-219.









Bowers, P. G., Golden, J. 0., Kennedy, A., & Young, A. (1994). Limits upon
orthographic knowledge due to processes indexed by naming speed. In V. W.
Berninger (Ed.), The varieties of orthographic knowledge: Theoretical and
developmental issues (pp. 173-218). Dordrecht: Kluwer Academic Publishers.

Bowers, P. G., Steffy, R., & Tate, E. (1988). Comparison of the effects of IQ control
methods on memory and naming speed predictors of reading disabilities. Reading
Research Querterly, 23, 304-309.

Bowey, J. A., Cain, M. T., & Ryan, S. M. (1992). A reading-level design study of
phonological skills underlying fourth-grade children's word reading difficulties.
Child Development, 63, 999-1011.

Bradley, L. & Bryant, P.E. (1983). Categorizing sounds and learning to read: A Causal
connection. Nature, 30, 419-421.

Bradley, L., and Bryant, P.E. (1985). Rhyme andReason in Reading and Spelling. Ann
Arbor: University of Michigan Press.

Brady, S. (1991). The role of working memory in reading disability. In Phonological
Processes in Literacy: A Tribute to Isabelle Y. Liberman, Brady, S., Shankweiler,
D. (Eds.), Hillsdale, NJ: Lawrence Erlbaum Associates.

Brady, S. (1997). Ability to encode phonological representations: An underlying
difficulty of poor readers. In Foundations of Reading Acquisition and Dyslexia:
Implications for Early Intervention, Blachman, B., et al. (Eds), Mahwah, NJ:
Lawrence Erlbaum Associates.

Brady, S., Poggie, E., and Rapala, M. M. (1989). Speech repetition abilities in children
who differ in reading skill. Language and Speech, 32, 109-22.

Brady, S., Shankweiler, D., and Mann, V. A. (1983). Speech perception and memory
coding in relation to reading ability. Journal of Experimental Child Psychology,
35, 345-367.

Brandes, P. J., & Ehinger, D. M. (1981). The effects of early middle ear pathology on
auditory perception and academic achievement. Journal of Speech and Hearing
Disorder, 46, 301-307.

Briscoe, J., Bishop, D. V. M., & Norbury, C. F. (2001). Phonological processing,
language, and literacy: A comparison of children with mild-to-moderate
sensorineural hearing loss and those with specific language impairment. Journal
of Child Psychology and Psychiatry, 42, 329-340.

Briscoe, J., Gathercole, S. E., & Marlow, N. (2001). Everyday memory and cognitive
ability in children born prematurely. Journal of Child Psychology and Psychiatry,
42, 749-754.









Brown, P. M. & Brewer, L. C. (1996). Cognitive processes of deaf and hearing skilled
and less skilled reader. Journal of Deaf Studies and Deaf Education, 1, 263-270.

Brown,L, Sherbenou, R. J., Johnsen, S. K. (1997). Test of nonverbal intelligence-3
(TONI-3). Pearson Education, Inc.

Bruck, M. (1992). Persistence of dyslexics' phonological awareness deficits.
Developmental Psychology, 28, 874-886.

Bruck, M., & Treiman, R. (1990). Phonological awareness and spelling in normal
children and dyslexics: The case of initial consonant clusters. Journal of
Experimental Child Psychology, 50, 156-178.

Bryant, P. E. (1990). Phonological development and reading. In P. D. Pumfrey, & C. D.
Elliot (Eds.), Children's difficulties in leing, spelling and writing Challenges
and responses (pp. 63-82). London The Falmer Press.

Bryant, P. E., Bradley, L., MacLean, M., & Crossland, J. (1989). Nursery rhymes,
phonological skills and reading. Journal of Child Language, 16, 407-428.

Byrne, B., & Fielding-Barnsley, R. (1989). Phonemic awareness and letter knowledge in
the child's acquisition of the alphabetic principle. Journal of Educational
Psychology, 81, 313-321.

Byrne, B., & Fielding-Barnsley, R. (1990). Acquiring the alphabetic principle: A case for
teaching recognition of phoneme identity. Journal of Educational Psychology, 82,
805-812.

Campbell, R., & Wright, H. (1989). Immediate memory n the orally trained deaf: Effects
of 'lipreadability' in the recall of written syllables. British Journal of Psychology,
80, 299-312.

Carrow-Woolfolk, E. (1999). Comprehensive Assessment of Spoken Language. Circle
Pines, MN: American Guidance Service.

Chall, J. (1996). Learning to Read. The Great Debate (1967). New York: McGraw Hill.

Chall, J. S. (1996). Stages of reading development. Fort Worth, TX: Harcourt Brace.

Chiat, S. (2001). Mapping theories of developmental language impairment: Premises,
predictions and evidence. Language and Cognitive Processes 16, 113-42.

Coley, J., & Bockmiller, P. (1980). Teaching reading to the deaf: An examination of
teacher preparedness and practices. American Annals of the Deaf 125, 909-915.

Compton, D. L., DeFries, J. C, & Olson, R. K. (2001). Are RAN- and phonological
awareness-deficits additive in children with reading disabilities? Dyslexia, 7, 125-
149.









Conrad, R. (1977). The reading ability of deaf school-leavers. British Journal of
Educational Psychology, 47, 138-48.

Cornwall, A. (1992). The relationship of phonological awareness, rapid naming, and
verbal memory to severe reading and spelling disability. Journal of Learning
Disabilities, 25, 532-8.

Cornwall, A. (1992). The relationship of phonological aware-ness, rapid naming, and
verbal memory to severe reading and spelling disability. Journal of Learning
Disabilities, 8, 532-538.

Cronin, V. and Carver, P., 1998. Phonological sensitivity, rapid naming, and beginning
reading. Applied Psycholinguistics 19, 447-461.

Davis, J. M., Effenbein, J., Schum, R., & Bentler, R. A. (1986). Effects of mild and
moderate hearing impairments on language, educational, and psychosocial
behaviours of children, Journal of Speech and Hearing Disorder, 51, 53-62.

Davis, J. M., Shepard, N. T., Stelmachowicz, P. G., & Gorga, M. P. (1981).
Characteristics of hearing-impaired children in the public schools. Part II.
Psychoeducational data. Journal of Speech and Hearing Disorders, 46, 130-137.

Delage, H., & Tuller, L. (2007). Language development and mild-to-moderate hearing
loss: does language normalize with age? Journal of Speech, Language, and
Hearing Research, 50, 1300-1313.

Denckla, M. B. & Cutting, L. E. (1999). History and significance of rapid automatized
naming. Annals of Dyslexia, 49, 29-42.

Denckla, M., & Rudel, R. (1974). Rapid automatizedd" naming of picture objects, colors,
letters, and numbers by normal children. Cortex, 10, 186-202.

Denckla, M., & Rudel, R. (1976). Rapid automatizedd" naming (R.A.N.): Dyslexia
differentiated from other learning disabilities. Neuropsychologia, 14, 471-479.

Dodd, B., & Hermelin, B. (1977). Phonological coding by the prelinguistically deaf.
Perception and Psychophysics, 21, 413-417.

Duara, R., Kushch, A., Gross-Gleen, K. Barker, W. W., Jallad, B., Pascal, S.,
Loewenstein, D. A., Sheldon, J., Rabin, M. Levin, B., Lubs, H. (1991).
Neuroanatomic differences between dyslexic and normal readers on magnetic
resonance imaging scans. Archives of Neurology, 48, 410-416.

Dunn, L. M., & Dunn, L. M. (1997). Peabody Picture Vocabulary Test-Third Edition:
Manual. Circle Pines, MN: American Guidance Services.

Dunn, L. M., & Dunn, L. M. (1997). Peabody Picture Vocabulary Test IIL Circle Pines,
MN: American Guidance Service.









Ehri, L. C. & Wilce, L. S. (1987). Does learning to spell help beginners learn to read
words? Reading Research Quarterly, 12, 47-65.

Ehri, L. C. (1989). The development of spelling knowledge and its role in reading
acquisition and reading disability. Journal of Learning Disabilities, 22, 349-364.

Ehri, L., Nunes, S., Willows, D., Schuster, B., Yaghoub-Zadeh, Z., & Shanahan, T.
(2001). Phonemic awareness instruction helps children learn to read: Evidence
from the National Reading Panel's meta-analysis. Reading Research Quarterly,
36, 250-287.

Eimas, P. (1974). Auditory and linguistic processing of cues for place of articulation by
infants. Perception & Psychophysics, 16, 513-521.

Elbro, C. (1996). Early linguistic ability and reading development: A review and a
hypothesis about underlying differences in distinctness of phonological
representations of lexical items. Reading and Writing: An Interdisciplinary
Journal, 8, 453-85.

Elbro, C. (1998). When reading is "readn" or somthn. Distinctness of phonological
representations of lexical items in normal and disabled readers. Scandinavian
Journal of Psychology. 39, 149-53.

Elbro, C., & Jensen, M. N. (2005). Quality of phonological representations, verbal
learning, and phoneme awareness in dyslexic and normal readers. Scandinavian
Journal of Psychology, 46, 375-384.

Elbro, C., Borstrom, I., & Petersen, D. K. (1998). Predicting dyslexia from kindergarten:
The importance of distinctness of phonological representations of lexical items.
Reading Reseaerch Quarterly, 33, 36-57.

Elbro, C., Nielsen, I., & Petersen, D. K. (1998). Predicting dyslexia from kindergarten:
the importance of distinctness of phonological representations of lexical items.
Annals of Dyslexia, 44, 205-226.

Elfenbein, J. L., Hardin-Jones, M. A., Davis, J. M. (1994). Oral Communication Skills of
Children Who Are Hard of Hearing. Journal of Speech and Hearing Research,
37, 216 226.

Engle, R. W., Cantor, J., & Turner, M. L. (1989). Modality effects: Do they fall on deaf
ears? Quarterly Journal of Experimental Psychology, 41A, 273-292.

Felton, R. H. & Wood, F. (1992). Cognitive deficits in reading deficits in reading
disability and attention deficit disorder. Journal ofLearning Disabilities, 22, 3-13.

Felton, R. H., Wood, F., Brown, I. S., & Campbell, S. K. (1987). Separate verbal memory
and naming deficits in attention deficit disorder and reading disability. Brain and
Language, 31, 171-184.









Fletcher, J. M., Shaywitz, S. E., Shankweiler, D. P., Katz, L., Liberman, I. Y., Stuebing,
K. K., Francis, D. J., Fowler, A., & Shaywitz, B. A. (1994). Cognitive profiles of
reading disability: Comparisons of discrepancy and low achievement definitions.
Journal of Educational Psychology, 85, 1-18.

Foorman, B. R., Francis, D. J., Beeler, T., Winikates, D., & Fletcher, J. M. (1997). Early
interventions for children with reading problems: Study designs and preliminary
findings. Learning Disabilities: A Multi-Disciplinary Journal, 8, 63-71

Fowler, A. E. (1991). How early phonological development might set the stage for
phoneme awareness. In Phonological Processes in Literacy: A Tribute to Isabelle
Y. Liberman, Brady, S., Shankweiler, D. (Eds.), Hillsdale, NJ: Lawrence Erlbaum
Associates.

Fowler, A. E., Doherty, B. J., & Boynton, L. (1995). The basis of reading skill in young
adults with Down syndrome. In L. Nadel & D. Rosenthal (Eds.), Down syndrome
living and learning in the community (pp. 182-196). New York: Wiley.

Foy, J. G. and Mann, V. (2001) Does strength of phonological representations predict
phonological awareness? Applied Psycholinguistics, 22, 301-325.

Foy, J. G., & Mann, V. (2001). Does strength of phonological representations predict
phonological awareness? Applied Psycholinguistics, 22, 301-325.

Friel-Patti, S., & Finitzo, T. (1990). Language learning in a prospective study of otitis
media with effusion in the first two years of life. Journal of Speech Hearing
Research, 33, 188-94.

Gathercole, S. E. (2006). Nonword repetition and word learning: The nature of the
relationship [Keynote]. Applied Psycholinguistics, 27, 513-543.

Gathercole, S. E., & Baddeley, A. D. (1990). Phonological memory deficits in language
disordered children: Is there a causal connection? Journal of Memory and
Language, 29, 336-360.

Gathercole, S. E., Hitch, G. J., Service, E. & Martin, A. J. (1997). Phonological short-
term memory and new word learning in children. Developmental Psychology 33,
966-979.

Gibbs, S. (2004). The skills in reading shown by young children with permanent and
moderate hearing impairment. Educational Research, 46, 17-27.

Gilbertson, M., & Kamhi, A.G. (1995). Novel word learning in children with hearing
impairment. Journal of Speech and Hearing Research, 38, 630-642.

Gillon, G. T. (2004). Phonological Awareness: from research to practice. New York:
Guilford Press.









Goldin-Meadow, S., & Mayberry, R. I. (2001). How do profoundly deaf children learn to
read? Learning Disabilities Research & Practice, 16, 222-229.

Goswami, U. (2000). Phonological and Lexical Processes. In Barr, R., Kamil, M.,
Mosenthal, P., and Pearson, D. (2000) Handbook of Reading Research, 3, 251-
268.

Goswami, U. (2000). Phonological representations, reading development and dyslexia:
Towards a cross-linguistic theoretical framework. Dyslexia, 6, 124-132.

Goswami, U. (2001). Developmental dyslexia. In N. J. Smelser & P. B. Baltes (Eds.),
International encyclopedia of the social & behavioral sciences. (pp. 3918-3921).
Amsterdam: Elsevier Sciences.

Goswami, U. (2003). Why theories about developmental dyslexia require developmental
designs. Trends in Cognitive Sciences, 7, 534-54.

Goswami, U., & Bryant, P (1990). Phonological skills and learning to read. Hove, East
Sussex: Lawrence Erlbaum Associates Ltd.

Gottardo, A., Stanovich, K. & Siegel, L. (1996). The relationships between phonological
sensitivity, syntactic processing and verbal working memory in the reading
performance of third-grade children. Journal of Experimental Child Psychology,
63, 563-582.

Griffiths, Y. M., & Snowling, M. J. (2002). Predictors of exception word and nonword
reading in dyslexic children: the severity hypothesis. Journal of Educational
Psychology, 94, 34-43.

Grigorenko, E. L. (2001). Developmental dyslexia: An update on genes, brains, and
environments. Journal of Child Psychology and Psychiatry, 42, 91-125.

Grushkin, D. A. (1998). Why shouldn't Sam read? Toward a new paradigm for literacy
and the deaf. Journal of Deaf Studies and Deaf Education. 3, 179-204.

Gupta, P. (2003). Examining the relationship between word learning, nonword repetition,
and immediate serial recall in adults. Quarterly Journal of Experimental
Psychology, 56, 1213-1236.

Halliday, L. F. & Bishop, D. V. (2006) Auditory frequency discrimination in children
with dyslexia, Journal of Research in Re, ing. 29, 213-228.

Halliday, L. F., & Bishop, D. V. (2005). Frequency discrimination and literacy skills in
children with mild to moderate sensorineural hearing loss. Journal of Speech,
Language, and Hearing Research, 48, 1187-1203.

Hansen, J., & Bowey, J. (1994). Phonological analysis skills, verbal working memory,
and reading ability in second-grade children. Child Development, 65, 938-50.









Hanson, V. L. (1982). Short-term recall by deaf signers of American Sign Language:
Implications for order recall. Journal of Experimental Psychology: Learniniig.
Memory, and Cognition, 8, 572-583.

Hanson, V. L. (1989). Phonology and reading: Evidence from profoundly deaf readers. In
D. Shankweiler & I. Lieberman (Eds.), Phonology and reading disability: Solving
the reading puzzle (pp. 69-89). Ann Arbor, MI: University of Michigan Press.

Hanson, V. L., & Fowler, C. A. (1987). Phonological coding in word reading: Evidence
from hearing and deaf readers. Memory and Cognition, 15, 199-207.

Hanson, V. L., & Lichtenstein, E. (1990). Short-term memory coding by deaf signers:
The primary language coding hypothesis reconsidered. Cognitive Psychology, 22,
211-224.

Hanson, V. L., & McGarr, N. S. (1989). Rhyme generation by deaf adults. Journal of
Speech and Hearing Research, 32, 2-11.

Hanson, V. L., Goodell, E. W., & Perfetti, C. A. (1991). Tongue twister effects in the
silent reading of hearing and deaf college students. Journal of Memory and
Language, 30, 319-330.

Hansson, K., Forsberg, J., Lofqvist, A., Maki-Torkko, E., & Sahlen, B. (2004). Working
memory and novel word learning in children with hearing impairment and
children with specific language impairment. International Journal of Language
and Communication Disorders, 39, 401-422.

Harris, M., & Beech, J. R. (1998). Implicit phonological awareness and early reading
development in prelingually deaf children. Journal of Deaf Studies and Deaf
Education, 3, 205-216.

Hartas, D. (2005). Language and Communication Difficulties, London: Continuum.

Hasenstab, M. S., & McKenzie, C. D. (1981). A survey of reading programs used with
hearing-impaired students. The Volta Review, 83, 383-388.

Hayiou-Thomas, M. E., Bishop, D. V. M. & Plunkett, K. (2004). Simulating SLI: General
cognitive processing stressors can produce a specific linguistic profile. Journal of
Speech, Language, and Hearing Research 47, 1347-62.

Hemmer, V.H. & Ratner, N.B. (1994). Communicative development in twins with
discordant histories of recurrent otitis media. Journal of Communication
Disorders, 27, 91-106.

Hickok, G., & Poeppel, D. (2000). Towards a functional neuroanatomy of speech
perception. Trends in Cognitive Sciences, 4, 131-138.









Hoffmeister, R. J. (2000). A piece of the puzzle: ASL and reading comprehension in deaf
children. In C. Chamberlain, J. P. Morford, & R. I. Mayberry (Eds.), Language
Acquisition by Eye (pp. 165-89). Mahwah, NJ: Earlbaum.

Hoien, T., Lundberg, L., Stanovich, K. E., & Bjaalid, L. K. (1995). Components of
phonological awareness. Reading & Writing: An Interdisciplinary journal, 7, 171-
188.

Holt, J. (1994). Classroom attributes and achievement test scores for deaf and hard of
hearing students. American Annals of the Deaf 139, 430-437.

Hulme, C., & Snowling, M. (1992). Phonological deficits in dyslexia: a reappraisal of the
verbal deficit hypothesis. In N. Singh & I. Beale (Eds.), Current perspectives in
learning disabilities, (pp. 270-331). New York: Springer-Verlag.

Hurford, D. P., Darrow, L. J., Edwards, T. L., Howerton, C. J., Mote, C. R., Schauf, J. D.,
& Coffey, P. (1993). An examination of phonemic processing abilities in children
during their first-grade year. Journal of Learning Disabilities, 26, 167-177.

Hynd, G. W., Semrud-Clikemand, M., Lorys, A. R., Novey, E. S., & Eliopulos, D.
(1990). Brain morphology in developmental dyslexia and attention deficity-
hyperactivity disorder (ADHD): Morphometric analysis of MRI. Archives of
Neurology, 47, 919-926.

Jarrold, C., Baddeley, A. D., & Phillips, C. E. (2002). Verbal short-term memory in
Down syndrome: A problem of memory, audition, or speech? Journal of
Speech,Language, and Hearing Research, 45, 531-544.

Jenkins, R., & Bowen, L. (1994). Facilitating development of preliterate children's
phonological abilities. Topics in Language Disorders, 14, 26-39.

Justas, S., & Wilkinson, G. S. (1984). Wide Range Achievement Test Revised.
Wilmington, DE: Justas Associates.

Kail, R. (1991). Developmental change in speed of processing during childhood and
adolescence. Psychological Bulletin, 109, 490-501.

Kail, R., & Hall, L. K. (1994). Procesisng speed, naming speed, and reading.
Developmental Psychology, 30, 949-954.

Kail, R., Hall, L. K., & Caskey, B. J. (1999). Processing speed, exposure to print, and
naming speed. Applied Psycholinguistics, 20, 303-314.

Katz, R. (1986). Phonological deficiencies in children with reading disability: Evidence
from an object-naming task. Cognition, 22, 225-257.

Katzir, T., Wolf, M., O'Brien, Kennedy, B., Lovett, M., & Morris, R. (2006). Reading
Fluency: The Whole is more than the Parts. Annals of Dyslexia. 56, 51-82.









Keith, R. W. (1986). SCAN. A Screening Test for Auditory Processing Disorders. San
Antonio, TX: The Psychological Corporation, Harcourt, Brace, Jovanovich, Inc.

Keith, R. W., Rudy, J., Donahue, P. A., & Katbamna, B. (1989). Comparison of SCAN
results with other auditory and language measures in a clinical population. Ear
and Hearing. 10, 382-386.

King, C.M., & Quigley, S.P. (1985). Reading and deafness. San Diego, CA: College-Hill
Press.

Kramer, J. H., Knee, K., & Delis, D. C. (2000). Verbal memory impairments in dyslexia.
Archieves of Clinical Neuropsychology. 15, 83-93.

Kushch, A., Gross-Glenn, K., Jallad, B., Lubs, H., Rabin, M., Feldman, E., & Duara, R.
(1993). Temporal lobe surface area measurements on MRI in normal and dyslexic
readers. Neuropsychologia, 31, 811-821.

Landerl, K., Wimmer, H., & Frith, U. (1997). The impact of orthographic consistency on
dyslexia: a German-English comparison. Cognition, 63, 315-334.

Larsen, J. P., Hoien, T., Lundberg, I., & Odegaard, H. (1990). MRI evaluation of the size
and symmetry of the planum temporale in adolescents with developmental
dyslexia. Brain Lanugage, 39, 289-301.

LaSasso, C. (1978). A national survey of materials and procedures used to teach reading
to hearing-impaired children. American Annals of the Deaf 123, 22-30.

LaSasso, C. (1987). Survey of reading instruction for hearing impaired students in the
United States. The Volta Review, 89, 85-98.

LaSasso, C., & Mobley, R. T. (1997). National Survey of reading instruction for deaf or
hard-of-hearing students in the U.S. The Volta Review, 99, 31-59.

Laybaert, J., & Charlier, B. (1996). Visual speech in the head: The effect of Cued-Speech
on rhyming, remembering, and spelling. Journal of Deaf Studies and Deaf
Education, 1, 234-248.

Leonard, C. M., Eckert, M. A., Lombardino, L. J., Oakland, T., Kranzler, J., Mkohr, C.
M., King, W. M., & Kreeman, A. (2001). Anatomical risk factors for
phonological dyslexia. Cerebral Cortex, 11, 148-157.

Leonard, C. M., Puranik, C., Kuldau, J. M., & Lombardino, L. J. (1998). Normal
variation in the frequency and location of human auditory cortex landmarks:
Heschl's gyrus: Where is it? Cerebral Cortex, 8, 397-406.









Leonard, C. M., Voeller, K. K., Lombardino, L. J., Morris, M. K., Hynd, G. W.,
Alexander, A. W., Andersen, H. G. Garofalakis, M., Honeyman, J. C., Mao, J.,
Agee, F. & Staab, E. V. (1993). Anomalous cerebral structure in dyslexia revealed
with magnetic resonance imaging. Archives of Neurology, 50, 461-469.

Leonard, L. (1989). Language learnability and specific language impairment in children.
Applied Psycholinguistics, 10, 179-202.

Leonard, L. (1995). Functional categories in the grammars of children with specific
language impairment. Journal of Speech and Hearing Research, 38, 1270-1283.

Leonard, L. (1998). Specific Language Impairment. MIT Press, Cambridge, MA.

Levelt, W. J. M. (1989). Speaking: from intention to articulation. Cambridge,
Massachusetts: The MIT Press.

Leybaert, J. (1993). Reading in the deaf: The roles of phonological codes. In M.
Marschark & M. D. Clark (Eds.), Psychological perspectives on deafness (pp.
269-309). Hillsdale, NJ: Lawrence Erlbaum Associates.

Leybaert, J. (1998). Phonological representations in deaf children: the importance of
early linguistic experience. Scandinavian Journal of Psychology, 39, 169-173.

Leybaert, J., & Alegria, J. (1990). Cued speech and the acquisition of reading by deaf
children. The Cued Speech Journal, 4, 24-38.

Liberman, I. Y., Shankweiler, D., & Liberman, A. M. (1989). The alphabetic principle
and learning to read. In D. Shankweiler & I. Y. Lieberman (Eds.), Phonology and
reading disability: Solving the reading puzzle (pp. 1-33). Ann Arbor: University
of Michigan Press.

Lichtenstein, E. H. (1998). The relationships between reading processes and English
skills of deaf college Students. Journal of Deaf Studies and Deaf Education, 3,
80-134.

Lieberman, I. Y., & Shankweiler, D. (1985). Phonology and problems of learning to read
the write. Remedial and Special Education, 6, 8-17.

Lindsay, P. H., Shapiro, A., Musselman, C., & Wilson, A. (1988). Predicting language
development in deaf children using subscales of the Leiter International
Performance Scale. Canadian Journal of Psychology, 42, 144-162.

Lovett, M. W., Steinbach, K. A., & Frijters, J. C. (2000). Remediating the core deficits
of developmental reading disability: a double-deficit perspective. Journal of
Learning Disabilities, 33, 334-358.

Lyon, G. R. (1998). Why reading is not a natural process. Educational Leadership, 55, 4-
19.









Lyon, G. R., Shaywitz, S. E. & Shaywitz, B. A. (2003). A definition of dyslexia. Ann
Dyslexia, 53, pp. 1-14.

Lyxell, B., Anderson, U., Borg, E., & Ohlsson, I. S. (2003). Working-memory capacity
and phonological processing in deafened adults and individuals with a severe
hearing impairment. International Journal of Audiology, 42, Suppl 1:S86-89.

Lyxell, B., Ronnberg, J. & Samuelsson, S. (1994). Internal speech functioning and
speechreading in deafened and normal hearing adults. Scandinavian Audiology,
23, 179-185.

Majerus, S., Amand, P., Boniver, V, Demanez, J. P., Demanez, L., & Linden, M. V.
(2005). A quantitative and qualitative assessment of verbal short-term memory
and phonological processing in 8-year-olds with a history of repetitive otitis
media. Journal of Communication Disorders, 9, 1-26.

Manis, F. R., Doi, L. M., & Bhadha, B. (2000). Naming speed, phonological awareness,
and orthographic knowledge in second graders. Journal of Learning Disabilities,
33, 325-333.

Manis, F. R., McBride-Chang, C., Seidenberg, M. S., Keating, P., Doi, L. M., Munson,
B., & Petersen, A. (1997). Are speech perception deficits associated with
developmental dyslexia? Journal of Experimental Child Psychology, 66, 211-235.

Mann, V. A. (1993). Phoneme awareness and future reading ability. Journal of Learning
Disabilities, 26, 259-269.

Mann, V. A. (1998) Language problems: A key to early reading problems. In B. Wong
(Ed.) Learning about Learning disabilities, 2ndEd. (pp. 163-202). Chicago.

Marchman, V. A. & Bates, E. (1994). Continuity in lexical and morphological
development: A test of the critical mass hypothesis. Journal of Child Language,
21, 339-366.

Marschark, M. & Harris, M. (1996). Success and failure in learning to read: the special
case of deaf children. In C. Cornoldi & J. Oakhill (Eds.), Reading comprehension
difficulties: Processes and intervention (pp 279 300). Hillsdale, NJ: Lawrence
Erlbaum Associates Inc.

Marschark, M. (1997). Raising and educating a deaf child. New York: Oxford University
Press.

Marschark, M. (2003). Cognitive functioning in deaf adults and children. In M.
Marschark & P. E. Spencer (Eds.), Oxford handbook of deaf studies, language,
and education (pp. 464-477). New York: Oxford University Press.

Marschark, M., & Everhart, V. S. (1999). Problem solving by deaf and hearing children:
Twenty questions. Deafness and Education International, 1, 63-79.









Marschark, M., & Lukomski, J. (2001). Understanding language and learning in deaf
children. In M. D. Clark, M. Marschark, & M. Karchmer (Eds.), Cognition,
context, and deafness (pp. 71-86). Washington, DC: Gallaudet University Press.

McConkey Robbins, A., Burton Koch, D., Osberger, M.J., Zimmerman-Phillips, S.,
Kishon-Rabin, L. (2004). Effect of age at cochlear implantation on auditory skill
development in infants and toddlers. Archives of Otolaryngology, Head, and Neck
Surgery, 130, 570-574.

McDougall, S., Hulme, C., Ellis, A., & Monk, A. (1994). Learning to read: The role of
short-term memory and phonological skills. Journal of Experimental Child
Psychology, 58, 112-133.

Merzenich, M. M., Jenkins, W. M., Johnson, P., Schreiner, C., Miller, S., & Tallal, P.
(1996). Temporal processing deficits of language-learning impaired children
ameliorated by training. Science, 271, 77-80.

Metsala, J. L. (1997). Spoken word recognition in reading in reading disabled children.
Journal of Educational Psychology, 89, 159-169.

Metsala, J. L. (1999). Young children's phonological awareness and nonword repetition
as a function of vocabulary development. Journal of Educational Psychology, 91,
3-19.

Metsala, J. L., & Walley, A. C. (1998). Spoken vocabulary growth and the segmental
restructuring of lexical representations: precursors to phonemic awareness and
early reading ability. In J. L. Metsala & L. C. Ehri (Eds.), Word Recognition in
Beginning Literacy, 89-120. Hillsdale, NJ: Erlbaum.

Misra M., Katzir, T., Wolf, M., & Poldrack, R. A. (2004). Neural systems of rapid
automatized naming in skilled readers: Unraveling the RAN-reading relationship.
Scientific Studies of Reading,. 8, 241-256.

Misra, M., Katzir, T., Wolf, M. & Poldrack, R.A. (2004). Neural systems for rapid
automatized naming (RAN) in skilled readers: Unraveling the RAN-reading
relationship. Scientific Studies in Reading. 8, 241-256.

Mody, M. (2003). Rapid auditory processing deficits in dyslexia: A commentary on two
differing views. Journal of Phonetics, 31, 529-539.

Mody, M., Studdert-Kennedy, M., & Brady, S. (1997). Speech perception deficits in poor
readers: Auditory processing or phonological coding? Journal of Experimental
Child Psychology, 64, 199-231.

Mody, R., Schwartz, G., Gravel, J.S., & Ruben, R. J. (1999). Speech perception and
verbal memory in children with and without histories of otitis media. Journal of
Speech, Language, and Hearing Research, 42, 1069-1079.









Moeller, M. P. (2000). Early intervention and language development in children who are
deaf and hard of hearing. Pediatrics, 106, 43-51.

Moeller, M. P., & Luetke-Stahlman, B. (1990). Parents' use of signing exact English: A
descriptive analysis. Journal of Speech and Hearing Disorders, 55, 327-38.

Moncrieff, D. W. (2002). Auditory Processing Disorders and Dyslexic Children.
http://www.audiologyonline.com/articles/pf article_detail. asp?article_id=369.

Moore, B. C .J. (1995). Perceptual Consequences of Cochlear Damage, Oxford
University Press, Oxford.

Morals, J., Alegria, J., & Content, A. (1987). The relationships between segmental
analysis and alphabetic literacy: An interactive view. Cahiers de Psychologie
Cognitive, 7, 415-38.

Morford, J. P., & Mayberry, R. I. (2000). A reexamination of "early exposure" and its
implication for language acquisition by eye. In C. Chamberlain, J. P. Morford, &
R. I. Mayberry (Eds.), Language Acquisition by Eye (pp. 165-89). Mahwah, NJ:
Earlbaum.

Morris, R. D., Stuebing, K. K, Fletcher, J. M., Shaywitz, S. E., Lyon, G. R., Shankweiler,
D. P. (1998). Subtypes of reading disability: Variability around a Phonological
Core. Journal of Educational Psychology, 90, 347-373.

Musiek, F. E., & Chermak, G. D. (1994). Three commonly asked questions about central
auditory processing disorders: Assessment. American Journal of Audiology, 3,
23-27.

Musiek, F. E., & Reeves, A. G. (1990). Asymmetries of the auditory areas of the
cerebrum. Journal of the American Academy ofAudiology, 1, 240-245.

Musselman, C. (2000). How do children who can't hear learn to read an alphabetic script?
A review of the literature on reading and deafness. Journal of Deaf Studies and
Deaf Education, 5, 9-31.

Muter, V., Hulme, C., Snowling, M., & Taylor, S. (1997), Segmentation, not rhyming,
predicts early progress in learning to read. Journal of Experimental Child
Psychology, 65, 370-396.

Nairne, J. S. (1990). A feature model of immediate memory. Memory & Cognition, 18,
251-269.

Nathan, L., Stackhouse, J., Goulandris, N., & Snowling, M. (2004). The development of
early literacy skills among children with speech difficulties: A test of the "critical
age hypothesis." Journal of Speech, Language, and Hearing Research, 47, 377-
391.









Neath, I., & Nairne, J. S. (1995). Word-Length effects in immediate memory:
Overwriting trace decay theory. Psychonomic Bulletin & Review, 2, 429-441.

Nemeth, S. (1992). The role of phonology and context in word recognition in hearing
impaired readers. Unpublished master's thesis. Montreal, Quebec, Canada:
McGill University.

Neuhaus, G. F., Foorman, B. R., Francis, D. J., & Carlson, C. (2001). Measures of
information processing in Rapid Automatized Naming (RAN) and their relation to
reading Journal of Experimental Child Psychology, 78, 359-373.

Nicholson, R. I., & Fawcett, A. J. (1990). A New Framework for Dyslexia Research?
Cognition, 35, 159-182.

Nielsen, D. C., & Luetke-Stahlman, B. (2002). Phonological Awareness: One key to the
reading proficiency of deaf children. American Annals of the Deaf 147, 11-19.

Nittrouer, S. (1996). Discriminability and perceptual weighting of some acoustic cues to
speech perception by 3-year-olds. Journal Speech and Hearing Research, 39,
278-297.

Nittrouer, S. (1996). The relation between speech perception and phonemic awareness:
Evidence from low-SES children and children with chronic OM. Journal of
Speech and Hearing Research. 39, 1059-1070.

Nittrouer, S., & Burton, L. T. (2005). The role of early language experience in the
development of speech perception and phonological processing abilities:
Evidence from 5-year-olds with histories of otitis media with effusion and low
socioeconomic status. Journal of Communication Disorders, 38, 29-63.

Norbury, C. F., Bishop, D. V. M., & Briscoe, J. (2001). Production of English finite verb
morphology: A comparison of SLI and mild-moderate hearing impairment.
Journal of Speech, Language, and Hearing Research, 44, 165-178.

Nozza, R. J., Bluestone, C. D., Kardatzke, D., Bachman, R. (1994). Identification of
middle ear effusion by aural acoustic admittance and otoscopy. Ear and Hearing.
15, 310-23.

O'Connor, R. E., Jenkins, J. R., Leicester, N., & Slocum, T. A. (1993). Teaching
phonological awareness to young children with learning disabilities. Exceptional
Children, 59, 532-546.

Oakhill, J. & Kyle, F. (2000). The Relation between Phonological and Working Memory.
Journal of Experimental Child Psychology, 75, 152-164.

O'Brien, B., Wolf, M., Morris, R., & Lovett, M. (2004). An Investigation of subtypes of
developmental dyslexia using taxometric classification. Annual Meeting Brain,
Neurosciences, and Education SIG Paper Session.









Olson, R. Forsberg, H., Wise, B., & Rack, J. (1994). Measurement of word recognition,
orthographic, and phonological skills. In G. R. Lyon (Ed.), Frames of reference
for the assessment of learning disabilities: New views on measurement issues
(pp. 243-277). Baltimore: Brookes.

Ostrander, C. (1998). The relationship between phonological coding and reading
achievement in deaf children: Is cued speech a special case? Unpublished
manuscript.

Padden, C., & Ramsey, C. (2000). Americans Sign Language and reading ability in deaf
children. In C. Chamberlain, J. P. Morford, & R. I. Mayberry (Eds.), Language
Acquisition by Eye (pp. 165-89). Mahwah, NJ: Earlbaum.

Paul, P. V. (1998). Literacy and deafness. Boston: Allyn and Bacon. Paul, P. V. (2001).
Language and deafness. San Diego, CA: Singular.

Paulesu, E., Demonet, J., Fazio, F., McCrory, E., Chanoine, V., Brunswick, N., Cappa,
S., Cossu, G., Habib, M., Frith, C., Frith, U. (2001). Dyslexia: cultural diversity
and biological unity, Science 291, 2165-2167.

Penhune, V. B., Zatorre, R. J., MacDonald, J. D., & Evens, A. C. (1996).
Interhemispheric anatomical differences in human primary auditory cortex;
probabilistic mapping and volume measurement from magnetic resonance scans.
Cerebral Cortex, 6, 661-672.

Perfetti, C. A. (1985). Reading ability. New York: Oxford University Press.

Perfetti, C. A.(1992). The representation problem in reading acquisition. In P. Gough, L.
Ehri, & R. Treiman (Eds.), Reading acquisition (pp. 107-143). San Diego:
Acadmeic Press.

Perfetti, C. A., & Sandak, R. (2000). Reading optimally builds on spoken language:
Implications for deaf literacy. Journal of Deaf Studies and Deaf Education, 5, 31-
50.

Peters, S. A., Grievink, E. H., van Bon, W. H., & Schilder, A. G. (1994). The effects of
early bilateral otitis media with effusion on educational attainment: a prospective
cohort study. Journal of Learning Disabilities. 2 7, 111-21.

Plapinger, D. S., & Sikora, D. M. (1995). The Use of Standardized Test Batteries in
Assessing the Skill Development of Children With Mild-to-Moderate
Sensorineural Hearing Loss. Language, Speech, and Hearing Services in Schools,
26, 39-44.

Porpodas, C. D. (1999). Patterns of phonological and memory processing in beginning
readers and spellers of Greek. Journal of Learning Disabilities, 32, 406-416.









Powell, D., Stainthorp, R., Stuart, M, Garwood, H & Quinlan, P. (2007). An experimental
comparison between rival theories of rapid automatized naming performance and
its relationship to reading. Journal of Experimental Child Psychology, 98, 46-68.

Ramus, F. (2001a). Dyslexia Talk of two theories. Nature, 412, 393-395.

Ramus, F. (2001b). Outstanding questions about phonological processing. Dyslexia, 7,
197-216.

Ramus, F. (2002). The neural basis of reading acquisition. In M. S. Gazzaniga (Ed.), The
Cognitive Neurosciences (3rd ed., pp. 815-824). Cambridge, MA: MIT Press.

Ramus, F. (2003). Developmental dyslexia: specific phonological deficit or general
sensorimotor dysfunction? Current Opinion in Neurobiology, 13, 212-218.

Ramus, F., Pidgeon, E., & Frith, U. (2003). The relationship between motor control and
phonology in dyslexic children. Journal of Child Psychology and Psychiatry, 44
(5), 712-722.

Ramus, F., Rosen, S., Dakin, S. C., Day, B. L., Castellote, J. M., White, S., & Frith, U.
(2003). Theories of developmental dyslexia: Insights from a multiple case study
of dyslexic adults. Brain, 126, 1-25.

Rayner, K., Foorman, B. R., Perfetti, C. A., Pesetsky, D. & Seidenberg, M. S. (2001).
How psychological science informs the teaching of reading. Psychological
Science in the Public Interest, 2, 31-74.

Rice, M., & Wexler, K. (1996). Toward Tense as a clinical marker of specific language
impairment in English-speaking children, Journal of Speech and Hearing
Research 39, 1239-1257.

Roberts, J. E., Burchinal, M. R., & Zeisel, S. A. (2002). Otitis media in early childhood in
relation to children's school-age language and academic skills. Pediatrics, 110,
696-706.

Roberts, J. E., Hunter, L., Gravel, J., Rosenfeld, R., Berman, S., Haggard, M., Hall, J.,
Lannon, C., Moore, D., Lynne, V., & Wallace, I. (2004). Otitis media, hearing
loss, and language learning: controversies and current research. Journal of
Developmental & Behavioral Pediatrics, 25, 110-122.

Roberts, J. E., Rosenfeld, R. M., & Zeisel, S. A. (2004). Otitis media and speech and
language: a meta-analysis of prospective studies. Pediatrics, 113, 238-48.

Rom, A., & Leonard, L. (1990). Interpreting deficits in grammatical morphology in
specifically language-impaired children: Preliminary evidence from Hebrew.
Clinical Linguistics and Phonetics, 4, 93-105.









Rosen, S. (2003) Auditory processing in dyslexia and specific language impairment: Is
there a deficit? What is its nature? Does it explain anything? Journal of Phonetics,
31, 509-527.

Ruben, R. J. (1972). The ear. In H. Barnett & A. Einhorn (Eds.) Pediatrics (15th ed., pp.
1881-92). New York: Appleton, Century, Crofts.

Rutter, M., & Yule, W. (1975). The concept of specific reading retardation. Journal of
Child Psychology and Psychiatry, 16,181-197.

Scanlon, D. M. & Vellutino, F. R. (1997). A comparison of the instructional backgrounds
and cognitive profiles of poor, average, and good readers who were initially
identified as at risk for reading failure. Scientific Studies ofReadint,. 1, 191-215.

Schirmer, B. R. (1994). Language and Literacy Development in Children Who are Deaf
Merrill: Merrill/ Macmillan Publishing Company.

Schirmer, B. R., & McGough, S. M. (2005). Teaching reading to children who are deaf:
Do the conclusions of the National Reading Panel apply? Review of Educational
Research, 75, 83-117.

Service, E., Maury, S., Luotoniemi, E. (2005). Forgetting and redintegration of
consonants and vowels in pseudoword lists. Memory, 13, 340-348.

Shankweiler, D., & Crain, S. (1986) Language mechanisms and reading disorder: A
modular approach. Cognition, 24, 139-168.

Share, D. L. (1995). Phonological recoding and self-teaching: Sine qua non of reading
acquisition. Cognition, 55, 151-218.

Share, D. L. (1999). Phonological recoding and orthographic learning: A direct test of the
self-teaching hypothesis. Journal of Experimental Child Psychology, 72, 95-129.

Share, D. L., Chalmers, D., Silva P. A., & Stewart, I. A. (1986). Reading disability and
middle ear disease. Archives of Disease in Childhood, 61, 400-401,

Sharma, A., Dorman, M., & Spahr, A. (2002). A sensitive period for the development of
the central auditory system in children with cochlear implants: Implications for
age of implantation, Ear & Hearing 23, 532-539.

Shaywitz, S. (1998). "Current concepts: Dyslexia." The New England Journal of
Medicine, 338, 307-312.

Shaywitz, S. (2003). Overcoming dyslexia: A new and complete science-based program
for reading problems at any level. New York: Knopf.









Shaywitz, S. E., Fletcher, J. M., Holahan, J. M., Shneider, A. E., Marchione, K. E.,
Stuebing, K. K., et al. (1999). Persistence of dyslexia: The Connecticut
Longitudinal Study at adolescence. Pediatrics, 104, 1351-1359.

Shaywitz, S., & Shaywitz, B. (2005). Dyslexia (Specific Reading Disability). Biological
Psychiatry, 57, 1301-1309.

Shaywitz, S., Fletcher, J., Holahan, J., Shneider, A., Marchione, K., Stuebing, K., et al.
(1999). Persistence of dyslexia: The Connecticut Longitudinal Study at
adolescence. Pediatrics, 104(6), 1351-1359.

Siok, W. T., & Fletcher, P. (2001). The role of phonological awareness and visual-
orthographic skills in Chinese reading acquisition. Developmental Psychology,
37, 886-899.

Snow, C. E., Bums, M. S., & Griffin, P. (1998). Preventing Reading Difficulties in Young
Children. Washington, DC: National Academy Press.

Snowling, M. & Hulme, C. (1989). A longitudinal case study of developmental
phonological dyslexia. Cognitive Neuropsychology 6, 379-401.

Snowling, M. J. (1998). Reading development and its difficulties. Educational and Child
Psychology, 15, 44-58.

Snowling, M. J. (2001). From language to reading and dyslexia. Dyslexia, 7, 37-46.

Snowling, M. J., Goulandris, N., Bowlby, M., & Howell, P. (1986). Segmentation and
speech perception in relation to reading skill: a developmental analysis. Journal
of Experimental Child Psychology, 41, 489-507.

Snowling, M., Bishop, D. V. M., & Stothard, S. E. (2000). Is pre-school language
impairment a risk factor for dyslexia in adolescence? Journal of ChildPsychology
and Psychiatry, 41, 587-600.

Snowling, M., Nation, K., Moxham, P., Gallagher, A., & Frith, U. (1997). Phonological
processing skills of dyslexic students in higher education: A preliminary report.
Journal of Research in Re, ding. 29, 31-41.

Spector, J. (1995). Phonemic awareness training: Application of principles of direct
instruction. Reading and Writing Quarterly, 11, 37-51.

Stanovich, K. E. (1980). Toward an interactive-compensatory model of individual
differences in the development of reading fluency. Reading Research Quarterly,
16, 32-71

Stanovich, K. E. (1994). Annotation: Does dyslexia exist? Journal of ChildPsychology
and Psychiatry, 35, 579-595.









Stanovich, K. E. (1998). Twenty-five years of research on the reading process: The grand
synthesis and what it means for our field. In T. Shanahan & F. Rodriguez-Brown
(Eds.), Forty-Seventh Yearbook of the National Reading Conference (pp. 44-58).
Chicago: NRC.

Stanovich, K. E., & Siegal, L. S. (1994). Phenotypic performance profile of children with
reading disabilities: A regresson-based test of the phonological-core variable-
difference model. Journal of Educational Psychology, 86, 24-53.

Stanovich, K.E. (1981). Relationships between word decoding speed, general name-
retrieval ability, and reading progress in first-grade children. Journal of
Educational Psychology, 73, 809-815.

Stanovich, K.E. (1986). Matthew Effects in Reading: Some consequences of individual
Differences in the acquisition of literacy. Reading Research Quarterly, 21, 360-
407.

Stanovich, K.E. (1988). Explaining the differences between the dyslexic and the Garden-
variety poor reader: The phonological-core variable difference modes. Journal of
Learning Disabilities, 21, 590-612.

Stanovich, K.E. (1992). Speculations on the causes and consequences of individual
Differences in early reading acquisition. In P. Gough, L. Ehri, and R. Treiman
(Eds.), Reading acquisition (pp. 307-342). Hillsdale, NJ: Lawrence Erlbaum
Associates.

Stanovich, K.E. (1993). Does reading make you smarter? Literacy and the development
of verbal intelligence. In H. Reese (Ed.), Advances in Child Development and
Behavior (Vol. 24, pp. 133-180). San Diego, CA: Academic Press.

Stanovich, K.E., and West, R.F. (1989). Exposure to print and orthographic processing.
Reading Research Quarterly, 24, 402-433.

Stelmachowicz, P.G., Pittman, A.L., Hoover, B.M., & Lewis, D.E. (2004). Novel-word
learning in children with normal hearing and hearing loss. Ear and Hearing, 25,
47-56.

Sterne, A., & Goswami, U. (1998). Deaf children can make rhymes: evidence from a
simple rhyme task. Surdite et access a la langue ecrite. 137-155.

Strong, M., & Prinz, P. (2000). Is American Sign Language skill related to English
literacy? In C. Chamberlain, J. P. Morford, & R. I. Mayberry (Eds.), Language
Acquisition by Eye (pp. 165-89). Mahwah, NJ: Earlbaum.

Svirsky, M. A., Robbins, A. M., Kirk, K. I., Pisoni, D. B., & Miyamoto, R. T. (2000).
Language development in profoundly deaf children with cochlear implants.
Psychological Science, 11, 153-158.









Svirsky, M. A., Stallings, L. M., Lento, C. L., Ying, E., & Leonard, L. B. (2002).
Grammatical morphologic development in paediatric cochlear implantation users
may be affected by the perceptual prominence of the relevant markers. Annals of
Otology Rhinology and Laryngology, 189(Suppl.), 109-112.

Swan, D., & Goswami, U. (1997a). Phonological awareness deficits in developmental
dyslexia and the phonological representations hypothesis. Journal of
Experimental Child Psychology, 66, 18-41.

Swan, D., & Goswami, U. (1997b). Picture-naming deficits in developmental dyslexia:
The phonological representations hypothesis. Brain and Language, 56, 334-353.

Tallal, P. (1980). Auditory temporal perception, phonics, and reading disabilities in
children. Brain and Language, 9, 182-98.

Tallal, P., Miller, S. L., Bedi, G., Byma, G., Wang, X., Nagarajan, S. S., Schreiner, C.,
Jenkins, W. M., & Merzerich, M. M. (1996). Language comprehension in
language-learning impaired children improved with acoustically modified speech.
Science, 271, 81-84.

Tallal, P., Miller, S., & Fitch, R. (1993). Neurological basis of speech: A case for the
prominence of temporal processing? In P. Tallal, A. Galaburda, R. Llinas, & C.
von Euler (Eds.), Annals of the New York Academy of Sciences: Vol. 682.
Temporal information processing in the nervous system (pp. 27-37). New York:
New York Academy of Sciences.

Tijms, J. (2004). Verbal memory and phonological processing in dyslexia. Journal of
Research in Re, din. 27, 300-3 10.

Torgesen, J. K. (1997). The prevention and remediation of reading difficulties:
Evaluating what we know from research. Journal ofAcademic Language Therapy,
1, 11-47.

Torgesen, J. K. (2000). Individual differences in response to early interventions in
reading: The lingering problem of treatment resisters. Learning Disabilities
Research and Practice, 51(1), 55-64.

Torgesen, J. K., & Bryant, B. (1994). Test of Phonological Awareness. Austin, TX: PRO-
Ed.

Torgesen, J. K., & Burgess, S. R. (1998). Consistency of reading-related phonological
processes throughout early childhood: Evidence from longitudinal-correlational
and instructional studies. In J. Metsala & L. Ehri (Eds.). Word Recognition in
Beginning Reading. Hillsdale, NJ: Lawrence Erlbaum Associates.

Torgesen, J. K., & Mathes, P. G. (2000). A basic guide to understanding, assessing, and
teaching phonological awareness. Austin, TX: Pro-Ed.









Torgesen, J. K., & Wagner, R. K. (1998). Alternative diagnostic approaches for specific
developmental reading disabilities. Learning Disabilities Research and Practice,
13, 220-232.

Torgesen, J. K., Rashotte, C. A., & Wagner, R. K. (1994). Longitudinal studies of
phonological processing and reading. Journal of Learning Disabilities, 2 7, 276-
286.

Torgesen, J. K., Wagner, R. K., & Rashotte, C. A. (1997a). The prevention and
remediation of severe reading disabilities: Keeping the end in mind. Scientific
Studies of Red, ing. 1, 217-234.

Torgesen, J. K., Wagner, R. K., Rashotte, C. A., Burgess, S., & Hecht, S. (1997b).
Contributions of phonological awareness and rapid automatic naming ability to
the growth of word-reading skills in second- to fifth-grade children. Scientific
Studies of Reading,. 1, 161-185.

Torgesen, J.K., Wagner, R. K., & Rashotte, C.A. (1999). Test of WordReading
Efficiency.Austin, TX: PRO-ED Publishing, Inc.

Treiman, R., & Bourassa, D. (2000). Children's written and oral spelling. Applied
Psycholinguistics, 21, 183-204.

Treiman, R., & Bourassa, D. (2000). The development of spelling skill. Topics in
Language Disorders, 20, 1-18.

Treiman, R., & Hirsh-Pasek, K. (1983). Silent reading: Insights from second-generation
deaf readers. Cognitive Psychology, 15, 39-65.

Trezek, J., & Malmgren, K. W. (2005). The Efficacy of Utilizing a Phonics Treatment
Package with Middle School Deaf and Hard-of-Hearing Students. Journal of Deaf
Studies and DeafEducation, 10, 256 271.

Vellutino, F. R. (1979). Dyslexia: Theory and research. Cambridge, MA: MIT Press.

Vellutino, F. R., Fletcher, J.M., Snowling, M. J., & Scanlon, D. M. (2004). Specific
reading disability (dyslexia): What have we learned in the past four decades?
Journal of Child Psychology and Psychiatry, 45, 2-40.

Vellutino, F. R., Scanlon, D. M., Sipay, E. R., Small, S. G., Pratt, A., Chen, R. S., &
Denckla, M. B. (1996). Cognitive profiles of difficult to remediate and readily
remediated poor readers: Early intervention as a vehicle for distinguishing
between cognitive and experiential deficits as basic causes of specific reading
disability. Journal of Educational Psychology, 88, 601-638.

Wagner, R. K., Torgesen, J. K., & Rashotte, C. A. (1994). Development of reading-
related phonological processing abilities: New evidence of bidirectional causality
from a latent variable longitudinal study. Developmental Psychology, 30, 73-87.









Wagner, R. K., Torgesen, J.K., & Rashotte, C.A. (1999). Comprehensive Test of
Phonological Processes. Austin, TX: PRO-ED Publishing, Inc.

Wagner, R. K., Torgeson, J. K., Laughon, P. L., Sommons, K., & Rashotte, C. A. (1993).
Development of young readers' phonological processing abilities. Journal of
Educational Psychology, 85, 83-103.

Wagner, R., & Torgesen, J. (1987). The nature of phonological processing and its causal
role in the acquisition of reading skills. Psychological Bulletin, 101, 192-212.

Wake, M., & Poulakis, Z. (2004). Slight and mild hearing loss in primary school
children. Journal of Paediatric Child Health, 40, 11-13.

Wake, M., Hughes, E., Poulakis, Z., Collins, C., & Rickards, W. (2004). Outcomes of
children with mild-profound congenital hearing impairment at 7-8 years: a
population study. Ear and Hearing. 25, 1-8.

Walley, A. (1993). The role of vocabulary development in children's spoken word
recognition and segmentation ability. Developmental Review, 13, 286-350.

Wandel, J. (1989). Use of Internal Speech by Hearing and Hearing-Impaired Students in
Oral, Total Communication, and Cued Speech Programs. Unpublished Doctoral
Dissertation, Columbia University, New York.

Wesseling, R. F., & Reitsma, P. (1998). Phonemically aware in a hop, skip, and a jump.
In P. Reitsma & L. Verhoeven (Eds.) Problems and interventions in literacy
development (pp. 81-94). Dordrecht: Kluwer.

Wiederholt, J. L., & Bryant, B. R. (2001). Gray Oral ReadingTests, Fourth Edition.
Austin, TX: Pro-Ed.

Williams, K. T. (1997). Expressive Vocabulary Test Circle Pines, MN: American
Guidance Service. Leonard, L. (1998). Children ii ith specific language
impairment. Cambridge, MA: MIT Press.

Wimmer, H. (1993). Characteristics of developmental dyslexia in a regular writing
system. Applied Psycholinguistics, 14, 1-33.

Wolf, M. & Bowers, P. G. (2000). The question of naming-speed deficits in
developmental reading disability: An introduction to the double-deficit
hypothesis. Journal of Learning Disabilities, 33, 322-324.

Wolf, M. (1997). A provisional, integrative account of phonological naming speed
deficits in dyslexia: Implications for diagnosis and intervention. In B. Blachman
(Ed.). Foundations of reading acquisition and dyslexia (pp. 67-93). Mahwah, NJ:
Lawrence Erlbaum Associates.









Wolf, M., & Bowers, P. G. (1999). The double-deficit hypothesis for the developmental
dyslexia. Journal of Educational Psychology, 91, 124.

Wolf, M., Bowers, P. G., & Biddle, K. (2000). Naming-speed processes, timing, and
reading: A conceptual review. Journal of Learning Disabilities, 33, 387-407.

Wolf, M., Goldberg O'Rourke, A., Gidney, C., Lovett, M., Cirino, P., & Morris, R.
(2002). The second deficit: An investigation of the independence of phonological
and naming-speed deficits in developmental dyslexia. Reading and Writing, 15,
43-72.

Woodcock, R. W. (1987). Woodcock reading mastery test Revised. Circle Pines,
MN:American Guidance Service.

Woodcock, R. W., & Johnson, M. B. (1989). Woodcock Johnson Psycho-Educational
Battery-Revised. Allen, TX: DLM Teaching Resources.

Woodcock, R.W. (1987). Woodcock Reading Mastery Tests Revised, American Guidance
Service, Circle Pines, MN.

Wren, S. (2005). The double-deficit hypothesis for decoding fluency. Personal
manuscript.

Yopp, H. K. (1988). The validity and reliability of phonemic awareness tests. Reading
Research Quarterly, 23, 159-177.

Yopp, H. K. (1992). Developing phonemic awareness in young children. Reading
Teacher, 45, 696-703.

Yoshinaga-Itano, C. (1999). Benefits of early intervention for children with hearing loss.
Otolaryngology and Clinic of North America, 32, 1089-1102.

Yoshinga-Itano, C., Sedey, A. L., & Coulter, D. (1998). Language of early- and later-
identified children with hearing loss. Pediatrics, 102, 1161-1171.

Ziegler, J. C. (2006). Do differences in brain activation challenge universal theories of
dyslexia? Brain and Language, 98, 341-343.









BIOGRAPHICAL SKETCH

Jungjun Park was born and raised in Seoul, Korea. Jungjun earned his Bachelor of Arts

degree in 1990 and a Master of Arts degree in Linguistics in 1994 from the Seoul National

University. After his doctoral coursework in Linguistics at the Seoul National University in 1997

followed by four years of teaching experience, Jungjun began his doctoral work at the

Department of Communication Sciences and Disorders at the University of Florida in August

2001 with a concentration in reading disorder and child language development. In August 2007,

Jungjun will join the faculty of the Department of Communication Sciences and Disorders at the

Baylor University in Waco, Texas as an assistant professor.





PAGE 1

1 CHARACTERISTICS OF PHONOLOGICA L PROCESSING, READING, ORAL LANGUAGE, AND AUDITORY PROCESSING SK ILLS OF CHILDREN WITH MILD-TOMODERATE SENSORINEURAL HEARING LOSS By JUNGJUN PARK A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Jungjun Park

PAGE 3

3 To my Godly wife, Mina; and to a bright future for my precious little princess, Hayoung.

PAGE 4

4 ACKNOWLEDGMENTS This research, from start to finish, spanned over approximately three y ears. First of all, I would like to acknowledge my wife (Mina) and daughter (Hayoung), precious family and friends, for their tireless support. Especiall y, Mina was more a part of this research than she realizes. I also acknowledge my four parents. I can never thank them enough for their support, encouragement, prayers, and deep faith in me. I cannot imagine how I could have finished this challenging task without their love. I would especially like to thank my Chair, Dr. Linda Lombardino for her countless hours of brainstorming, editing, encouragement, and pr ofessional advising she offered so patiently. I extend sincere gratitude to Drs. Bonnie J ohnson, Alice Holmes, and Ratree Wayland, my committee members, for their prof essional support and guidance. I also would like to acknowledge several pe ople who contributed a great deal of love, support, and encouragement. The faculty and staff of the University of Florida's Department of Communication Sciences and Disord ers provided remarkable educ ation, financial support, and training that were necessary for conceptualiz ing and completing this research project. I would especially like to ac knowledge my department chair, Dr. Chris Sapienza, for her remarkable patience with me during this unexp ectedly lengthy process and for her incessant encouragement. I extend special thanks to outst anding teachers that I met during my studies in Gainesville, including Drs. William Brown, Howard Rothman, Kenneth Logan, Scott Griffiths, Kenneth Gerhardt, Lori Altmann, Rahul Shrivastav, and Mini Shrivastav. I thank Drs. Shannon Brumfield and Judith Winga te, my clinical supervisors, for providing such a wonderful opportunity for me to finish the practicum hours. In addition, I am also thankful for the friendship and support of my coll eagues in the Reading Research Laboratory at the University of Florida: Sue Ann Eids on, Rebbecca Wiseheart, Cynthia Puranik, Heeyoung

PAGE 5

5 Park, and Sunjung Kim. They have made the experience bearable, and often times even enjoyable. I am also indebted to many of th e staff members, Ms. Idella King, Mrs. Debbie Butler, Ms. Cassie Mobley, Mrs. Addie Pons, Mr David Fleming, and Mr. Neal Musson. I also thank three research assistants, Heeyoung Park, Nicole Leshin and Taprini Spence, for their valuable assistance in completing a wide variety of tasks associated with the research. Thank yall! Financial assistance for my doc toral study and this research were graciously provided by the Department of Communi cation Sciences and Disorders at th e University of Florida, Doorae Academic Foundation, and the Dissert ation Research Grant by the College of Arts and Sciences of the University of Florida. I also would like to express my sincere appreciation to the children and parents involved in this res earch for giving their time and energy so that others may benefit. I thank Ms. Lori Lazarus, a fascinating speech pathologist serving a lot of hard-of-hearing children, for her initial supports for this study. Al so, I want to thank Drs. Melissa Riess and Ronald Kelley for their great devotions to da ta collection in the areas of audiology. I also acknowledge the staff members of the Exception Education Departments whom I worked with for data collection. This degree would not have been possible without the he lps of all of them. The following space has been saved for me to thank a special group of friends that God has brought into my life: my brothers and sisters at North Central Baptist Church in Gainesville. I thank them for the special love and prayers for my family and this uneasy project. Especially, I extend my deepest thanks to Mr. Jerre Brannen, Mr s. Pat Brannen, and my pastor Calvin Carr for their continuous words of wisdom caring heart, and prayers. Last and yet greatest, I thank Jesus Christ for His amazing mercy.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4LIST OF TABLES................................................................................................................. ..........9LIST OF FIGURES.......................................................................................................................11ABSTRACT...................................................................................................................................12 CHAP TER1 INTRODUCTION..................................................................................................................14Background of the Study........................................................................................................14Rationale and Significance of the Study.................................................................................16Study Objectives.....................................................................................................................18Brief Definition of Terms.......................................................................................................192 LITERATURE REVIEW.......................................................................................................21Introduction................................................................................................................... ..........21Phonological Core Deficit Theory.......................................................................................... 21Components of Phonological Processing...............................................................................23Phonological Awareness................................................................................................. 24Dimensions of phonological awareness................................................................... 25Factors related to order of difficulty........................................................................ 26Phonological Memory..................................................................................................... 27Rapid Naming..................................................................................................................28Cognitive Model of Phonological Processing................................................................. 29Theories on Impaired Phonology............................................................................................ 31Phonological Representation Hypothesis........................................................................ 32Lexical restructuring theory..................................................................................... 33Application of repr esentation hypothesis.................................................................35A challenging observation........................................................................................36A critical summary................................................................................................... 38Double-Deficit Hypothesis..............................................................................................39RAN as a phonological processing skill.................................................................. 39Three subtypes..........................................................................................................40Differentiation from phonology............................................................................... 41Linkage between RAN and reading process............................................................43Auditory Perceptual Deficit Hypothesis..........................................................................45Hearing loss, Phonology, and Literacy...................................................................................46Phonological Processing, Language Skills, and Hearing Loss........................................ 47Otitis media with effusion (OME)...........................................................................48Sensorineural hearing loss........................................................................................52

PAGE 7

7 Hearing Loss and Literacy...............................................................................................54Conductive hearing loss........................................................................................... 54Sensorineural heari ng loss and literacy....................................................................57Impaired phonology in children with hearing loss................................................... 58Summary..........................................................................................................................59Conductive hearing loss........................................................................................... 59Permanent SNHL.....................................................................................................603 METHODS AND MATERIALS........................................................................................... 66Introduction................................................................................................................... ..........66Setting and Participants....................................................................................................... ...67Recruitmen t Setting......................................................................................................... 67Selection Criteria.............................................................................................................68Participants......................................................................................................................69Participants with normal heari ng and reading skills (NH group).............................69Participants with hearing impairment (HI group).................................................... 70Dyslexic group (RD group)......................................................................................71Matching Variables.........................................................................................................73Procedure................................................................................................................................73Materials.................................................................................................................................74Audiologic Measures....................................................................................................... 74Puretone and speech audiometry.............................................................................. 75Auditory processing tests......................................................................................... 76Middle ear function (tympanometry)....................................................................... 77Literacy (Reading and Spelling)...................................................................................... 78Phonological Processing Skills........................................................................................ 81Standardized Oral Language Tests..................................................................................83Interrater Reliability......................................................................................................... .......85Research Questions and Hypotheses...................................................................................... 86Category I (Group Effect)............................................................................................... 86Category II (Relationships among Measures)................................................................. 86Category III (Regression Question).................................................................................87Treatment of the Data.......................................................................................................... ...874 RESULTS...............................................................................................................................93Descriptive Statistics......................................................................................................... .....94Demographic Data........................................................................................................... 94Audiologic Ability Measures.......................................................................................... 94Oral Language, Phonology, and Reading........................................................................ 95Inferential Statistics......................................................................................................... .......95Question Category I: Group Comparisons of Language, Phonology, and Reading........ 95MANCOVA for oral language measures................................................................. 97MANCOVA for phonology.....................................................................................99MANCOVA for literacy measur es (reading and spelling)..................................... 102Question Category II: Correlations among Measures................................................... 105

PAGE 8

8 Question 1: Phonology and reading....................................................................... 105Question 2: Phonology and hearing loss................................................................ 108Question 3: Phonology a nd oral language.............................................................. 109Question 4: Auditory processi ng, phonology, and reading measures....................109Question 5: Reading and oral language.................................................................. 110Question Category III: Explorat ory Hierarchical Regressions...................................... 111Regressions on word-level reading........................................................................112Regression on passage reading rate and accuracy.................................................. 115Regression on spelling........................................................................................... 116Role of rapid naming in further regression analyses.............................................. 117Summary:...............................................................................................................1185 DISCUSSION.......................................................................................................................155Introduction................................................................................................................... ........155Oral Language Skills Findings............................................................................................. 155Group Comparison........................................................................................................ 156Interrelationships between Phonological Processing and Language Skills................... 158Phonological short-term memory and vocabulary................................................. 158Grammatical knowledge and phonological processing..........................................160Phonological Processing Findings........................................................................................ 162Hearing Loss and Phonological Representation............................................................ 165Phonological Representation and Short-term Memory Skill......................................... 166Intact Rapid Naming...................................................................................................... 167Literacy Findings..................................................................................................................169Group Comparisons....................................................................................................... 169Correlation and Regression Findings............................................................................ 172Summary...............................................................................................................................176Clinical Implications.......................................................................................................... ...179Limitations.................................................................................................................... ........180APPENDIX A INFORMED CONSENT LETTER FO R PARENTS AND GUARDIANS ........................ 182B QUESTIONNAIRE FORM..................................................................................................186C SCORING SHEETS............................................................................................................. 193LIST OF REFERENCES.............................................................................................................196BIOGRAPHICAL SKETCH.......................................................................................................221

PAGE 9

9 LIST OF TABLES Table page 3-1 Matching variables (Grade, Age, Gender, and non-verbal intelligence)........................... 89 3-2 List of tests used......................................................................................................... ........90 3-3 List of research hypotheses................................................................................................91 4-1 Background information and basic auditory skills of individuals in the HI group. ......... 119 4-2 Descriptive statistics for the NH and HI groups on PTA, SRT, and W RS scores...........120 4-3 Mean pure-tone thresholds for all tested frequencies (NH and HI groups only). ............ 121 4-4 Descriptive statistics of auditory processing variables. ................................................... 124 4-5 Descriptive statistics and estimated adjusted m eans of oral language measures (MANCOVA)..................................................................................................................125 4-6 ANOVA tables for univariate ANCOVAs for each oral langu age measure................... 126 4-7 Pairwise comparison results based on adjusted group means.......................................... 126 4-8 Descriptive statistics and estimated adjusted groups m eans for phonological processing measures......................................................................................................... 128 4-9 Summary of three univariate ANCOVAs on the phonological measures from the CTOPP. ......................................................................................................................... ...129 4-10 Summary of post hoc pairwise comparisons of phonologica l m easures (CTOPP subtests)............................................................................................................................130 4-11 Descriptive statistics and estimated adju sted group m eans for literacy measures........... 134 4-12 Summary of univariat e ANCOVAs on the reading and spelling m easures..................... 139 4-13 Summary of post hoc pairwise comparisons of eight reading and spelling m easures..... 140 4-14 Correlations between the degree of hearing loss and phonological m easures, partialing out for age and grade.......................................................................................141 4-15 Partial correlation matrix betw een phonology and reading (HI group). .......................... 141 4-16 Partial correlation matrix betw een phonology and reading (NH group). ........................ 142 4-17 Correlation matrix between better ea rs PTA and phonology, partialing out age (in months). ...........................................................................................................................143

PAGE 10

10 4-18 First-order correlation matrix betwee n oral language and phonology m easures............. 144 4-19 Correlation matrices among auditory processing m easures, and phonology and reading scores, partialing out for age, gr ade, duration of hearing aid use, and nonverbal intelligence...................................................................................................... 145 4-20 Correlation matrix between oral language and reading m easures................................... 146 4-21 Hierarchical regr essions of variables separately predicting untim ed word reading (the Word Identification subtest on the WRMT-R)......................................................... 147 4-22 Hierarchical regr essions of variables separately predicting untim ed nonword reading (the Word Attack subtest on the WRMT-R).................................................................... 148 4-23 Hierarchical regr essions of variables separately predicting tim ed word reading (the Word Efficiency on the TOWRE)....................................................................................149 4-24 Hierarchical regr essions of variables separately predicting tim ed nonword reading (the Word Decoding on the TOWRE).............................................................................150 4-25 Hierarchical regr essions of variables separately predicting tim ed passage reading rate on the GORT-4......................................................................................................... 151 4-26 Hierarchical regr essions of variables separately predicting tim ed passage reading accuracy on the GORT-4.................................................................................................152 4-27 Hierarchical regr essions of variables separately predicting scores on the Spelling on the W RAT........................................................................................................................153 4-28 Summary of hierarchical regressions on tim ed reading measures, controlled for phonological awareness and shor t-term memory scores (s howing the results from Step 3 only)......................................................................................................................154

PAGE 11

11 LIST OF FIGURES Figure page 2-1 Three different areas of phonological processing.............................................................. 61 2-2 Phonological loop.......................................................................................................... ....61 2-3 Word recognition th rough phonological route. .................................................................. 62 2-4 Model of (underspecified ) phonological representation. ................................................... 63 2-5 Two different aspects of phonological representations ..................................................... 64 2-6 Simplified model of visual naming.................................................................................... 65 4-1 Mean PTA thresholds for the left ear............................................................................... 122 4-2 Mean PTA thresholds for the Right ear........................................................................... 122 4-3 Histogram of 19 hearing impaired chil dren by better ear pure-tone average. ................. 123 4-4 Clustered box plot of oral langua ge m easure for the NH and HI groups........................ 127 4-5 Clustered boxplot for phonological awar eness measures for the NH, HI and RD groups.. .............................................................................................................................131 4-6 Clustered boxplot for phonological shor t-term memory for the NH, HI, and RD groups...............................................................................................................................132 4-7 Clustered boxplot for rapid nami ng for the NH, HI, and RD groups. ............................. 133 4-8 Clustered boxplot for untimed word-level reading ( Word Id entification and Word Attack subtests on the WRMT-R)....................................................................................135 4-9 Clustered boxplot for untimed word-level reading ( Word Effic iency and Word Decoding subtests on the WRMT-R)...............................................................................136 4-10 Clustered boxplot for conn ected -text reading fluency ( GORT-4 Rate and Accuracy)... 137 4-11 Clustered boxplot for passage compre hension (WRMT-R) and spelling (WRAT). ....... 138

PAGE 12

12 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CHARACTERISTICS OF PHONOLOGICA L PROCESSING, READING, ORAL LANGUAGE, AND AUDITORY PROCESSING SK ILLS OF CHILDREN WITH MILD-TOMODERATE SENSORINEURAL HEARING LOSS By Jungjun Park August 2008 Chair: Linda J. Lombardino Major: Communication Sciences and Disorders The principle aim of the study was to examine the effect of mild to moderate SNHL on 19 childrens performances on a range of spoke n language, phonological processing skills, and literacy (reading and spelling). Performances of two controls groups (29 normally developing and 30 dyslexic children), matched for age, gr ade, and non-verbal intelligence were also analyzed to investigate the HI groups relative weaknesse s and strengths in phonological processing ability and its relati on to their literacy skills. It is well known that a lack of natural a nd complete linguistic i nput in early childhood contributes to significant language delays even in children with mild to moderate hearing loss. Although they may be helped by hearing aids, thes e children usually remain unable to extract enough auditory information to develop spoken language with the ease and efficiency of a normally hearing child. Due to this, the deve lopment of appropriate phonological processing skills which are necessarily required for the early reading development is a considerable challenge for children with hearing loss. This study was planned with an assumption that literacy development would require the same acquisiti on of phonological processi ng skills, whether a childs hearing is impaired or normal.

PAGE 13

13 A set of MANCOVAs revealed that HI groups performances on the measures of spoken language (receptive and expressive vocabul ary and morphosyntactic knowledge) and phonological processing tasks were significantly delayed when compared to their normal controls. The HI groups phonologi cal processing skills were selectively depressed such that only phonologically-based tasks (phonolog ical awareness and phonologica l short-term memory) were seen significantly lower than the normal cont rols but their rapid naming skills were seen preserved despite hearing loss. Th is finding suggests that auditory perceptual distortion due to SNHL can only affect phonologically-based com ponent and would not have negative impact on extraphonological processing ability (RAN). It was found that the HI groups literacy skills were significantly lower than their normally hearing controls. The necessary role of phonologi cal processing skills was also confirmed by the correlational and regression analyses which s uggest that phonological processing skills are important correlates and predictors of hearing-impaired childrens reading and spelling as well. The strongest predic tor of reading a nd spelling was the Elision subtest on the CTOPP. These findings contradict previous studi es which argued that reading sk ills of hearing-impaired would not necessarily require the support from phonologi cal processing ability. Impressively, RAN measures were not asso ciated with any of untimed reading tasks, but had significant correlations only with timed reading tasks. RAN also additively contributed independent variance to timed reading measures after p honologically-based variables were controlled for. Lastly, assuming that hard-of-hearing individuals, like hearing individuals, will also benefit from direct phonologica l instruction, the findings impl y that effective programs and strategies for teaching hearing-impaired children these skills may be th e key to obtaining higher levels of reading achievement for this population.

PAGE 14

14 CHAPTER 1 INTRODUCTION Background of the Study The purpose of this study is to investigat e the strengths and weaknesses of skills in reading, phonological processing, and oral language of children with congenital mild to moderate sensorineural hearing loss in comparison to t hose of dyslexic and normal readers with intact hearing ability. It is widely acknowledged that the critical requirement of written la nguage skills is the acquisition and development of language and conc omitant cognitive skills at as early an age as possible. Among various tiers of linguistic knowledge, the phonological component is now known as the most important foundation for the deve lopment of early literacy skills. Over the last several decades, an intense research focu s has been placed on re lations between phonological processing capacities and reading performan ces measured by word recognition, reading comprehension, reading fluency, and spelling. Ev idence from both typically developing and atypically developing children demonstrates that the quali ty of a childs phonological representations is important for their subsequent progress in literacy. This relationship has been found across all languages studied, for both persons with typical reading ability (e.g. Bradley & Bryant, 1983; Hoien, Lundberg, Stanovich, and Bjaalidet, 1995; Siok & Fletcher, 2001), and persons with reading disabili ties (e.g. Bradley & Bryant, 1983 ; Bruck, 1992; Landerl, Wimmer, & Frith, 1997; Porpodas, 1999). It is generally accepted that dyslexia is characterized by developmental weaknesses in establishing phonological repres entations of speech sounds. The phonological core deficit theory (Stanovich, 1988) argues that children with dyslexia find it difficult to represent mentally the sound patterns of the words in their language in a detailed and specific way. Research to date has supported the causal connections between

PAGE 15

15 phonological representation a nd reading acquisition. Thus, research supports that the presence of normal phonological processing capac ity is a hallmark characteris tic of good readers while its absence is a consistent charac teristic of poor readers (Hurford, Darrow, Edwards, Howerton, Schauf, & Coffey, 1993; Mann, 1993). Based upon the phonological core deficit hypothesis of literacy development, if childrens phonological representations are adve rsely affected for some reas on, we can hypothesize deficits in their phonological skills will result in defici encies in reading skill. Many students with congenital hearing impairment can have problem s with reading and wri ting because of their difficulty in acquiring and/or manipulating the structural characteristics of phonological component of language. It is assumed that hard -of-hearing children have constrained access to the input of speech sounds of their first language during the early critic al period of language acquisition. Because of the possibl e distortion of the incoming acoustic signal, it is hypothesized that the phonological processing capacity of the children with hearing loss and the emerging literacy-related cognitive skills resulting from it would also be negatively affected (Briscoe, Bishop, & Norbury, 2001; Nittrouer, 1996; and Nittrouer and Burton, 2005). According to Nittrouer and Burton (2005), early auditory experience associated with normal hearing capacity facilitate s the development of language-s pecific perceptual weighting strategies, which are believed to be critical fo r accessing phonetic entiti es of a given language and constructing a language-spec ific phonological structure. In turn, this knowledge of a languages phonemic inventory a nd phonological structure, based on normal auditory perception, will allow for the appropriate development of phonological processing skills such as efficient temporary storage (verbal working memory), le xical retrieval from l ong-term memory, and a variety of meta-phonological skills. In this respect, peripheral h earing loss which is assumed to

PAGE 16

16 result in alteration or distorti on of auditory input, can provide a way of testing the phonological processing deficit hypothesis. That is, can the th eory of phonological core deficit be applied to hearing impaired children, for wh om possible phonological depressi on may be associated with a peripheral auditory deficit? Rationale and Significance of the Study Links between hearing loss, language, and reading perform ance have been widely investigated in children w ith conductive hearing loss caused by otitis media with emission (OME) (Friel-Patti & Finitzo, 1990; Mody, Studdert-Kenne dy, & Brady, 1997). Researchers tried to address the effects of frequent OME on childrens speech, language, and reading abilities. However, findings are inconsistent across studies. Also, ther e is a large body of research investigating oral and written language of children with SNHL, but most of this research focuses on children with severe to profound losses who rely heavily on visual information. Thus, scant research efforts have been directed to studying the contribution of phonological processing skills to literacy development in children with mild to moderate levels of SNHL. To date, only a few studies have investigat ed the effect of potentially impaired phonological skills in children w ith mild-to-moderate SNHL on reading and related cognitive skills (Briscoe et al., 2001; Delage & Tuller, 2007; Gibbs, 2004; Halliday & Bishop, 2005). Furthermore, the results of previ ous research are contrary to ou r expectations based on findings from other populations. That is hard-of-hearing children with depressed phonological processing skills do not show the pervasive difficulties with language and literacy. For example, Briscoe et al. (2001)s study shows that even with significantly depressed phonological skills associated with hearing loss, reading skills are possible to some degree. In Briscoe et al. study, nineteen children with mild and moderate SNHL were merged together to comprise a single experimental

PAGE 17

17 group and their reading and phonology sk ills were compared with those of children with specific language impairment (SLI). However, among the total 19 children in this group, 13 children had mild hearing losses and only three children had moderate SNHL. This skewed distribution of hearing loss levels would not adequately represent the popula tion of children with mild-tomoderate SNHL. As expected, this might be th e reason the subjects reading performance was comparable to that of normally hearing children. Seemingly, due to this limitation, statistica l analyses conducted for group comparisons may not have captured the potential adverse e ffects of phonological depression associated with hearing impairment. Gibbs (2004) also reported a similar pattern. Fifteen children with mild-tomoderate bilateral SNHL were compared to nor mally hearing controls on reading performance. The reading abilities of the child ren with SNHL were indistinguishable from their hearing peers and reading abilities were not si gnificantly associated with th e degree of hearing impairment. However, their phonological skills were inferior to those of the normally hearing controls even though they had a significant nega tive correlation with hearing loss level. Gibbs concluded that reading might be possible even without well developed phonological sk ills. Thus, further research strictly focusing on the literacy and reading-related cognitive skills in children with mild-to-severe SNHL with appropriate methodology is warranted. In these studies, the hearing-impaired childre ns success with reading in spite of depressed phonological processing skills was notable and une xpected in the theore tical framework of phonological core deficit. These unexpected findi ngs raise a critical question regarding the precise role of phonological skills in written language acquisition. That is, according to the phonological processing deficit hypothesis, im paired phonology would lead to compromised literacy ability since phonological capacity is regarded as a n ecessary, although not a sufficient

PAGE 18

18 condition for reading. Moreover, no previous studies compared hard-of-hearing childrens phonological and literacy skills to the same skills in children who have dyslexia. Such a comparison could provide insight in to the cognitive skills of two groups of children with reading disabilities due to di fferent etiologies. In addition, very few studies investigated the effects SNHL has on phonological, language, and reading skills using a wide range of tests. To provide a more precise view of the developmental characteristics of hearing impaired childrens reading and phonological skills, a comprehensive battery of tests was used in this study. That includes: (1) oral language (receptive and expressive vocabulary, grammatical knowle dge), (2) phonological processing (phonological awareness and short-term memo ry, verbal working memory, ra pid naming), (3) reading skills (word recognition and nonword decoding in both timed and untimed manner, paragraph reading accuracy/speed, reading comprehension, and spelling). Finally, auditory processing tasks were also used to determine if hearing impaired child rens reading skill is associated with low-level auditory perceptual processing, a skill needed to perceive spee ch events for brief temporal durations. Study Objectives Objective 1: To provide a com prehensive da ta set on a range of oral language, phonology, auditory processing, and literacy skil ls in hard-of-hearing children. Objective 2: To investigat e the strengths and weakne sses in reading, phonological processing, and oral language of hearing-impaired children and to compare their abilities to those of children who are typical readers and children who have dyslexia, a specific reading impairment. Objective 3: To overcome some methodologica l limitations mentioned earlier by collecting a set of data which well represent the population and by using a better statistical approach for analysis. Objective 4: To address the uni versality of phonological core de ficit theory. The theory is that the contributive and necessary ro le of phonological processing capacity for literacy found in normal hearing children will also be obs erved in children with hearing loss, that

PAGE 19

19 is, the theoretical framework known as the phonol ogical deficit hypothesis predicts that the more severe the initial hearing loss, the more impaired reading skills will be. This issue is of great theoretical importance in unde rstanding associations between impaired phonological processing skills a nd reading achievement in children with hearing loss. Clinically, if this hypo thesis is supported in this study, the results will serve to guide principles for reading programs for hearingimpaired children in school settings. Brief Definition of Terms Research literature in the area of phonological proces sing and phonological awareness entail high ly technical language. The following definitions can be used as a guide for the subsequent discussion of these complex concepts. Phonological representation: Use of arbitrary symbols (oral or writ ten) to represent experience or concepts (e.g., words or graphic symbols like "$"). Phonological processing: The use of phonology or sounds of language to process verbal information in oral or written form in shor tand long-term memory (Wagner & Torgesen, 1987). Components include awareness and codi ng (i.e., coding sounds for storage in memory and retrieval of sounds from memo ry codes) of verbal information only (Cornwall, 1992; Hurford et al., 1993; Torg esen et al., 1990; Vellu tino & Scanlon, 1987a; Wagner & Torgesen, 1987). Phonological coding: "The representation of information about the sound structure of verbal stimuli in memory" (Torgesen et al., 1990, p. 236). Phonological recoding: Translati on from either oral or wri tten representation into a soundbased system to arrive at the meaning of words in the lexicon (stored vocabulary) in longterm memory (Wagner & Torgesen, 1987). Phonological units: Refers to the size of the sound (e.g., phon emes, onset-rimes, syllables, word). Phonetic recoding: Translati on of verbal information into a sound-based system for temporary storage in working memory for processes such as decoding unfamiliar words in fluent reading, or during th e beginning reading processes of blending and segmenting. Phonological awareness: Conscious awareness of the sound segments in language (e.g., syllable, shared rimes, or phonemes) and abil ity to manipulate sound (e.g., move, combine, and delete). Phonemic awareness: Awareness of phonemes, discrete individual sounds that correspond to individual letters. Spector (1995) pointed out that many term s have been used for this ability, including phonemic awareness, phonetic analysis, auditory analysis, phonological reading, phonological processing, and linguist ic awareness. Typically, phonological

PAGE 20

20 awareness is used as a genera l term and phonemic awareness is used to refer specifically to awareness at the phoneme level. Decoding: Translating individual letters and/or groups of letters into sounds to access the pronunciation of a word. Lexical access: Access to intern al dictionary in memory. Retrieval: Accessing coded information from short-term or long-term memory. Phonemes: Individual sounds, smallest unit of sound. Grapheme: Written symbols or letters of the alphabet; arbitrary, abstract, and usually without meaning; the written equivalent of phonemes. Grapheme-to-phoneme correspondence: Linka ges between discrete phonemes and individual letters or graphemes. Onset-rime: Two-part division of words into units th at are smaller than syllables; onset is the first division of a single phoneme or consonant cluster (e.g., /br/ in bright), rime is the last division with multiple phonemes (e.g., /ight/ in bright). Alphabetic understanding: Understanding that letters represent sounds and that whole words have a sound structure c onsisting of individual sounds and patterns of groups of sounds.

PAGE 21

21 CHAPTER 2 LITERATURE REVIEW Introduction The prim ary goal of this chapter is to provide a revi ew of previous stud ies related to the phonological core model of readi ng disabilities is provided. The nature of weakened phonological processi ng and its possible cau sal role to later literacy development is considered through a review of current explanatory models or theories of the interface between reading and phonology. Theo ries reviewed will include: (i) Phonological representation hypothesis and lexical restructuring hypothesis; (ii) double-deficit hypothesis, and (iii) auditory processing deficit hypothesis. Integr ated summary of each theory will be provided in each section. The chapter also reviews studies dealing with language and reading development of hearing impaired children. Before that, a brief theoretical framework of phonological processing is given in the first section. Phonological Core Deficit Theory Developm ental dyslexia (DD) refers to the inab ility to acquire proficient reading skill and is a prevalent learning disability affecting be tween 5 and 15% of child ren in school. Vellutino (1979), one of the earliest studie s in reading disorder, reporte d that children with reading problem have systematic difficulties on tasks with verbal demands, whereas they performed at the similar level with normal readers on non-verbal tasks. Over the last two decades, a large body of c onverging evidence now indicates that dyslexia stems from an underlying deficit in the phonologi cal processing system, s uggesting that deficits in processing the sounds of language explai n a significant proportion of beginning reading problems and correlated problems w ith older readers even though de bates still remain regarding whether a single, phonological core deficit or other cognitive deficits lead to reading failure

PAGE 22

22 (Beitchman & Young, 1997; Mody, 2003; Shay witz, 1998; Snowling, Nation, Moxham, Gallagher & Frith, 1997) or underspecified, poor phonological representations (Elbro, 1996; Fowler, 1991; Hansen and Bowey, 1994; Mets ala, 1997; Snowling, Goulandris, Bowlby, & Howell, 1986; Swan and Goswami, 1997a,b). The strong association between the phonologi cal deficits and dyslexia led Stanovich (1986) to propose that dyslexia should be de fined as a core phonologi cal deficit. In the phonological core-variable difference model of dyslexia (Stanovich & Siegel, 1994), poor phonology is related to poor reading performance, ir respective of IQ. An important advantage of the core phonological deficit defini tion of dyslexia is that it make s sense in terms of what is known about the normal acquisition of reading. That is, it has been known that phonological awareness skills measured in preschool is an excellent predictor of later literacy performance, even after the substantia l effects of IQ are controlled. It is well understood that the ability to reflect upon the sound structure of words at the phonemic level is cr itical to the development of the alphabetic principle that allows children to decode novel words that they have not seen before. There is now converging evidence that the core deficit in read ing disability is at the level of phonological awareness, letter-s ound decoding and limitations of verbal short-term memory (Ehri, Nunes, Willows, Schuster, Yaghoub-Zade h, & Shanahan, 2001; Fletcher, Shaywitz, Shankweiler, Katz, Liberman, Stuebing, Francis, Fowler, & Shaywitz ,1994; Foorman, Francis, Beeler, Winikates, & Fletcher, 1997; Morris, St uebing, Fletcher, Shayw itz, Lyon, Shankweiler, 1998; Scanlon & Vellutino, 1997; Shaywitz et al., 1999; Stanovich, 1988, 1993; Stanovich & Siegel, 1994; Torgesen, Wagner, & Rashotte, 1997; Vellutino, Fletcher Snowling, & Scanlon, 2004). Shaywitz (2003) describes phonological awareness as an in clusive term that includes

PAGE 23

23 all levels of awareness of the s ound structure of words. It also is used to refer to the earliest stages of developing an awareness of the parts of words, such as sensitivity to rhyme or noticing larger parts of words such as syllables (p.144). Letter-sound decoding is the process of converting the written symbols on the page to the smallest unit of speech sounds called phonemes (Shaywitz, 2003). There is also eviden ce that dyslexic children have trouble with long-term verbal learning and the retrieval of phonological information from long-term memory. Word-finding difficulties are often seen clinically and deficiencies in lexical retrieval can be manifested by rapid naming tasks. In a recent summary of what has been learne d about dyslexia in the past four decades, Vellutino et al. (2004) reviewed th e support behind a number of theo ries that have been proposed as the underlying cause of dyslexia. Citing findings from the research litera ture, Vellutino et al. found that there is growing consensus that the most influential cause of difficulties in learning to read is the failure to acquire phonological awareness and skill in alphabetic coding (p.12). More specifically, weak phonological coding has been identified as the central cause of reading disability in most impaired readers (Archer, Gleason, & Vachon, 2003; Ehri et al., 2001; Lyon, Shaywitz, & Shaywitz, 2003; Ramus, Rosen, Dakin, Day, Castellote, White, & Frith, 2003; Vellutino et al., 2004). Components of Phonological Processing Phonological processing is defined as the use of phonologica l inform ation (i.e., the sounds of ones language) in processing written and oral language. Phonological processing is defined as the use of phonological information (i.e., the s ounds of ones language) in processing written and oral language (Wagner & To rgesen, 1987). It sometimes has b een used interchangeably with phonological awareness, but th ese two concepts are best seen as distinct from one another. Three major components of phonological processing defic its have been identif ied: (a) phonological

PAGE 24

24 awareness, (b) phonological recoding in lexical access, and/or (c) phonetic recoding to maintain information in working memory (Wagner & Torgesen, 1987). More specifically, rese arch with school-age children has identified three interrelated phonological processing abilities that are important for read ing and writing: phonological awareness, phonological memory, and efficiency of phonological access to lexical storage. How the various phonological processing abil ities are related to each other and what roles they play in literacy development are issues of considerab le theoretical and practical importance. Phonological awareness (PA) refers to one s ability to detect or manipulate the sounds in his or her oral languag e (for review, see Anthony & Fran cis, 2005). PA encompasses phoneme awareness, the ability to manipulate individua l sounds (phonemes) in words, and rudimentary phonological awareness skills, such as judging whether two words rhyme. Phonological memory (PM) refers to the coding of information in a sound-based representation system for temporary storage. PM is utilized during all cognitive tasks that involve processing sound information. Individuals PM capacity is often operationalized by audito ry span tasks, like digit span. Rapid naming (RAN) refers to the efficiency of re trieving phonological codes fro m memory. Individual differences in efficiency of retrieving phonologically stored information from memory are typically measured by performance on rapid autonomic naming tasks in which individuals verbally identify common objects, letters, or numbers as quickly as possible. Phonological Awareness Stanovich viewed phonological awareness, as conscious access to the phonem ic level of the speech stream and some ability to cognitively manipulate representations at this level (1986, p 362). Basically, phonological awar eness is a multilevel skill of breaking down words into smaller units (Hoien, Lundberg, Stanovich, & Bjaa lid, 1995). It refers to an individuals awareness of or sensitivity to the sound struct ure, or phonological struct ure, or segments of

PAGE 25

25 differing level in spoken words. Torgesen (1997) defines phonological awareness as the ability to notice, think about, or mani pulate the sounds in language. Dimensions of phonological awareness Based on recent phonological theories focusing on the hierarchical sound structure of a word, phonological awareness can be described in term s of different phonol ogical tiers such as syllable, onset-rime, and phonemes. Stanovich (1994) viewed phonologica l awareness as the ability to deal explicitly and segmentally with sound units ranging from syllable, onset-rime, and phonemes. Gillon (2004:5-9) subdivided it into th ree subtypes; syllable awareness, onset-rime awareness, and phonemic awareness (Figure 2-1). A variety of measures have been used to assess individuals know ledge of these three differing levels such as auditory discri mination, blending, segmenting, deletion, isolation, rhyming, substitution, sound categorization, tappi ng, reversing order of sounds, and word to word matching. Especially, regarding phoneme-level of awareness, Adams (1990) describes five different types of manipulative sk ill in terms of abilities: (i) To do oddity tasks (comparing and contrasting the sounds of words for rhyme and alliteration); (ii) to hear rhymes and alliteration as measured by knowledge of nursery rhymes; (iii) to blend and split (segment) syllables; (iv) to perform phonemic segmentation (such as counting out the number of phonemes in a word), and (v) to perform phoneme manipula tion tasks (such as adding, dele ting a particular phoneme and regenerating a word from the remainder). Similarly, research has shown that phonologica l awareness dimensions can be validly and reliably measure through a variety of tasks (Y opp, 1988). According to Yopp, the dimensions of phonological awareness are represen ted by a range of difficulty. Fr om easiest to hardest the range of difficulty is as follows: (a) rhyme, (b) auditory discrimination, (c) phoneme blending,

PAGE 26

26 (d) word-to-word matching, (e) sound is olation, (f) phoneme counting, (g) phoneme segmentation, and (h) phoneme deletion. Factors related to order of difficulty A num ber of properties of phonological units have been found to affect the degree of difficulty of phonological awareness tasks including: (a) the positi on of the unit in words (i.e., first, middle, or last); (b) degree of abstraction; (c) characteristics of tasks used; and (d) size of sound unit and related acoustic features. Position: Research points to the differential difficulty for initial, medial, and final positions, with initial and final positions be ing easier than middle (Byrne and FieldingBarnsley, 1989). Presence of semantic content: Degree of se mantic abstraction also affects difficulty. Different from phonemes or rhymes, real word s are less abstract entity due to their semantic entity. Thus, words are recognized and manipulated naturally without less instruction. Instead, phonemes are: (a) the sm allest phonological unit, (b) not acoustically pure (not easily isolated), (c ) independent of meaning, and (d ) abstract and arbitrary. Task characteristics: The characteristics of tasks involved decide difficulty as well. Adams (1990) indicated that most young chil dren can rhyme but not delete. Size of unit: Differential difficulty among units of varying size can be explained by their acoustic features. Spector (1995) conjectured that we do not hear discrete pure phonemes because they overlap in speech chain; rather, we hear in syllables. Therefore, tasks that require sensitivity to phonemes may be more complex and necessarily more difficult than those that require manipulation of syllables. Similarly, syllable segmentation is easier and often develops without inst ruction in contrast to phone me segmentation. Liberman & Shankweiler (1985) reported th at in groups of four-year-old children, none could segment by phoneme whereas about 50% could segment by syllables Processing memory: Memory capacity also contribute to difficulty related to phonological awareness since each phonological awaren ess task requires a certain amount of memory capacity to temporarily hold material fo r later processing. For example, phonemic awareness tasks can be divided into diffe rent categories depending on the memory processes and the number of operations requi red. For example, when asked what sounds are heard in fish (segmentation), only one ope ration of pulling apart sound is needed (i.e., /f/, /i/, and /sh/) while holding the word in s hort-term memory. In contrast, when asked to delete the first sound from fish (deletion), a child should (1) segment the sounds, (2) identify the beginning sound, and (3) crea ting a new chain of phonemes using the remaining sounds filling the memory slot previously occupied by the omitted sound.

PAGE 27

27 Phonological Memory Phonological m emory refers to recoding visually (letters) or orally (speech) presented words into a temporary phonological buffer to mainta in efficiently for subsequent verbal tasks such as rote repetition of words/nonwords or co nscious manipulation of parts of a word. This process is called verbal shortterm storage or temporary rete ntion of phonological representation. Baddeley and his colleague figured out human memo ry model which best f its various cognitive behaviors. They extended the previo us short-term memory, a passive storage of verbal input, to a more active working space (i.e., working memory ) serving lots of c ognitive tasks such as phonological awareness, sentence an alysis, retrieval of lexical information from a LTM, or decoding of words. Old research on short-term memory (S TM) has focused on temporary holding of information, rather than on the processes or tr ansformation of information in general cognition. However, for tasks which require more than simple retention of information, a different memory module is needed. For Baddeley a nd Hitch (1974), the working memo ry is a workspace and its capacity can be divided between processing and storage. The storage portion was referred to as the phonemic buffer or phonological loop and the processing portion was referred to as central executive. Tasks such as verbal rehe arsal which extracts the information from the phonemic buffer, segmenting, blending, reordering, and substitution were all regarded as the functions attributed to the central executive, a processing system. Fi gure 2-2 is a simplified display of Baddeley (1986)s tripartite framew ork, which had consolidated the position of the working memory model in cognitive psychology. Th e frame consists of two modality-specific, limited-capacity storage systems called slave systems and a limited-capacity general processing system called central executive Repetition of words, nonwords, or sentence and digit span tasks (i.e., serial recall of digits) are used to measure the efficiency of phonological recoding in

PAGE 28

28 working memory (i.e., they measure phonologica l short-term memory). Among these, nonword repetition is seen as the best measure of short term memory since semantic cues would be available to facilitate rote repetition of nonsense words. Rapid Naming A great deal of research over the past 30 years on childrens phonol ogical awareness has brought us virtually everything we need to know about identifying children who lack phonological awareness and teach ing them to de velop this knowledge. However, another capacity was shown significantly compromised in dyslexic children; slow lexical access (rapid automatized naming, RAN in short). Wagner and Torgesen (1987) used a specific terminology for this process: phonological recoding in lexical access It is defined as getting from a written word to its lexical referent by converting the graphemes into sound-based representation system. In general, coding process involves translating stimuli from one form to another (e.g., from letters to auditory). For effi cient word recognition, phonological coding ability should be automatized or fluent. Since Wagner and Torgesen (1987)s study, a ne w line of research have suggested that automatized lexical access and re trieval may significantly affect ease of reading acquisition. Rapid naming and lexicality tests are two tasks commonly used to measure ability to fluently code letters into phonological representations. OConnor, Jenkins, Leicester, & Slocum (1993) reported that significant differe nce on reading and spelling m easures between lowand highskilled readers could be expl ained by differences in rapid letter naming. Cornwall (1992) similarly indicated that students that had ra pid rates of letter nami ng did better in word identification and passage reading speed and acc uracy than those with lower rates of rapid naming. Wren (2005) indicated that impaired ski lls in RAN is not easy to identify at younger age, and improving RAN and visual processing speed is considerably more difficult than helping

PAGE 29

29 children develop PA. To the theorists, it ha s been an important question whether the phonological awareness and the impaired RAN repres ent the so-called doubledeficits or whether they are two manifestations of the same underlying disorder (Tijms, 2004). Unfortunately, less is known about cognitive processing model of RAN. First of all, visual stimuli (letter sequence or pictures) s hould be converted into a sound representation to get access to permanent lexi cal item (retrieval). Acco rding to Ramus (2001b), after visual or auditory signal is processed to form sub-lexical phonological representation, with which no lexical, semantic information is yet linke d, retrieval system will start to find a lexical target of which the phonological form is well matching with the sublexical phonetic entity. After retrieval is successfully executed, semantic component of the target item will be activated or pulled out to temporary working memory space a nd used for subsequent verbal tasks such as comprehension. Cognitive Model of Phonological Processing Figure 2-3 p rovides is a cognitive model of phonological processing and word recognition, which is a refined version of Levelt (1989)s model of speech production and Ramus (2001b) model of lexical access. This m odel integrated orthographical lexical access (reading), visual lexical access (object recogni tion), and auditory lexical access (speech-based word recognition) to explain reading. Each box stands for distinct levels of repr esentations and arrows stand for a specific conversion, or translation between different modules. Mental lexicon is divided into three parts (lexical meaning, phonological form, and orthographi cal form). When children perceive ambient speech signal, it is first encoded as non-specific manner, that is, as non-speech signal (arrow 1). In the model, this non-specific si gnal is called acoustic representation. At a later stage, this must be encoded as a speech signal (arrow 2). Si nce no lexical entry is accessed yet, the model

PAGE 30

30 uses a term of sub-lexical phonol ogical representation at this level. This sub-lexical representation has also to be converted into a phonological representation for a specific lexical item. So, the arrow 3 between sub-lexi cal and phonological form represents auditory word recognition Auditory word recognition requires the finding of lexical item whose phonological form matches the sub-lexical sound representa tion. As indicated by Ramus (2001a:201), the phonological lexicon is a permanent, long-term st orage for word forms (LTM), whereas the sublexical phonological represen tation is a short-term storage for whatever can be represented in a phonological manner, that is, words or non-sense wo rds. In this regard, Ramus model can be extended into a new framework, in which the lo ci of the phonological memory (STM/LTM) and awareness can also be established. When a word is very familiar to a listener (i.e., cat ), a procedure called whole-word recognition will happen without any rehearsal in work ing memory area, that is, the child may not have to manipulate or rehear se the input sequence of sounds to figure out the ultimate target in lexicon and will instantly map the s ub-lexical sound form with the corresponding phonological form (arrow 3). However, when the incoming sub-lexical phonological representation is new or unfamiliar ( e.g., peruse), the child needs to analyze it into smaller parts (syllable, intra-syllable, or phonemes) and use this parsing information ( e.g. per + use ) for lexical retrieval by manipulating or processing them. Since these manipul ation tasks are quite different from rote repetition usually done in ST M, we need working or processing memory, as a functionally separate module. Working memory is a locus where retention and phonological manipulation (awareness tasks) occur. The input from the STM in sub-lexical component can be sent to WM for further manipulation or phonologi cal rehearsal need to process unfamiliar words or non-words (arrow 4). WM also receives its input from the le xical representation. From the

PAGE 31

31 mental lexicon, phonological form of a lexical item can be sent to WM for further awareness tasks (arrow 5), but WM is not a place for pe rmanent storage like lexical storage. For phonological awareness tasks, WM consists of its own retention resource needed for temporary processing (storage capacity) and manipulation func tion (processing capacity). In this respect, WM is displayed as separated from sub-lexical and lexicon. Of importa nce is the fact that phonological processing components (sub-lexicon, WM, phonological awaren ess skill) are the main gate which incoming auditory signal mu st enter for successful lexical access and acquisition. Theories on Impaired Phonology One of the robust findings in literacy developmen t research is that ch ildren with im paired reading skill show concurrent weakness in meta-phonological skills (phonological awareness tasks). Indeed, a childs PA knowledge has been described as th e best single predictor of reading performance (Lieberman, Shankweiler, & Liberman, 1989). Because of this, it was assumed that weak phonological awareness is a causal factor in reading and spelling difficulties. Logically, however, a strong predictive power of a relevant factor does not necessarily imply its causal relation to a resulting co ndition. Rather, it is still quite unclear in what way aspects of weakened phonological awareness skills are causally linked to reading impairment. Of equal significance is the fact that deficits in other tw o areas of phonological processing skills have been also identified in children with reading disorder: (1) in efficient activation of long-term phonological coding in le xicon for rapid naming tasks, and (2) impaired short-term memory skills necessary for immediate or rote repetition of sequences of words or sentences. Considering these general impairments in phonologi cal components, research pointed to a more central phonological module as the possible source of dyslexia (B rady, 1997; Mody, 2003; Snowling, 2001; Swan & Goswami, 1997a).

PAGE 32

32 Thus, the precise nature of the phonological proc essing difficulties in reading disorder is a central topic of current psycholinguistic rese arch on dyslexia (Elbro & Jensen, 2005). Various hypotheses have been suggested about the nature and origin of phonologica l processing difficulty in dyslexia. How these problems relate to each other (interrelation), the extent to which weakened phonology is a cause of reading difficulty (causality) continue to be explored, but the exact mechanisms by which dyslexics weakened phonological processor impact each relevant skill of oral or written language skills such as impaired decoding or spelling, delayed word recognition, and reading comprehens ion is not clearly documented. In this section, three theoretical approaches will be shortly reviewed regarding the possible causes of impaired phonological skills in dyslexic population: phonological representation deficit, lexical underspecification hypothesis, double-deficit hypothe sis, and auditory perceptual deficit theory. Phonological Representation Hypothesis One suggestion is that the dys lexic readers phonological repr esentations of lexical item s may be less well specified than normal children (Katz, 1986; Fowler, 1991; Elbro, 1996; Foy & Mann, 2001; Griffith & Snowling, 2002). Problems rela ted to establishing complete, full, clear, or precise phonological representations in childrens speech-based coding system or long-term lexical memory have frequently be en mentioned as a possible cause of diverse phonological difficulties of dyslexia. This hypothesis derives from theories of th e development of spoken word recognition and production. When children first begin to acquire le xical items during infancy, each word is coded in terms of certain semantic and phonological features. For example, for them, Daddy may refer to a person of a certain sex/size and it w ould appear different from doggy. These features

PAGE 33

33 will be more fine-grained, specified, and augm ented over time. So, children will soon understand the difference between Daddy and doggy or Debbie. Phonological distinctness refers to the magnitude of the difference between a representation and its neighbors and it is an aspect of the static quality of phonological material. This hypothesis states that a lack of distinct ness and/or segmental specificity in dyslexic childrens developing phonologica l representations supporting s poken word recognition and production is causally linked to their impair ed phonological processing skills (Goswami, 2000; Snowling et al., 1986, Elbro, 1996). Thus, it should be distinguished from dynamic phonological processes such as conscious manipulation, ve rbal rehearsal, phonol ogical retrieval, and articulation, all using phonetic se gments already represented. For example, due to the poor phonological en coding skill, only parts of the phonetic material of the input can be st ored in lexicon (e.g. sub for subway or croco for crocodile ), or relatively full but not complete representati on can be provided (e.g. cro?dile), where the question mark indicates that any unspecified se gment can be inserted which fits with the phonotactic rules of English (e.g. crowdile, cropodile ). To sum, inaccurate, underspecified, indistinct, or low-quality phonological representation of in coming sounds may hinder higher level of phonological proce ssing (Snowling, 2001; Swan & Goswami, 1997b; Tijms, 2004). Lexical restructuring theory A few evidence for weakened phonological repres entation com e from a theory of lexical learning strategy. Metsala and Walley (1998) and ot her studies suggested a theory of lexical maturation, lexical reconstructio n theory (Fowler, 1991; Walle y, 1993). This theory suggests that childrens phonological represen tation of words is re-presented a number of times in terms of different aspects, that is, from whole words to syllables, and phonemes.

PAGE 34

34 It is interesting to note that childrens word recognition strate gies change with the increase of vocabulary; as vocabulary grows, initial ho listic representations are gradually re-structured, and ultimately, phonemes. In this process, freque nt or familiar or early-acquired words will be encountered many times and so will experience mo re re-structuring than less frequent words. Also, children should have more experience for phonologically ambiguous words, that is, words with many similar-sounding neighbors than words with few phonological neighbors. Another important aspect of this theory with regard to phonological under-specification is that the phoneme would not be an integral aspe ct until certain period of language growth (Eimas, 1974), but rather it emerges as a representati on unit via continuing spoken language experience as children experience further lexical restructur ing processes. So, the degree to which segmental (i.e., phoneme-mediated) representation has taken place will be in turn thought to determine childrens abilities in phonological awareness, which is essential for phoneme-based decoding process and resulting word-level reading. It is assumed that each childs lexical restructuring would be different in terms of rate and accu racy. Thus, based on a huge body of research on phonological awareness deficit in dy slexia, it is hypothesized that dyslexic children will show delayed or slow developing lexical restructuring. Similarly, Snowling et al. (1986) and Mets ala (1997) provided a few evidences. The former study reported that, when appropriate le xical contexts were pr ovided, dyslexic children were better at recognizing phonol ogically ambiguous words, suggesting a more holistic lexical presentation. The latter study used a speech gating task, in which small segments from onset of words are presented via headphones (e.g., /f/, /fu/, or /fud/ for fudge ). Dyslexic children needed more acoustic information than age-matched cont rols to recognize words when the target word had few similar-sounding neighbors, suggesting that their phonologi cal representation is not

PAGE 35

35 totally decomposed into phoneme s when compared to the cont rols. This was interpreted by Metsala (1997) to evidence a delayed segmen t-based specification of such words. Application of representation hypothesis The explanatory power of the phonological re presentation hypothesis is pronounced when it com es to weaknesses of other phonological tasks. First of all, poor phonological representation would easily explain dyslexic childrens typical problems in phonological awareness. Since the child only knows that the sound af ter initial stis c onsonant, but not which, he or she would not efficiently delete [r] and say sting when asked to say string without saying [r]. Or the child would not easily say ring by deleting stwhen asked to say string without saying [st] since he or she is now aware of th e actual consonant after st(Elbro & Jensen, 2005). In general, phonemic manipulation tasks (segmenting, blending, etc.) would be very ha rd if incoming segments are underspecified or inco rrectly specified. Second, the distinctness hypothesis (or representation hypothesis) may also account for the picture naming speed deficits, a strong predictiv e factor of reading disorder. For successful lexical access, it is easier to ge t access to a phonological representa tion that is well specified and clearly separated from its neighbors than to unstable or underspecified representations (e.g., Katz, 1986, Bowers & Wolf, 1993). In this way, the representation hypothesis may explain the naming speed deficit. Third, low phonological distinctness may also be linked to the delayed decoding capacity of dyslexic children. Decoding involves the following steps and each step has its own requirement for successful decoding.

PAGE 36

36 Step 1. Consecutive grapheme-to-phoneme conversion : each letter should be converted into its corresponding phoneme (i.e., p [p], r [r], i [i], n [n], t [t]). Step 2. Representation accuracy (distinct c oding): each sound linked to a certain orthographical pattern should be accurate ly coded in the decoders phonological representation in a fixed order. Step 3. Short term memory: every phoneme obtai ned from step 2 should be held in a phonological buffer temporarily, waiting to be blended to produce a whole word print. Step 4. Blending (phonological manipulation): each phoneme will be ultimately put together to form a final representation of a word. If any of these procedures are impaired, decoding problem can occur. Especially, efficient short term retention and blending process would be totally dependent upon the quality of phonological representation of each phoneme converted from input letters. It is expected that step 3 and step 4 will be damaged by underspeci fied input. Likewise, all phonology-based tasks utilizing underspecified input re presentation will be deteriorated in a differing degree. These chain-like negative effects of inaccurate phono logical representation on related phonological tasks are summarized in the Figure 2-4. A challenging observation Based upon the fact that dyslexic children s low perform ances on phonological processing tasks are related to processing incoming sound information, Goswami (2000) indicated that the input-based representation is the major obstacle for further literacy development. So, low scores on input measures of speech perception should di fferentiate dyslexic children from controls; Dyslexic children should find it more difficult to discriminate between different phonemes in speech (phoneme perception), they should find it difficult to recognize spoken words (auditory lexical decision), and they need more phonological information or clearly articulated input for accurate spoken words (speech gating).

PAGE 37

37 However, lots of studies found that such speech recognition-related skills do not consistently characterize individu al dyslexic children even tho ugh all these difficulties have been reported in group or population studies (Manis Doi, & Bhadha, 1997; Mody, Studdert-Kennedy & Brady, 1997). Instead, one consensus has been documented that dysle xic children showed significantly impaired scor es on tasks tapping their production skills such as word finding difficulties (rapid naming), rote repetition of non-words (STM for nonsense word repetition), and less distinctness of vowel production (See Figure 2-5). This subtle observation raises some challenge s to the theory of phonological representation deficit. Is dyslexia an inputprocessing deficit (representation deficit)? Or, are other factors necessary for phonological production such as in sufficient verbal memory resource or phonological awareness capacity asso ciated with dyslexia (post-i nput processing deficit)? In spite of the above observation, Goswami (2000) ar gued that it would be logically possible that our current measures are failing to tap the main representation deficits causing underspecified representation, or that compensation for early repr esentation deficits have already occurred in some children after remediation. Even though this theory is right, the fundamental difficulties measuring childrens inner phonological representation in their mental system would make experimental design hard. According to Elbro (1998), studies of the quali ty of phonological represen tations are scarce. In addition, with a good deal of possible evidence for re presentational or auditory perceptual deficit hypothesis, the fine-grained explanation of th e relationship between underspecified phonological representation and associated phonological tasks (phonological awareness, working memory, lexical retrieval) still remains to be resolved to reveal the exact mech anism responsible for the apparent impairment seen at th e production level as indicated.

PAGE 38

38 A critical summary There is now strong and highly convergent evidence in suppor t of weak phonological coding as an underlying cause of dyslexia. Dysl exic childrens im paired phonology-based skills such as phonological awareness an d alphabetic decoding are believ ed to be due to the weak coding skill. Other problems such as slow lexical storing and/or lexical ac cess, impairment seen in short-term memory tasks are also viewed as stemming from the phonological underspecification (Vellutino et al ., 2004). In turn, difficulties in wo rd storing and retrieval can impair the beginning readers ability to establ ish a stable connection between the sublexical phonological codes constructed from oral or grap hical stimuli and permanent phonological codes in the lexicon, which will ultimately impede the r eaders efficient word id entification. Fluency in word identification is a critically important pr erequisite for adequate reading comprehension (Perfetti, 1985). Compromised short-term phonologi cal memory can also impair the readers decoding capacity, for which each phoneme converte d from letters needs to be blended in an orderly manner. Thus, phonologica l coding weakness was hypothesi zed to be a core deficit which could cause a collection of phonology-based difficulties observed in dyslexics such as word identification (retrieval), phonological aw areness, decoding of letter sequences, rapid naming, vocabulary learning, and nonword repetition. This hypothesis now raises an essential ques tion of why dyslexia is associated with underspecified phonological repr esentation. Very little work has been done on the possible causes of poor representation. Metsala (1999)s we ll-known lexical restruct uring theory is in line with the phonological representation hypothesis. In this theory, th e unit of phonological representation gets smaller star ting from whole utterance to phrase, words. For some unknown reason, this lexical restructuring appears to take place at a lower rate in dyslexic children than others. But, this hypothesis also does not r eadily explain why dysle xics weak phonological

PAGE 39

39 awareness is sustained after thei r vocabulary grows at a near nor mal rate and the theory itself does not provide the root cause of delayed lexical restructuring in dyslexic children. Double-Deficit Hypothesis Deficits other than phonological ones characterize individuals with dyslexia. Im paired performance on rapid serial naming tasks distinguis hes individuals with dyslexia from those with other developmental disorders, like attention-deficit disorder (Felton, Wood, Brown, Campbell & Harter, 1987; Felton & Wood, 1992). Research suggests that for most children, there are two main aspects that drive the development of fluent decoding skills (Lovett, Steinbach, & Frijters, 200 0). First, children should have good meta-phonological awareness skill to identify or manipulate sounds within speech. Secondly, children should be able to proc ess visual or orthographical information very rapidly to be a good reader. This skill is a pa rt of rapid automatic naming (Torgesen, Wagner, Rashotte, Burgess, & Hecht, 1997) and it is report ed that children with slow retrieval time for pictures or objects have similar delayed nami ng with letters or prin ted words. This slow processing of visual information to activate lexical storage will put some students at a disadvantage when it comes to reading (Lovett et al., 2000). RAN as a phonological processing skill Many research studies have dem onstrated that RAN makes a contribution to reading that is independent of the contribution of other predic tors of reading ability such as phonological awareness and memory (Bowers, 1989; Bowers, Steffy, & Tate, 1988). Nonetheless, RAN has often been placed within the phonological processing domain, along with phonological awareness (both synthesis and analysis) and verbal working memory (Wagner et al., 1994; Torgesen, Wagner, Rashotte, Burgess, and Hech t, 1997). This idea is based on Denckla and Rudel (1974)s early suggestion that RAN is basically a phonol ogical processing skill. Those

PAGE 40

40 who believe that RAN is a component of phonological processing, alongside phonological awareness and memory, define R AN as the "efficiency of phonologica l code retrieval" (Wagner, Torgesen, Laughon, Sommons, & Rashotte, 1993; Ve llutino, Scanlon, Sipay, Sm all, Pratt, Chen, & Denckla, 1996). Torgesen and his colleagues re garded naming speed into a part of the phonological family (Torgesen et al, 1997b). Researchers who conceptualized naming speed as a phonological process suggested that naming speed, like any other linguistic tasks (e.g., expressive vocabulary), involves accessing and retrieving a phonol ogical code stored in longterm memory. One of the reasons for this approach is associated with sufficient amount of correlation between serial naming and performance on other phonological processing tasks and based on that, it is argued that se rial naming tasks should be incl uded as part of assessment of childrens phonological pr ocessing abilities. However, others have argued that defic its in visual naming speed and phonological processing are distinct and dissociable aspe cts (Bowers & Wolf, 1993; Wolf & Bowers, 1999). For them, RAN is not a subsidiary component of phonological processing and the evidence based on correlations between speed naming and other phonological skills is insufficient reason to categorize and subsume naming speed under phonology. This opinion that RAN represents functions separate from the phonological proces sing domain stems from the facts that: (1) RAN consistently makes a unique contribution to reading beyond phonologi cal awareness and memory; (2) some group of poor readers show a sp ecific clinical characte ristics (i.e., adequate decoding skill, but slow reading speed). Three subtypes Im paired RAN skills, about which there is more controversy, concerns deficiencies in visual naming speed, that is, impairments in ra pidly accessing and retrieving names for visually presented symbols (letters, numbe rs, pictures of objects, or an imals) even though the names are

PAGE 41

41 quite well known to the child. Wolf and Bowers (1999) proposed a neurocognitive model of reading disability, the Double Deficit Hypothesis (DDH), along at least 2 dimensions of dyslexia subtypes with unique causes of reading failure According to the DDH, these subtypes are classified according to the pr esence or absence of phonological a nd naming-speed deficits, where one subtype exhibits both deficits. Readers within the phonological deficit subtype are characterized by marked difficulty decoding words and exhibit little phoneme awareness of the sound structure of words. Such individuals have special difficulties wi th pseudo-word reading. They are typically identified by significantly low scores on te sts requiring blending phonemes into words or pronouncing parts of words by removing one or more phonemes (phoneme elision tasks). Readers within the second, naming-speed-deficit subtype are markedly slower at serially naming a sequence of visually presented stimuli, such as letters or numbers. This deficit is believed to encompass problems with the automatic or rapid word retrieval necessary for the development of fluent readi ng and ultimately comprehension. Readers with problems on both phonological and naming speed tasks (i.e., more than one standard deviation below average) fall with in a double-deficit subt ype. These individuals regularly represent the most seve re cases of reading failure. According to the DDH, as compared with phonological processing ability, rapid serial naming speed contributes unique variance to read ing rate and fluency measures (Bowers, 1995), and is a stronger predictor of r eading ability in more orthographi cally transparent languages like German (Wimmer, 1993) and Dutch. These findi ngs suggest a unique role for the processes underlying the rapid recognition and retrieval of visually presented linguistic stimuli, and that they may not be subsumed under the phonol ogical core deficits of dyslexia. Differentiation from phonology Intr insic cognitive complexity of RAN: Th e rationale for emphasizing the difference between naming speed and phonology is based on the complexity of cognitive procedure involved in them. Wolf, Bowers, and Biddle (2000), using a model of simple letter naming task, argued that serial naming and its internal complexity go beyond phonological processes. For

PAGE 42

42 them, visual naming represents a demanding arra y of attentional, perceptual, conceptual, memory, lexical, and articulatory processes and all of these components places heavy emphasis on precise time requirements. The following is a simplified version of the model they used to illustrate the complex processes involved in visual naming (Figure 2-6) Step 1 (Activation of attention system): Naming task requires the participants to focus on the task with high level of attention. Step 2 (Visual processes): Based on adequate level of attention, the letters visual information will be processed and this will al low for identification process by integrating information about the present stimulus (letter B) with known mental, visual representation. Step 3 (Integration processes): After Step 2, lexical information will be integrated with accumulated visual information. Step 4 (Lexical access/retrieval): Lexical information, esp ecially the targ et letters phonological code, will be accessed and retrieved. Step 5 (Production): Motor planning and ar ticulation will translate this phonological information into an articulated name. This model exemplifies both the importance of access to the phonological code in naming and the fact that phonological processes repres ent only one subset of the multiple processes involved in naming. Also, the authors emphasized the importance of speed requirement needed for fast naming and this is marked as PSR, whic h is required in every step of whole model. Different from single letter rec ognition, serial letter naming add to this model the extra demands of rapidity, maintenance of a ttention, and serial processing. In sum, the models inherent complexity, the extent of processing speed demands, and the addition of ra pid rate and seriation make naming-speed task a quite different cognitive task from phonology (Wolf, Bowers, & Biddle, 2000:393). Other reasons for differentiation: In addition to the cognitive difference, there are other evidence supporting the differentiation betw een naming speed and phonology. First, the

PAGE 43

43 correlation found between phonological measures and RAN tasks are modest. There is variability in this finding; it is reported that correlation is higher in younger readers and it gets weaker in relatively older readers since naming speed appr oaches somewhat automatic rates in average readers between Grade 1 and 2 (Biddle, 1996). Secondly, there are different patterns of re lationships with reading subskills that characterize naming speed and phonological variables. Bower and her colleagues found that phonological tasks strongly predic ted word/nonword decoding, but not word and text reading speed. Rather, naming speed predicted latency (r esponse time) of word identification and speed of text reading (Bowers, 1993; Bowers & Sw anson, 1991). In Bower (1995), only naming speed contributed to speed on reading measures. After finding similar results, Cornwall concluded that naming speed may represent unique aspects of the reading process as opposed to an overall phonological ability (1992:537). Given this ev idence, to subsume naming speed only under phonology will obscure its separate role in predicting reading disorder. Linkage between RAN and reading process The relation between phonologi cal processing skills (awa rene ss, memory) is easily understandable. Specifically, learning to d ecode through grapheme -to-phoneme conversion requires an acute awareness of and an ability to analyze or blend the sounds within words. For this to be successful, normally developing memory skill must be associated with age-appropriate decoding skills. In contrast, no su ch straightforward conceptualiza tion exists to explain how the processes underlying naming speed affect word identification and nonword decoding. Wolf et al. (2000) proposed two speculative hypotheses on the possible roles of orthographic processing. Orthographic pattern recognition: Slow na ming speed can prevent the appropriate amalgamation of the connections between phoneme s and orthographic patterns at sublexical level. When a reader does not efficiently merge phonemes decoded from each letter due to slow access to phonological codes, it will surely reduce the effici ency of word recognition.

PAGE 44

44 Second route involves the importance of the le arned association between letters in the development of orthographic representation. Th at is, if a beginning reader is slow in identifying individual letters (as indexed by RAN) then single letters in a word will not be activated in sufficiently close temporal proximity to allow the child to become sensitive to the knowledge about orthogra phic (letter) pattern s (Bowers, Golden, Kennedy, and Young, 1994). Similarly, Bowers and Wolf (1993) and Wo lf & Bowers, 1999) suggested that deficits in visual naming speed disrupt reading acquisition by inhibiting growth in the connections between phonemic and orthographic patterns at word and subword levels of representation during word identification learni ng. The possible delay and ineffici ency in establishing this relationship is believed to reduce the qua lity of orthographic codes in memory. Not all scholars seems to accept the view that re ading-related cognitive deficits are caused exclusively or primarily by phonol ogical limitations. Double deficit theory suggests that naming speed deficits are caused by disruption of a p recise timing mechanism that may influence temporal association of visual and phonological counterparts, thereby impairing the childs ability to detect and repres ent orthographic pattern. This compromised connection between visual and phonological pattern of a word cannot be explaine d only by indistinct phonological representation. As discussed above, more general cognitive problem such as visual pattern matching may be involved in impaired RAN Current research may not be able to establish a unifying framework with regard to the differences in the psycholinguistic mechanism and underlying causes between these two components. By positing two separate deficits (DDH) cognitive pathways (visual vs. decodingbased), the single-deficit appro ach (phonological representation theory) may not be the ultimate answer. It may have to be posited that there ar e multiple causes of reading disorder and there are other cognitive areas that cross many domains including visual, auditory, and, possibly, motor processing. All of these functions need to be developed simultaneously with a normal phonology for successful literacy.

PAGE 45

45 Auditory Perceptual Deficit Hypothesis Researchers on the opposite side suggested th at phonological deficit should be understood in the con text of more general sensory or learning impairments (Nicolson and Fawcett, 1990; Tallal, Miller, & Fitch, 1993). This auditory perceptual theory has been proposed by Tallal and her colleagues (Tallal, 1980; Tallal et al., 1993). They have argued that dyslexic children have particular difficulties in processing rapidly chan ging or transient acoustic events, and that the ability to process rapid successive informati on is fundamental to setting up the phonological system. It is posited that children with language -learning disabilities and dyslexia process sounds more slowly than the average children and this may diminish their ability to distinguish phonemes. Evidence for this theory comes from studies that measure how mu ch time (inter-signal interval, known as ISI) children need between two sounds before recognizing that there is more than one sound. The children showed deficits in comparison to controls when one stimulus rapidly followed another in both a temporal order judgment paradigm (TOJ) and a samedifferent discrimination paradigm. Similar deficits were th en observed in 8 out of 20 dyslexic children (Tallal, 1980). ISI is called timing threshol d and it is interpreted as the time it takes for nerve cells to fire and process a sounds acoustic features and th en recover to pick up the next sound. While the average child has an average timi ng threshold in the tens of milliseconds range for simple tones, children with reading disabilities have time thresholds measuring hundreds of milliseconds. For example, because the difference be tween ba and ad occur within tens of milliseconds, children who need longer time to detect changes may not hear the difference. If they cannot discriminate phonetically simila r sounds, this may cause underspecified phonological representations of words in lexic on. Theoretically, it was argued that a rapid processing deficit could affect literacy because ef ficient processing of transient information is

PAGE 46

46 critical for phoneme perception and fully speci fied phonological repres entation of words in permanent lexicon. This stable, correct phonological specification is regarded as the basis of competent phonological processing (phonemic awar eness, short tem retention, and lexical access), which are necessary for successful reading. This auditory processing deficit theory has become so dominant that a remediation package based on the elongation of brief perceptual cues has been developed and is administered to thousands of children (Merzenich, Jenkins, Johnson, Schreiner, Mi ller, & Tallal, 1996). However, this theory has lost favor despit e its logical appeal (Goswami, 2003). Common criticisms are that: (i) positive findings are difficu lt to replicate; (ii) that only sub-groups of dyslexics are affected; (iii) that when positive relationships are found they are more robust in control groups, (iv) and that when auditory de ficits are found they te nd to be small (Ramus, 2003; Rosen, 2003). In response to the criticism, th e proponents of this theo ry argue that, by the time children are diagnosed with dyslexia around age 9, their brains may have compensated for the auditory deficit, but early deficit may have laid the foundati on for trouble with other subtle phonological processing skills su ch as phonemic awareness. Hearing loss, Phonology, and Literacy It is now a well established consensus that phonological representa tion capacity plays a crucial role in language development and langua ge-based cognitive functions such as verbal memory, speech perception, metalinguistic awareness, and literacy in normal hearing children. If, for some reason, children s phonological representation is adversely affected, it is hypothesized that skills mentioned above clos ely associated with appropriate phonological foundation would be compromised. As far as h earing loss is concerned, the most important question is to what extent ch ildrens phonological capacity will be damaged by limited auditory input (Leybaert, 1998).

PAGE 47

47 It should be mentioned that most of previ ous research have focused on children with profound hearing loss and very few is known about th e literacy or literacy-re lated cognitive skills in children with mild-to-moderate hearing loss. Allen & Schoem (1997) reported that even hardof-hearing children, that is, chil dren with only mild-to-moderate hearing loss, read at lower median levels than do hearing children. Another limitation of the previous research on the literacy development of heard-of-h earing children is that evaluations of reading skills were done in very limited areas. Thus, understanding th e developmental change of a wider and comprehensive range of literacy and related c ognitive skills is not yet seen in the current literature. Phonological Processing, Language Skills, and Hearing Loss Childrens poor word learning can be explained by either m isrepresented phonological features or insufficient memory resources (R amus, 2001b). The quality of incoming auditory input will be limited in children with heari ng losses and this might cause the phonological component to be affected especially during the so-called sensitive or critical period of language acquisition. This time-specific delay in phonological component may be the cause of persistent language impairment a nd later literacy wea kness in children with HL. Under-specified phonological input or lack of enough storage cap acity can adversely affect normal lexical growth. If a child has a permanent or tempor ary limitation of auditory input, his overall phonological processing skills can be impacted by way of various routes. Therefore, different from normal-hearing children, hard-of-hearing childrens compromised phonological processing skills can have a negative impact on later vocabulary, grammatical knowledge, and overall language competence.

PAGE 48

48 Otitis media with effusion (OME) Fluctuating conductive hearing loss associated w ith repeated episodes of otitis m edia with effusion (OME) occurs most often during the firs t three years of life, a time that is most important for age-appropriate language developmen t. OME is the presence of the fluid in the middle ear that is not infected. OME typically cau ses a mild-to-moderate hearing that lasts as long as the fluid persists and the variability in auditory input due to OME has been hypothesized to disrupt childrens ability to code phonologica l information accurately into their phonological representation system (working memory), build ing up inaccurate sound representation in the lexical storage (long-term memory). Several studies have tried to relate an early history of OME to language development during the first 3 years of life to examine the eff ect of OME, which usually occur during critical period of spoken language acquisition. It is posited th at early normal audito ry functions would facilitate the acquisition of language-specific sp eech perceptual strate gy. In turn, normal access to incoming sounds would also facilitate: (1) the phonological coding via high-quality resolution of acoustic features; (2) stable formation of phonological representation of a word in lexical memory; (3) efficient word retrieval, and (4) enhanced phonological short-term memory (Nittrouer and Burton, 2005). One correlation study reported that children with OME episodes during the first 2 years of life scored slightly lower in expressive language, but caught up by second grade (Roberts, Burchinal, & Zeisel, 2002). Early impact of OME (first 2 years) on expressive language skills seem s to shed light on the importance of cr itical time window for language acquisition, beyond which children do not show fast developmental ra te. But, this possible delay in expressive language skills seems to be re solved as children enter 2nd grade level if children are provided

PAGE 49

49 appropriate audiologic or speech-language inte rvention. This suggests that OMEs effect on language development is not substantial. Similar observation can be found in a very recent study. Majerus, Amand, Boniver, Demanez, Demanez, & Linden (2005) reported that 8-year-old children with OME had strictly normal performance for (1) expressive/recepti ve vocabulary, (2) verbal STM (new word learning/serial words recall), but (3) slight decrea se of performance was found in phonological processing/ awareness level (nonword repetition, a rhyme judgment task) when compared with normal-hearing controls. In Roberts et al. ( 2002)s study, students caught up with the hearing peers language skills by the second grade and this period approximates to the mean age of 8, which was reported to be the chronological age when no continuing language delay was documented in Majerus et al. (2005). However, different observation was document ed in Nittrouer et al. (2005). This study analyzed 5-year-old childrens phonological awareness, verbal s hort-term memory, and sentence comprehension. Children with OME showed impaired phonological awareness and verbal working memory. Different from other previous studies, which reported no significant risk for language development (cf. Majerus et al., 2005, Briscoe et al., 2001), this research found significant delay in language (sentence comprehension) and i ndicated that impaired phonological processing was associated with later language delay (syntax). Nittrouer and her colleagues suggested that this different resu lt should be reinterpreted in the context of methods used. It is indicated that most of previous studies repor ting no significant language impaired in children with OME used parental checklists (e.g., MCDI), or non-in-depth standardized assessment tools (e.g., CELF), which they say may not assess in-depth language skill. With this in mind, it can be

PAGE 50

50 expected that Majerus et al. (20015s documenta tion of normal language skills found in children with impaired phonological processing skills might be changed if we use in-depth language tests. Nittrouer and the colleagues also suggested that no group difference between normalhearing and hard-of-hearing childr en might be explained by the fact that the majority of children in both OME and non-OME groups came from lo w-SES group. Because of high correlation between SES and language skills the possible language difference might have been masked by similarly low language skills in both groups. For this reason, careful control of the SES factor is required for exact evaluation of later language, or literacy capac ity. Finally, the mean age of participants also appears to be an importa nt factor for this outcome discrepancy. As mentioned in Majerus et al. (2005), overa ll evidence for residual STM impairment in children with OME remains equivocal depending upon different tasks used and subject-related factors such as family supportiveness, quality of later audiologic, spee ch-language intervention, and socio-economic status (SES), which usually were not reflected in a measurable way. According to Majerus et al. (2005), hearing loss can affect verb al STM or long-term lexical memory in three different ways: (1) residual pe rceptual deficit might reduce the accuracy of phonological coding (initial represen tation in the model) in STM area; (2) this poor phonological coding can cause poor phonological memory because of insufficient acoustic resolution; or (3) limited perceptual input during in fancy may lead to less structured phonological form in mental lexicon, which is related to later weak vocabulary level. Phonological repres entation of a lexical item with weak and low acoustic resolution would hinder efficient both sh ort-term memory tasks and fast lexical retrieval for ageappropriate comprehension/production. Majerus et al. (2005) re ported that OME does not have a dverse effect on verbal STM, and phonological processing ability would be subtly impaired. They went on to say that this might

PAGE 51

51 indicate the delay of phonological processing tasks in children with OME cannot be related to their normal verbal STM capacity. In this st udy, nonword repetition sk ill was shown to be slightly impaired, but it is not sure why this task was not classified as ve rbal STM. In Norbury, Bishop, & Briscoe (2001) and Baddeley et al. (1 998), it is clearly i ndicated that nonword repetition is best understood as a measure of the capacity of the phonological short-term memory. If we follow this standpoint, the data in this study confirm that part of verbal STM skill (nonword repetition) is significantly decreased. In this respect, fu rther in-depth analysis of the memory skills tapped into by different task s is strongly recommende d, without which it is meaningless to discuss the effect of heari ng loss on phonological verbal memory (i.e., STM, WM, or LTM). Interestingly, verbal STM skill in children w ith a history of OME was shown significantly delayed or quite normal according to the age of participants. In contrast to Majerus (2005), Nittrouer et al. (2005) reported de layed verbal STM. The average age of participants in this study was 5 (age range: 4;11-5;11 months) and the mean age of subjects in Ma jerus et al. (2005)s study, which reported normal verbal STM capacity, was 8 years. Considering the fact that the subjects in Majerus and colleagues (2005) study showed strictly normal vocabulary skill, it is conceivable that limitation in verbal STM in early ages reported does not seem to have negative impact on later language functioning. That is, ea rly history of OME can have adverse impact on initial verbal STM capacity and this prompt lo ss of memory skill seems to be recovered. This could be the reason why childrens language does not show significant delay later on (Majerus et al., 2005; Briscoe et al, 2001) at the age of 7 to 8 years (cf. second grade). It could be that intensive audiologic or speech-language thera py combined with good family support and SES helped children to catch up with the language skills of normal-hearing peers.

PAGE 52

52 Sensorineural hearing loss To date, substantial am ount of research has been conducted to investigate the effect of OME on later language. Thus, much less is known about the development of phonology and language in children with permanent mild-to-mo derate SNHL (Stelmachowicz, Pittman, Hoover, & Lewis, 2004; Briscoe et al ., 2001; Gilbertson & Kamhi, 1995; Plapinger and Sikora, 1995; Davis et al., 1986). Briscoe et al. (2001) assessed three phonological processing skills: phonological discrimination, phonological awarene ss (onset-rime detection), and STM (nonword repetition) in 4 groups of children (SLI, SNHL, and 2 control groups). Children with SNHL showed scored significantly weakened phonological processi ng and phonological STM skills than a control group matched on chronological age. But, no di fference was observed in SNH and CA control groups on vocabulary and sentence comprehens ion (syntax). Impressively, considerable individual variation within the SNH group was found; nearly 50% of the SNH group classified impaired group showed phonological impairment associated with poorer vocabulary than remaining children who had no impaired phonolo gy and vocabulary, but both groups language skills were within normal range. It is notable that the unimp aired group showed relatively m ilder hearing loss. This might suggest a causal link between the degree of hear ing loss (impaired phonology) and the resulting decrease in language skill in the impaired subgroup. But, the overall result is that impaired phonological processing due to SNHL can occur w ithout clinically signi ficant deficits in language. As criticized by Nittrouer et al. (2005), this might be related to shallow assessment of language skills, which may not find existi ng language difference in deep level. Gilbertson and Kamhi (1995) were concerne d about phonological skills and new word learning in children with SNHL. Substantial va riability was found on m easures of vocabulary

PAGE 53

53 and word learning, that is, half of the children with SNHL sh owed normal development and the other half of the children (n =10) performed more poorly than the first group. Based on this observation, they argued that ch ildren with SNHL can be divided into two distin ct subgroups: (1) one lower-functioning group with impaired wo rd learning, vocabulary; (2) one higher functioning group with normal language ability. They went on to say that hearing level was not significantly related to word-learning or measur es of language skills si nce the higher-functioning group had poorer SRT scores than the lower-f unctioning group. It is argued that the higherfunctioning groups better language (lexical) skill is related to their better phonological memory (new word learning, word repetition). This had led them to the conclusion that one out of every two children with a hearing loss might be considered as language-impaired. However, for the following reasons, this s eemingly important conclusion needs to be carefully interpreted. First, the data show that low-functioning children ha ve better rapid naming skill which is a measure of phonological long-te rm memory, but this observation was not discussed at all. In contrast, the better STM function (word re petition) of the higher-function group was regarded as a crucial factor for effi cient lexical acquisition. Secondly, the scores in grammatical understanding test did not yield a significant language gap between two groups. Also, the small number of subjects in both groups (n=10 and 9) also needs to be considered in terms of statistical power. Thirdly, it is argued th at this intra-group differe nce is not related with the severity of hearing loss, but no explanation regarding the possible link between hearing loss and lower functioning groups impaired phonologi cal memory (STM) was provided. Lastly, as noted by the authors, substantial amount of variance was observed in the performance of children. The delayed vocabulary in their lower-functioning group may have resulted from a set of other extraneous factors such as poor fa mily factor (responsive ness, supportiveness), low

PAGE 54

54 quality of intervention, inappropr iate educational program, higher age of identification, short duration of hearing aid wearing, lo w SES, and so on. All of these factors might have contributed to the group difference. Considering the inap propriate data interpretation and the variance seemingly linked to other uncontro lled factors, vocabulary difference may not be a definitely reliable criterion for suggesti ng that one out of every two hearing impaired children has concomitant language disorder. Hearing Loss and Literacy Conductive hearing loss Links between audition, language, and literacy ha ve been widely studied in children with conductive hearing loss caused by OME (Friel-P atti, 1990; Mody, Schwartz, Gravel, & Ruben, 1999). Despite a considerable num ber of studies conducted during the past three decades on whether children with frequent OME score lowe r on measures of speech, language, and reading than children without such a histor y, the literature is still controversial. Mo reover, there are very few comprehensive studies on the effect of OME-related hearing loss on literacy development and the results of research, if any, are quite he terogeneous. One problem of most of studies is that very limited areas of reading or underlying skills of reading are assessed. For example, one longitudinal study with subjects from second graders measured only one area for reading assessment: letter-naming (iden tification) skill an d concluded that no correlation was found between a history of OME occurred up to 4 years of age and literacy (Roberts et al., 2002). This is surely a problem since the subjects in this study were at age range of 5 to 7 years of age, when most of children would master letter naming. Peters (1994) conducted a longit udinal study testing the effect of middle ear infection on reading and spelling. Subjects were tested on non-word reading, word recognition, reading comprehension, and sentence identifi cation at ages of two, four, a nd seven. The results indicated

PAGE 55

55 that ear infections had a si gnificant effect on spelling, but not on reading. Some authors reported difficulties with phonological skills with unimpaired general language skills in children with OM (Peters, 1994), but others did not reveal any difference (Roberts, Burc hinal, & Zeisel, 2002). A recent meta-analysis by Roberts, Rosenfeld, and Zeisel (2004) reported no significant difference a number of standardized language measures at preschool age and OME -related hearing loss. Share and Chalmers (1986) is also a long itudinal study on the relationship between ear infection and reading. They reported no significant effect of middle ear infection on reading ability. Hemmer and Ratner (1994) studied the im pact of middle ear in fection on literacy using six pairs of twins through a longitudinal obser vation. One of each pair had a repeated ear infection and one did not. Middl e ear infection was shown to have a negative effect on vocabulary skill, but had no effect on speech, ot her linguistic skills, and reading. Abu-Rabia (2002) sampled 49 first graders all from low SES families and 11 children with a history of at least one episode of middle ear disease we re assessed for a wide range of phonological awareness skills and pseudo words reading. No significant differences were found on the phonological awareness tasks, and non-word readi ng task. These findings accord with the above results, suggesting that middle ear infection does not adversely impair childrens phonological sensitivity and reading development. In contrast, different result was reported by Nittrouer (2005), in which four groups of second graders participated. It is asserted that conductive hearing loss may have adverse effect on language when it occurs in association with othe r social or health risk factors. In this study, one group, the control, was from the middle cla ss and had no ear infection; the second group, also from the middle class, had a history of ear infection; the third and the fourth groups of low SES, was with or without ear in fection history, respect ively. The rationale of the study was that

PAGE 56

56 middle ear infection causes tempor ary hearing loss, whic h Nittrouer expected would affect the amount of exposure to language. Nittrouer hypothes ized that limited language experience would account for the variance in phonological awareness skills. As stated above, the underlyi ng hypothesis of this study wa s that phonological awareness can only be developed via explicit exposure to acoustical speech sounds. Significant difference in the performance on phonological awareness tasks was found between the two groups from the middle class, but no differences was reported between groups with low SES, i.e., both groups from low SES families performed at a low level. Nittrouer explained that this lack of group difference in samples from low SES families is due to the fact that slight exposure to print of these children in their homes provided a negative environment against normal phonological and reading development. So, regardless of the h earing status, the possible group difference might have been masked by simultaneous impairment due to the low SES. Similarly, Yoshinaga-Itano (1999) reported adverse overall language outco mes for children attending early intervention services even when the loss was classified as m ild, with outcomes more st rongly related to age of diagnosis than to severity of loss. Similarl y, Davis, Elfenbein, Schu m, and Bentler (1986) reported receptive vocabulary, ve rbal ability, and reasoning to be more than one standard deviation below the mean, even in children with mild hearing loss. As seen from the literature, the effects of fluctuating hear ing loss related to middle ear infection, more specifically the OME, on phonology, language, a nd literacy are still not explicitly determined. However, the overall im pression is that mild or moderate conductive hearing loss does not severely impair childrens phonologica l skills. Based on this observation, it can be expected that skills of language and literacy associat ed with conductive hearing loss would not be significantly compromised. Howe ver, reflecting upon possible delay though not

PAGE 57

57 statistically significant, it is strongly suggested that further studies include the assessment of wider and deeper range of phonol ogical capacities, reading, a nd cognitive skills underlying reading process in the children w ith mild or moderate conductive hearing loss for evidence-based intervention design. Sensorineural hearing loss and literacy Reading dev elopment in children with permanent SNHL has received relatively much less attention than conductive hearing loss related to middle ear disease. In contrast to children with fluctuating hearing loss, very scant research effort s have been directed to literacy development in children with permanent SNHL. Although there is a large body of research documenting the language and literacy outcomes of children with SNHL, most of this was concentrated on children with profound hearing impairment. So, mu ch less is known about literacy development for these children, even thought it can affect not only hearing level thresholds, but also frequency discrimination (Moore, 1995) and fu rther research strictly focu sing on the reading and readingrelated cognitive skills in this population is warranted. Briscoe et al. (2001) compared 5to 10-year -old children with mild-to-moderate SNHL, children with specific language impairments an d children with no hearing or no languagelearning difficulties. In common with the SLI group, mean scores of children with hearing loss were significantly poorer on tests of phonological STM (non-word repetition), phonological discrimination, and phonological awareness than chr onological age controls But, no differences were observed in SNH and age contro l group on language (vocabulary, sentence comprehension), digit/sentence recall (STM), and literacy (reading comprehension, nonword decoding, and word recognition). That is, while the data revealed little overall difference of the reading abilities between the hearing impair ed and normally hearing children, the hearing

PAGE 58

58 impaired childrens phonological skills were signifi cantly inferior to children in the age-control group. Interestingly, Briscoe and colleagues study also reported that there was considerable individual variation within the SNH group. Th at is, nearly 50% of the SNH group showed phonological impairment associated with poorer expressive and receptive vocabulary and literacy. In addition, this subgr oup showed higher hearing thres holds than remaining children without phonological impairment. Thus, it was suggested that there was a link between vocabulary (language) and literacy and phonological skills associated with hearing loss. Strikingly, all three areas of reading (word/nonword reading, comprehension) were shown quite normal (99.5, 91.6, and 95.3, respectively), revea ling no significant between impaired and unimpaired subgroups. Overall, this study suggest ed that major problems in nonword repetition and depressed phonological component occurred without clinically significant deficits in wider language and literacy abili ties in children with mild-to-moderate SNHL. Impaired phonology in children with hearing loss In Briscoe and her colleagues (2002) study, it was hypothesized that children with hearing im pairment might show phonological impairment sim ilar to those seen in SLI and this was well observed. But, this expected link between compromised damaged phonology and other language and literacy measures was not found in children with SNHL. The hearing impaired childrens success w ith reading in spite of depressed phonological processing skills is very notable. It was suggested that even though slight language deficiencies (receptive/expressive vocabulary) were observe d, phonological depression linked with auditory limitation would not impede reading skills of children with mild-to-moderate SNHL and at least some reading may be possible even without closely a ssociated phonological skills (Gibbs, 2004). This led the authors to concl ude that auditory deficit can compromise phonological skills, but

PAGE 59

59 impaired phonological skill does not necessarily lead to seriou s impairments in the reading performance of hearing impaired children. Theoretically, this argument poses a very important aspect of impaired phonological component found in children with hearing loss si nce the argument itself is the converse of numerous studies of children with dyslexia, wh ich all agree in indica ting that poor phonological capacity will compromise the de velopment of word reading. A very similar phenomenon was reported in Gibbs (2004)s study. Fifteen children with mild-to-moderate bilateral SNHL were comp ared to normally hearing controls on their performance on reading and underpinning skills. Gi bbs showed that heari ng impaired childrens reading abilities were indistinguishable from their hearing peers, while showing phonological skills that were not as good as the controls. S o, Gibbs (2004) supports the findings of Briscoe et al. (2001) and offers a similar challenge to the necessity of phonological skills in the development of early reading and the universality of phonological core deficit theory. Summary The following is a short summary of current l iterature about the eff ect of hearing loss on phonology, language, and literacy. Conductive hearing loss While there have been a substantial num ber of studies of profoundly deaf childrens reading, the literacy of children with mild -to-moderate fluctuating conductive and permanent SNHL has relatively not received enough attention. Results from the studies conducted during the past three decades on the effect of OMErelated hearing loss on speech, language, a nd reading are still controversial. However, the overall impression is that mild or moderate conductive hearing loss does not impair childrens phonological and linguistic knowledge in a way that their reading level is significantly delayed or depressed. Thus, we anti cipate that skills of language and literacy associated with conductive hearing loss would not be significantly compromised.

PAGE 60

60 Permanent SNHL In contrast to children with fl uctuating hearing loss, very scan t research h as been directed to literacy development in children with permanent SNHL. So, much less is known about literacy development in this population. Furthe r research strictly focusing on the reading and reading-related cognitive skills in this population is warranted. The data revealed little difference of langua ge and reading abilitie s between the hearing impaired and normally hearing children, but the hearing impaired childrens phonological skills were significantly inferior to children in the age-cont rol group. The hearing impaired childrens success with reading in spite of depressed phonological processing skills was noted. Thus, it is assumed that even though slight language deficiency (receptive/expressive vocabulary) was observed, phonologi cal depression linked with auditory limitation would not impede reading skills of hearing impaired children. Furthermore, possible weakness in parts of language areas (receptive vocabulary) does not seem to have significantly adverse impact on later reading ca pacity in this population. Based on this, phonological strength would be regarded as a sufficient condition for normal literacy, but not as a necessary condition for age-appr opriate written language development in hearing-impaired children. Children with less significant hearing loss or t hose with relatively more auditory access to phonemes did not readily use phonological inform ation for reading words in the same way as hearing readers. This preferential use of the visual route was manifested by more dependence upon sight word recognition strategy in children with mild or moderate hearing loss. Further studies which can investig ate hearing impaired childrens use of sight word use would reveal an interesting fact which would be of clinical importance. To tap childrens phonological memory skills (STM, WM, and LTM), a consistent, theorybased theoretical framework should be used to categorize each memory task. Without this, the interpretation of data could be misleading. A further recommendation is that the socio-economi c status factor should be considered as an influential factor fo r reading achievement. Finally, two issues can be mentioned regardi ng the research design. First, a longitudinal study design with longer period of observation for reading development is quite warranted to see the trend of literacy development in children with hearing loss. Second, childrens hearing levels (aided or unaid ed) should be representative of the population to be studied. That is, the possible group difference associated with hearing level effect can be obscured when the distribution of hearing loss is skewed.

PAGE 61

61 Figure 2-1. Three different areas of phonological processing (Gillon, 2004). Figure 2-2. Phonological loop (Baddeley, 1986) Phonological processing Retrieval of phonological information stored in long-term memory Phonological awareness (syllabl e, intra-syllable, phoneme) Short-term coding (representation) of phonological information Central executive Visuo-spatial STM Phonological (auditory) loop Rote Retention Processing system phonological awareness verbal rehearsal Pattern matching (lexical retrieval)

PAGE 62

62 3 4 3 5 3 3 3 4 3 5 Figure 2-3. Word recognition through phonologi cal route (revised from Ramus, 2001) Speech [ C A T ] Acoustic re pr esentation Perceived as Non-speech signal Sub-lexical phonological representation (words or non-words) Short-term retention memory Processed as speech signal Phonological form (phonological long-term memory) Working memory (can process both words and nonwords) 1 2 PROCESSING CAPACITY (phonological awareness) 1. meta-phonological 2. manipulation tasks STORAGE CAPACITY

PAGE 63

63 Figure 2-4. Model of (underspeci fied) phonological representation Permanent lexical storage S T R I N G (real word) S T A T I N (nonsense word) Underspecified representation S P R ? N G S T A ? I N Weak representation in shortterm phonological buffer Underspecified representation would hinder long retention Phonemic blending needed for nonword repetition or word decoding would fail. Phonological form Semantic content Due to insufficient phonological cue, access to phonological information will be ham p ered. Ultimate activation of semantic content will be delayed (slow word recognition and reading comprehension) Lexical retrieval Short term memory Working memory Baddeleys working memory

PAGE 64

64 Input representation deficit Post-input processing deficit Figure 2-5. Two different aspects of phonol ogical representations (Goswami, 2000). Speech perception (recognition) Phoneme perception Lexical decision Speech gating Speech production Rapid Automatized Naming (slower activation of lexical information in lexical retrieval Non-word repetition (STM) Less distinctness of vowel production Phonological representation of words

PAGE 65

65 Figure 2-6. Simplified model of vi sual naming (Wolf et al, 2000:394) Attention activation Visual processing *PSR (shape of letter B) Inte g ration p rocesses *PSR Corresponding lexical information *PSR phonological representation orthographical representation Lexical process *PSR [Access to and retrieval of] phonological code -semantic code Motor planning *PSR Articulation (naming) *PSR PSR= processing speed requirement

PAGE 66

66 CHAPTER 3 METHODS AND MATERIALS Introduction Despite the substantial num ber of empi rical studies on phonological development and phonology-based reading skills in children with d eafness, little is known about hard-of-hearing childrens skills in similar areas. Most previous studies concur that hearing impaired childrens phonological processing skills could be adversely affected duri ng the critical period of language development possibly due to decreas ed acuity of peripheral audition related to congenital hearing loss. However, impaired phonological processing skill has not been reported to negatively impact reading and relate d cognitive skills (Briscoe et al., 2001; Gibbs, 2004). This study is designed to further explor e relationship between phonological processes, spoken language, and reading. This cross-sectional study addressed th e following primary research question: What are the effects of mild to moderate sensorin eural hearing loss on th e phonological processing skills, oral language, and reading ability? It was hypothesized that hearing loss would affect phonological awareness, rapid naming, verbal me mory, vocabulary and grammar, and reading skills (reading fluency and comprehension). To i nvestigate this question, the investigator used a theoretical framework of phonological core de ficit hypothesis, estab lished in the study of developmental reading disorders. A wide range of tests in the areas of r eading and spelling, oral language, phonological processing skills, auditory processing ability, and basic auditory skills was used to accomplish this objective. Reading skills were measured for word reading under timed and untimed conditions, spelling, oral read ing fluency, and reading comprehension. Also, a range of phonological processes were measured, includ ing (1) phonological awareness (blending and

PAGE 67

67 deletion), (2) phonological short-te rm memory (digit span, nonwor d repetition), and (3) rapid naming skills (digits and letters). Further, audito ry processing skills were investigated using a subset of tests from a screening test of central auditory processing abilit y. Basic auditory skills were checked by obtaining thresholds for stimu li consisting of pure-tones and speech signals. The research methods presented in this se ction are addressed under the headings of recruitment setting, participants and selection criteria, procedures, trea tment of the data, and research questions. Setting and Participants The purpose of this section is to describe th e instructional settings where the study took place and to provide a d escription of the participants. Demographic and audiologic characteristics of the participants ar e provided for three groups of subjects. Recruitment Setting Three groups of children between 7 to 12 years o f age, were recruite d for the study: (1) 19 children with mild to moderate hearing loss m easured by their better ears pure-tone average (PTA); (2) 29 children with normal hearing and reading skills; and (3) 30 dyslexic children with normal hearing ability. The University of Florid a Institutional Review Board (IRB-02) approval (UFIRB 2006-U-0535) was received to recruit participants and c onduct the research. Data were collected from October 2006 through October, 2007. For the group of children with hearing loss, 2,500 flyers and recruiting letter s containing information about th e study and contact information were sent out to public elementary schools locate d in the following ten c ounties in north central Florida region: Alachua, Clay, D uval, Gilchrist, Lake, Marion, Orange, Putman, St. Johns, and Volusia Counties. Around fifty parents or guard ians of hearing impair ed students expressed interest in the study thro ugh reply letters, emails, or phone call s. All students were enrolled in public elementary schools in the central Florid a area. Participants with hearing loss were

PAGE 68

68 recruited from several sources including contracts with th e Exceptional Student Education department (ESE) of each county, advertisements posted at hospitals and several churches located in Gainesville, FL., and personal acqua intances. For the hearing impaired subjects, directors of ESE department were contacted to help distribute re cruiting letters to schools which had special program for hard-of-hearing childr en. Letters were sent to the families through classroom teachers. Of the 69 (2.8%) children w ho returned permission forms or expressed their interests, 50 (70%) were qualifie d to participate in the study. All participants with normal hearing and reading ability were recruited from twelve elementary schools located in Gainesville, FL through the department of research at the Alachua County School Board. Selection Criteria Potential pa rticipants were required to satisfy the following criteria. Each participant must be a nativ e speaker of American English. Each participants non-v erbal intelligence screened by the Test of Nonverbal Intelligence-3 (TONI-3, Brown L., Sherbenou R. J., & Johnsen S. K., 1997) must be equal to or more than 80. All participants were required to have normal tympanograms and no signs of middle ear infection at the time of testing. All parents of participating ch ildren were required to read and sign the informed consent letter. Childrens hearing loss must be binaural sensorineural and no t associated with any other sensory impairments, neurological and/or neurodevelopmental disturbances. When potential subjects were identif ied, children with monaural SNHL and/or cochlear implant devices were excluded. Hearing loss must be congenital and the particip ants must wear hearing aids for both ears at the time of testing. Exception was allowed only when the hearing loss was very mild (26 dB HL to 30 dB HL) and the participant s word recognition scores for both ears were 100% at the most comfortable level (MCL). All hearing-impaired particip ants were required to atte nd a mainstream school and use speech mode as their primary communication mode.

PAGE 69

69 Participants Parental consent was obtained for each child pa rticipant at the onset of the investigation. When a fa mily agreed to have their child partic ipate in the study, the ch ilds parent or guardian provided informed consent as required by the co urse of the study and a completed questionnaire (Appendix A). The information in this form wa s used to check the subjects language and reading ability, the status of hearing loss, age at identification of hearing loss, age at initial audiologic intervention, duration and frequency of speech language services, developmental change of patterns or types of hearing loss, family history of hearing loss, and any known etiologic cues. Children were not compensated fo r their time. Instead, the participants were provided a letter of test sc ores within three weeks. The participants in the presen t study initially consisted of two groups of students between 7 and 12 years of age: (1) A sample of 29 norma lly developing children, a nd (2) a sample of 19 children with mild to moderate binaural SNHL. Et hnicity representation fo r the students in the HI and the NH groups was predominantly Caucasia n (n=38), followed by Asian American (n=5), Hispanic (n=3), and African-American (n=2). After the data collection of these two groups was over, an additional archival data set of thirty dyslexic children was incorporated for the purpose of comparison and formed a third experimental group. Participants with normal hearin g and reading skills (NH group) For the control group of children with norm al hear ing ability and reading skills, thirty eight potential participants responded. Af ter removing nine students who did not satisfied all inclusive and exclusive criteria for the following reasons the group was reduced to twenty nine (boys=16, girls=13). None of the children participating in the control group were known to have histories of speech, language, or hearing problems, or a ny type of exceptional educational needs.

PAGE 70

70 Three students had bilingual background (one for Spanish and two for Korean language) According to the questionnaire form, four ch ildren were reported to have developmental reading difficulties. One of them actually co mpleted the protocol for the study, but due to this problem, the data could not be included in the final data. Two students who also completed the protocol were excluded from the data since their ages were either lower or higher th an the suggested research plan. The mean age (in months) and mean grade (in years and months) of this NH group were 111;9 (SD= 14.0) and 3.8 (SD=1.15), respectively. Participants with hearing impairment (HI group) Thirty-one children with prelinguistic, sensorineural hear ing loss were contacted. The childrens parents or guardians we re asked if their children had b een or were currently receiving speech or language therapy. Only f our participan ts who enrolled in the study were not receiving speech and language therapy. Al l subjects in HI group had binaur al sensorineural hearing with better ears hearing thresholds in the mild to mo derate range (26 to 70 dB HL) as assessed by the pure-tone average (PTA). All children wore ear-level hearing aids, with only one subject aided monaurally. Two categories of hearing loss we re identified in this sample, on the basis of the childs PTA in the better ear. Mild hear ing loss defined as having a PTA of 20 40dB HL and moderate hearing loss was defined as a PTA of 41 dB HL Audiometric assessments revealed that 6 children had mild hearing loss, whereas 13 children met the criteria for moderate hearing loss. Information about hearing loss, audiologic s ervices received, and c linical histories were determined through interviews with at least one parent or guardia n and/or a parental questionnaire form provided in the recruiting le tter packet. Among these thirty-one potential participants, twelve children who did not satisfy the inclusiv e and exclusive criteria were excluded from the study.

PAGE 71

71 Four children were reported to ha ve only monaural hearing loss. Three children had bilingual background and th ey were all Spanish-speaking participants. Three subjects were reported to wear cochlear implants for at least one ear. Two children completed the protocol for the st udy; however, their data were excluded from final data since their hearing losses were found to have developed after four and three years of age, respectively. Nineteen hearing impaired children included in the final selection (mean age [in months] = 110.8, SD = 19.3, mean grade [in years and months ] = 3.4, SD = 1.61, boys = 11, girls = 8) were designated as the experimental group (HI group). Th ey were all enrolled in public elementary schools that provided oral program ming for hard-of-hearing students and emphasized use of their residual audition for the developmen t of speech and language skills. Dyslexic group (RD group) To enable a rigorous investigation of the st rengths and the weaknesses of the HI groups performance on phonological processes and literacy sk ills, the investigator decided to add a third group of children with specific reading disorder w ho have no auditory perceptual limitations due to cochlear damage. That is, the HI ch ildren performances on phonological processing and literacy were compared to those of age and gr ade-level matched dyslexic controls of similar nonverbal ability. Many dyslexic children have a history of language difficulty (Rutter and Yule 1975) and dyslexia is conceptualized either as a mild form of language impairment, affecting only the phonological system, or as a residual problem that remains when oral language difficulties have resolved (Aram, Ekelman, and Natio n 1984; Scarborough and Dobrich, 1990). Based on the fact that the auditory system is crucial for the development of language, many researchers have suggested that for at least so me of the children with phonologic dyslexia, there may be a disorder within the a uditory system that has disrup ted normal acquisition of language.

PAGE 72

72 However, unlike hearing impaired children, the di sruption is not occurring at the periphery, but at certain point in the ascending auditory pathway or the cortical level, through intrahemispheric, interhemispheric or associative connections (Moncr ieff, 2002). There is evidence to indicate that dyslexic readers have abnormalities within some of the auditory structures necessary for language development, including symmetry diff erences of the planum temporale (Hynd, Semrud-Clikemand, Lorys, Novey, and Eliopul os Hynd, 1990; Kushch, Gross-Glenn, Jallad, Lubs, Rabin, Feldman, and Duara, 1993; Larsen, Hoien, Lundberg, and Odegaard, 1990; Leonard, et al. 1993), abnormal portions of the co rpus callosum (Duara, Kushch, Gross-Gleen, Barker, Jallad, Pascal, Loewenstein, Sheldon, Rabin, Levin, Lubs, 1991; Hynd, et al. 1995), and Heschl's gyrus in the right hemisphere (Le onard, et al. 1998; Musiek & Reeves 1990; Penhune, Zatorre, MacDonald, and Evens, 1996). Therefore, these children with specific read ing disorders would serve as an informative comparison group when investigating the impact of congenital peripheral hearing loss on the development of phonological processing and relate d literacy skills in comparison to the one associated with cortical or central disruption of auditory proce ssing. Currently, there is no indepth study that has looked at the different characteristics of hearing impaired and dyslexic childrens performances on phonol ogy and reading measures. Thirty dyslexic childrens data were selected from an archival data set that has been collected from 1999 to 2003 from the Dyslexia C linic Program at the University of Florida Speech and Hearing Clinic (mean age[in months] = 116.6, SD = 20.3, grade[in years and months] = 3.96 SD = 1.62, boys = 18, girls = 12 ). A ll subjects were diagnosed with dyslexia, a specific reading disabilit y. Some parts of variables were not available for this RD group for the following variables: (1) basic audiometry (p uretone and speech), (2) auditory processing

PAGE 73

73 measures, and (3) oral language measures. However, normal hearing abiliti es were confirmed for all dyslexic participants through hearing screen on the day of testing by a certified audiologist in UFSHC or by parent report of norma l hearing based on previous testing. Matching Variables Nor mally developing (NH) and dyslexic subjects (RD) were selected to match age, grade, and sex of the hearing impaired subjects (HI). Statistical comparisons revealed that these variables were well matched on chronological age, grade, and gender for the groups (Age: F[2,75] = .778, p = .463; Gender: F[2,75] = .011, p = .989; Grade: F[2, 75] = .818, p = .445). A non-verbal IQ measure was selected as a covariate in data analyses to reduce any existing effect of intelligence on reading skills. This test was only administered to children in the NH and HI groups. A criterion nonv erbal IQ of 80 as assessed by the TONI-3 was required. Controlling for nonverbal IQ was deemed n ecessary because a univariate one-way ANOVA confirmed a significant gr oup effect (F[1,45] = 3.406, p = .042). Therefore, controlling for this nonverbal IQ measure was ju stified (See Table 3-1). Procedure The experim ental tasks were completed in two separate sessions in a quiet room in the UFSHC and the speech hearing clinic in the Univ ersity of Central Florid a. In the first session, which lasted for about 90 minutes, general informa tion about the test and the test procedure was explained to students and their parents or guard ians. Upon successful conf irmation of eligibility, each participant signed a written consent for the data to be used for research purposes and completed all audiologic testing. In a second sessi on that followed within two weeks, tests were administered in the areas of phonology, oral la nguage, and reading for about 120 minutes.

PAGE 74

74 Because of decreased speech intelligibility, he aring impaired childrens responses on all of phonological processing tasks were recorded digi tally using Olympus digital voice recorder (Model No. DS-40) and phonetically tran scribed by the author for scoring. The author and two other research assistants administered tests to all participants. Two research assistants were senior undergraduate students in speechlanguage pathology. Both assistants were trained by the i nvestigator in test administration. For reading and language tests, the author provided specialized in struction to the assistants and 50% of training sessions were supervised by a certified speech pa thologist. To ensure reliable test administration and scoring, whenever any deviation from the protocol o ccurred, additional instruction was provided until that assistant was able to demonstrate comp lete compliance with th e testing protocols. All audiologic tests were conducted by three do ctoral students enrolled in the Doctor of Audiology program (AuD) of UF and one certifie d audiologist from UCF. All tests were administered in a soundproof su ite using recorded material. Materials Each NH and HI participant was given tests individually in four separate dom ains, including (1) auditory function, (2) phonological processes, (3) or al language, and (4) reading and spelling from September 2006 through Octo ber 2007. However, RD subjects' data on auditory function and oral langu age were not available from th e archival database. Table 3-2 is a list of all the tests that were used. Descriptions of the test instruments are presented in detail in the next section. Audiologic Measures All subje cts were examined otoscopically to rule out the presence of fluid in the middleear-cleft. Each listener was tested in one 60-min session and received all te sts in the above listed sequence. Both left and right ears were tested to determine the better ear. The audiologic test

PAGE 75

75 battery consisted of (a) a pure-tone threshold test at 250, 500, 1000, 2000, 3000, 4000, and 8000 Hz, (b) a speech recognition threshold test, (c) word recognition test, (d) two central auditory processing tests, and (d) tympanometry. Puretone and speech audiometry The hearing acuity of the children was assess ed in a conventional m anner since all of participants were 7 years or ol der. The air-conducted thresholds were examined in a sound proof booth at 0.24 8 kHz with a GSI 61 clinical audiometer calibrated according to the ISO-389 standards (1985), employing in sert earphones (E-A-R-TONE 3A Insert Earphone) and TDH-39 headphones. A standard audiometric staircase procedure (5 -dB step size; down 10 dB, up 5 dB rule) was used to obtain pure-tone air conduction thresholds. The intens ity recorded as threshold was the lowest intensity at which two correct respons es were given (response in the presence of a stimulus tone) out of four pres entations. Children responded to te st stimuli by hand-raising. Bone conduction thresholds were obtained in a similar manner if air conduction thresholds were 20 dB HL at any frequency. PTA (pure-tone average at the frequencies of 0.5, 1 and 2 kHz in the better ear; right ear if equal hearing) define d each participants hearing threshold. Speech reception thresholds and word recognition scores in percentile were measured via live voice using CID W-22 spondaic word lists (Aud ited Revised Auditory Tests CD). For the NI group, stimuli were provided at the most comfortable level (MCL) as determined in the speech reception thresholds (SRTs). For hearing impaired children, unaided thresholds were obtained. Listeners wore various heari ng aids ranging from basic anal og to high performance digital technology. The subjects in the NH group had pure-tone thres holds of 15 dB HL or better at all octave frequencies from 250 to 8000 Hz with an exception of one male child, whos e right ear pure-tone

PAGE 76

76 threshold for 2 KHz was 21 dB HL. All heari ng subjects also had ex cellent word-recognition scores ranging from 880% to 100% for the CID W-22 word lists (M = 99.0, SD = 2.95). Hearing impaired subjects' pure-tone thresholds at 0.25, 1, and 2 kHz ranged from 0 to 13 dB HL and their word recognition scores measured at the MCL level were moderately good (Right ear: M = 75.2, SD = 29.6, Left ear: M = 86.0, SD = 21.9). Auditory processing tests Based on recommendations by Musiek & Cherma k (1994), four comm only used central auditory tests were administered to all participan ts in the NH and HI groups: (1) One subtest, the Dichotic Digit test, was selected from the Central Aud itory Processing Tests developed by Frank Musiek, and (2) three subtests were selected from the SCAN-C/A screening test for auditory processing disorders (SCAN-C for children aged 3 to 11 years, SCAN-A for subjects aged 12 and more). Musieks (1983) Dichotic Digits test is a test of binaural integration in which pairs of digits are delivered simultaneously to each ear at the MCL, with each ea r receiving a different digit pair. This test has good valid ity data and is simple and quick to administer and to score. Subjects were given three practi ce items to ensure their understa nding of the task. All stimulus items consisted of monosyllabic digits from 1 to 10 (excluding 7) spoken by a male speaker of General American English. Stimuli were routed from a CD player through a two-channel audiometer meeting ANSI (1996) standards a nd were delivered via TDH-49 headphones at an intensity of 40 dB SRT. Twenty stimulus items consisting of four digits each were presented. The test was administered following standard clinical recommendations (Musiek, 1983) using a free-recall paradigm. Performance was scored as a function of percent correct for each ear. The minimum subject age for this test is 7 years (Musiek, 1983). Administration time was approximately 5 to 6 minutes.

PAGE 77

77 The SCAN is a screening test for auditory pr ocessing disorders (Kei th, 1986). It may be used to identify potential factors related to poor social skills, language use, and academic performance in children from 3 to 11 years of ag e. This test was developed to determine possible central nervous system dysfunction by assessing auditory maturation, (b ) identify children at risk for auditory-processing or receptive language pr oblems who may require additional audiological or language testing, and (c) id entify children who may benefit from specific management strategies (Keith, Rudy, Donahue, & Katbamna, 1989). For this study, first three SCAN subtests were gi ven in the order stipulated by standard test format (Filtered Words ([FW], Auditory Figure Ground [AFG], Competing Words [CW]). In the Filtered Words subtest, the subject is asked to repeat words that sound muffled. The test stimuli consist of one syllable words that have been low-pass filtered at 500 Hz. Two practice words and 20 test words are presented to each ear. In the Auditory Figure-Ground subtest, the subject's ability to understand words in the presence of backgr ound noise is measured. One syllable words were recorded in the presence of multi-talker speech babble noise at th e 0 dB signal-to-noise (S/N) ratio. Two practice words and 20 test words were pres ented to each ear. In the Competing Words subtest, the subject hears two words simultane ously--one word presented to each ear. The test stimuli consist of one syllable word pairs pr esented to the right and left ears. First, two practice word pairs and 15 word pairs are presented. The subject is instructed to repeat the words presented in each ear, repeating the word heard in the right ear first. Th en, a second set of two practice word pairs and 15 word pairs are presented, with the subject repea ting the word heard in the left ear first. Middle ear function (tympanometry) All par ticipants were required to pass a tympanometry screening (i.e., type A with equal to or more than 0.2 ml compliance) to ensure normal eardrum and middle ear functions. Acoustic

PAGE 78

78 immittance measures used to determine middle ear function were obtained by using a commercially available middle ear analyzer (G rason Stadler, Model GSI-33). Tympanograms were obtained in both ears on all participants during the first sessi on. Any tympanogram for which tympanometric width could not be calcul ated (i.e., no measurable peak) resulted in a rescheduling of the participant for testing at a later date. Children were considered to have normal middle ear function when their tympanometric width was 250 daPa in both ears (criterion based on Nozza, Bluestone, Kardatzke, & Bachman, 1994). Only one normally hearing participant had middle ear infection at the time of testing. One month la ter, this subject was retested after the infec tion had been treated. Literacy (Reading and Spelling) All subjects in the three groups (NH, HI, and RD) were adm inistered exactly the same set of tests in the areas of phonol ogical processing and literacy (re ading and spelling). To test reading and spelling, eight standard ized subtests from four published tests were used. To test phonological processing skills, six st andardized subtests were ta ken from one published test. A description of each test is provided below. Woodcock Reading Mastery Test Revised (WRMT-R; Woodcock, 1987): Three subtests of the WRMT were administered to assess: (1) untimed phonemic decoding skills for real words and nonwords, and (2) pa ssage reading comprehension. The Word Identification and Word Attack subtests comprise 106 and 45 pronounceable real and pseudo words of increasing complexity, respectively. Especially, in the Word Attack subtest, five unfamiliar words are presented at a time on one page, and the examinee is asked to read them aloud as accurately as possible. Pronunciation rules are provided for the examiner to de termine the accuracy of th e childs responses. This test is discontinued when six errors are made. The Word Attack subtest evaluates the childs ability to use phonic skills to determine the correct pronunc iation of unfamiliar words while reading aloud

PAGE 79

79 letter combinations that form nonsense words. The Passage Comprehension subtest requires participants to read a segment of prose with a mi ssing word and say word(s) to fill in the blank(s) in the printed paragraph. The WRMT-R test record allows for raw scores to be convert ed to age equivalent scores, grade equivalent scores, and standard scores ( M 100, SD 15). The WRMT-R was selected for two reasons: (a) to measure word and nonword reading normed on the same sample, and (b) to use large numbers of test items in order to re duce the likelihood of idiosyncratic word knowledge causing lack of reliabil ity (Olson, Forsberg, Wise, & Rack, 1994). The WRMT-R has more items than other tests with a similar format. Content and concurrent validity ar e well documented in the test manual (Woodcock & Johnson, 1989). Internal reliability coefficients of the WRMT-R calculated based on split half reliability for 1st through 3rd grade ranged from .91 to .98 (M = .94; Woodcock, 1987). Wide Range Achievement Test-3 (WRAT-3; Justas & Wilkensen, 1993): The WRAT-3 includes three subtests that measure reading, sp elling, and arithmetic skills. The WRAT-3 is an academic achievement test that has been shown to have good correlation with the Wechsler Individual Achievement Test. For the purpose of this study, only the Spelling subtest was used to assess the ability to spell single words. Children were asked to write single words on test form after listening to the target word followed by a sa mple sentence. Children are asked to try as hard as they can to spell every word. For each item, targ et word is spoken first in isolation and then in a sentence in which the word is stressed. Finall y, the word is spoken again. Standard scores ( M 100, SD 15) are provided for 32 age groups. Internal consistency reliability figures in the range of r .86 to .91 are reported for children ages 7 to 13 years. The inclusion of a spelling

PAGE 80

80 measure is based on the strong associati on between early spel ling ability, phonological awareness, and beginning reading (Ehri & Wilce, 1987). Test of Word Reading Efficiency (TOWRE; Wagner, Torgesen, & Rashotte, 1999): The TOWRE was given to measure abil ity to pronounce both sight words (Sight Word Efficiency subtest) and nonwords ( Phonemic Decoding Efficiency subtest). In the Sight Word Efficiency subtest, 104 context-free single words of increasing complexity in terms of phonological structure are given. Participants were asked to read aloud as many words as possible in 45 seconds. Similarly, the Phonemic Decoding Efficiency subtest of the TOWR E was administered to quantify rapid nonword decoding skill. The subtest presents 63 pronounceable pseudo-words and participants were aske d to read as many nonwords as possible in 45 seconds. Gray Oral Reading Tests-4 (GORT-4; Wiederholt & Bryant, 2001) : The GORT-4, a measure of reading fluency (accuracy and rate) wa s administered individually to each participant to test reading fluency only. Both the child a nd the examiner were audio-recorded with an Olympus DS-40 Digital Voice Recorder for more accurate scoring procedure. Especially, recording was necessitated especially because of the decreased speech inte lligibility of hearing impaired children in fast passage reading task. The GORT-4 is designed for children aged 7;0 to 18;11 (in months). It consists of 13 passages. Each passage has one paragraph that is centered on a single topic. Across the test, there is an increase of length and complexity of paragraph, sentence, grammatical structures, and vocabulary content. The te st yields raw scores, st andard scores, percentiles, and grade-equivalent scores. The fluency assessment is a co mposite score of two components: a Rate (i.e., the time taken to read each passage) and an Accuracy (number of deviations from print). The mean for the

PAGE 81

81 two subtest components is 10, with a standard deviat ion of 3. They were instructed to read each story aloud as quickly and accurately as possible. Phonological Processing Skills Comprehensive Test o f Phonological Processing (CTOPP; Wagner, Torgesen, & Rashotte, 1999). Phonological coding is an oral language skill and consists of the analysis and synthesis of phonemes converted from visual s timuli of letters. Beginning readers who have deficits in phonological coding seem to have difficulty naming letter s of the alphabet, identifying sounds for alphabet letters, segmenting word s into phonemes and syllables, and applying knowledge of letter-sound correspondence to decode words (V ellutino, et al., 1996). For example, phonological coding involves analysis such as recognizing that the first sound of the word ball (/b /) can be replaced with /t / to produce the word tall Phonological coding abilities associated with this process of changing ball to tall include: (1) letter-sound correspondence, (2) phonemic awareness and segmentation, and (3 ) working with information in phonological working memory. For this study, six subtests of the CTOPP were used, including the Elision, Blending Words, Memory for Digits, N onword Repetition, Rapid Letter Naming and Rapid Digit Naming subtests (Form A). The CTOPP is an individually administered, norm-referenced measure that is used to evaluate a wide range of aspects of an indivi duals phonological processing. A three-part model, based on earlier studies in this area has been presented by the test developers (Torgesen & Wagner, 1998; Wagner & Torgesen, 1987). That is, three pairs of scores were combined to produce composite scores: Elision and Blending Words for Phonological Awareness composite Memory for Digits and Nonword Repetition for Phonological memory composite and Rapid Letter Naming and Rapid Digit Naming for Rapid Naming composite scores, respectively. The following is a short description of each of these components.

PAGE 82

82 Phonological awareness: analysis and synthesis of the sound structure of oral language. The order of progression of phonological awareness starts with syllables and moves toward smaller units of speech sounds (Adams 1990). Phonological awareness provides individuals with the ability to break words into sylla bles and component phonemes, to synthesize words from discrete sounds, and to learn about the distinctive features of words (Torgesen & Wagner, 1998). Phonological memory: coding information phonologically for temporary storage in working or short-term memory. Phonological sh ort-term memory invol ves storing distinct phonological features for short peri ods of time to be "read off" in the process of applying the alphabetic principle to word identification. Rapid naming: efficient retrieval of a series of names of objects, colo rs, digits, or letters from long-term memory. Rapid naming of verbal material is a measur e of the fluid access to verbal names, in isolation or as part of a series, and related efficiency in activating name codes from memory (Wagner, Torgesen, & Rashotte, 1999). The Elision subtest required participan ts to delete a phoneme in e ither the initial, final, or medial position from a real word and then produce a new word (e.g., Say tiger without saying /g/. [tire]). The Blending subtest required isolated syllables and phonemes to be combined into a word (e.g., What word do these sounds make: /h m/ /er/? [hammer]). In the Rapid Letter and Digit Naming tests, children are shown a visual displa y of randomly presented items and asked to name them in sequence as quickly as they can. The Rapid Digit Naming test uses numbers 2, 3, 4, 5, 7, and 8. Seventy-two numerals are presented on two pages in four rows, with a space after each. The time in seconds to name the 72 items in the display is recorde d. Standardized scores are provided for individuals ages 5 to 24 years on a s cale ranging from 0 to 20, with an average of 10. The authors were able to establish criter ion predictive validity with a sample that represented ethnic, gender, and ag e variations. Reliability of th e CTOPP was investigated using estimates of content sampling, time sampling, a nd scorer differences. Most of the average internal consistency or alternate forms reliability coefficients exceed .80 according to the test manual. The test/retest (time sampling) coeffici ents range from .70 to .92. It is known that the

PAGE 83

83 magnitude of the coefficients report ed from all the reliability studies suggests that there is limited error in the CTOPP and that examiners can ha ve confidence in the results. Reliability coefficients for the rapid-naming subtests are r .82 for letter and r .87 for digits. Standardized Oral Language Tests The Peabody Picture Vocabulary Test-III (L M. Dunn & Dunn, 1997) and Expressive Vocabulary Test (W illiams, l997) were administer ed to assess lexical knowledge and word retrieval. Both tests were administered accordi ng to the guidelines provided in the testing manual. The EVT and the PPVT-III were standardized on the same population of 2,725 examinees ranging in age from 2-6 to 90. All hard -of-hearing participants were wearing hearing aids for both ears and given the same oral inst ruction as hearing subjects. Throughout the data collection, no subjects had diffi culty understandi ng the task. Peabody Picture Vocabulary TestIII (PPVT-III, L. M. Dunn & Dunn, 1997): The PPVT-III is a standardized test of receptive lexical knowledge. Each easel page contains four pictures. Participants are require d to choose the picture drawing fr om four choices on a page that best depicts a word orally presented by the te st administrator. The normative sample included 2725 persons. And while the original PPVT wa s standardized only on white children from Tennessee, the normative sample of the PPVT-III was selected to match the data of the 1994 US Census. The sample was stratified with each age group by gender, race/ethnicity, geographic region, and SES. Only individua ls who were determined to speak and understand English were included in the testing. The alpha reliabilities for the 25 standardized age groups were between .92 and .98 with a median reliability of .95 for both forms A and B. The split -half reliabilities for the 25 age groups ranged from .86 to .97, with a median of .94 for both forms. The alternate forms reliabilities range from .88 to .96 with a median correlation of .94.

PAGE 84

84 The Expressive Vocabulary Test (EVT; Williams, l997 ): The EVT is a test of expressive vocabulary (lexical knowledge and word retrieval) requiring picture labeling and synonym tasks. For the 38 items, the tester points to a pictur e and asks a question. On the 152 synonym items, the examiner presents a picture and stimulus word(s) within a carrier phrase. The examinee responds to each item with a one-word answer. All stimulus pictures are in full color, carefully balanced for gender and ethnic representation. The EVT reliability analyses indicate a high degree of internal consistency. Sp lit-half reliabilities range from .83 to .97 with a median of .91. Alphas range from .90 to .98 with a median of .95. And test-retest studies w ith four separate age samples resulted in reliability coefficients rang ing from .77 to .90, indicating a strong degree of test score stability. The Comprehensive Assessment of Spoken Language (CASL, Carrrow-Woolfolk, 1999 ) was used to measure the knowledge of grammatical knowledge. The CASL is an individually and orally administ ered oral language assessment batt ery for ages 3 through 21. This test measures four main areas of oral language such as (a) lexical/ sema ntic, (b) syntactic, (c) supralinguistic, and (d) pragmatic sk ills. In the Syntactic Tests, five subtests are included: Syntax Construction, Paragraph Comp rehension, Grammatical Morphemes, Sentence Comprehension, and Grammaticality Judgment. For this study, the Grammatical Mo rphemes subtest was selected to investigate hearing impaired subjects' sens itivity to morphological markers and syntactic constituents for major grammatical information, including tense, voice, number, modality, person, pronoun, comparative, and lexico-conceptual know ledge. The examiner read one pair of words or phrases that demonstrated an analogy, then read the first word or phrase of a second pair. The examinee must complete the analogy of th e second pair (e.g., Skate is to skated, as talk is to _____.).

PAGE 85

85 The rationale for selecting this subtest is ba sed on some previous theoretical suggestion regarding phonetic sali ence of grammatical morphemes in English language. For example, Leonard (1998) suggested that many inflections in English have low phonetic salience and that this factor in combination with reduced sp eed of processing underl ies the problems with inflectional morphology described in SLI (Surf ace hypothesis). Even though the surface feature of phonetic salience alone cannot account for th e difference in difficulty (Hayiou-Thomas, Bishop, & Plunkett, 2004; Rice & Wexler, 1996), it is well assumed that peripherally depressed hearing acuity would be a negati ve factor for efficient and normal perception of inflectional markers which have low phonetic salience. T hus, we hypothesized that hearing-impaired children's performance on this specific test would show significant di fference from that of normally hearing subjects. Interrater Reliability To ensure test reliability, whenever any deviat ion from the protocol of test administration occurred, additional instruction was provided until that assistant was able to demonstrate stabilized skill. The test results and the b ackground data were fed into SPSS for Windows package 15.0. The first author and two research assi stants, who assisted in the test administration, score conversion and coding, met regularly to discuss any problematic matters. Interrater reliability procedures were conducted to determine the reliability of scoring and coding procedures. To determine interrater reliability at the end of data collection peri od, a trained reliability coder, a doctoral student of speech pathology, ch ecked on all scores (raw and standard scores) and coded data of 30 randomly selected particip ants (38% of the whol e subjects). This coder independently obtained childrens raw scores on all tests, conv erted them to standard and percentile scores, and checked the coded numbers in the SPSS. The reliability coder and the test

PAGE 86

86 administrator were blind to each others scori ng. The coder and the test administrator conducted an item-by-item comparison of their respective re sponses to each item administered in the battery. A reliability score was calculated for all variables by dividing the number of agreements by the number of disagreements plus agreemen ts and multiplying this score by 100. Interrater analysis showed 99% agreement for data codi ng (disagreement on 18 co ding errors out of 2,100 numeric codes) and 97% agreement for scor e conversion (disagreement on 12 out of 420 standard scores). Research Questions and Hypotheses The following three m ain categories of re search questions were investigated. Category I (Group Effect) Question : Are there any significant group effects on phonology, oral la nguage, and reading skills? Hypotheses : It was hypothesized that children with m ild to moderately severe SNHL would demonstrate significantly lower performance th an the comparison groups (NH and RD) on the measures of (a) auditory processing skills, (b ) phonological processing (phonological awareness, verbal memory, and rapid naming), (c) oral lang uage (receptive and expressive vocabulary and grammar) and (d) reading (word/nonword r eading, passage reading fluency, and comprehension). Category II (Relationships among Measures) Question : What are the interrelationship am ong r eading achievement, hearing ability, phonology, oral language skills, and audito ry processing skills? Which phonol ogical processes are the most strongly correlated w ith reading skills? Hypotheses: It was hypothesized that there would be significant association among variables of phonological processing, oral languag e, and reading skills of hear ing impaired subjects. This

PAGE 87

87 hypothesis was explored primarily using correlat ional analyses using a matrix of Pearson product-moment correlation coefficients between the variables. Category III (Regression Question) Question : What are the contributions of phonologica l and auditory processing skills in predicting hard-of-hearing child rens reading achievement? (1) Specifically, which measure of phonological proc essing skills is the best predictor of reading skills? (2) How much of unique variance of reading performance (reading comprehension, reading fluency, word/nonword reading) is expl ained by phonological processing measures? A full list of specific null hypotheses for the resear ch questions to be te sted in the study is presented in Table 3-3. Treatment of the Data All variables were entered into Microsoft Ex cel spreadsheets and analyzed with the SPSS 16.0. Descriptive statistics were calculated for all variables: dem ographic, audiologic, auditory processing, oral language, phonological processing, and r eading/spelling skills. Three statistical methods were used to analy ze the data. First, to measure group effect, a series of multivariate analysis of covariance (MANCOVA) tests was carried out on the scores of auditory processing, phonological processing, and r eading skills of the three groups (NH, HI, and RD) to investigate any significant group effect (Category I). A set of MANCOVAs was also conducted to compare oral language skills of childr en with normal and impaired hearing ability (NH and HI only). The grade scores (in months) and/or the non-verbal intelligence score (TONI3) served as the covariates or control variables. Secondly, to measure relationshi p among variables, a series of partial correlation analyses was conducted to determine significan t relationships among variables.

PAGE 88

88 Finally, to examine the unique variance explai ned selected explorat ory variables, block hierarchical regression analyses were performed to determine which predictor variables could most effectively account for variance in the following dependent variables (4 word/nonword reading measures; 1 reading fluency measures; and 1 reading comprehension). The predictor variables examined were: (1) b ackground information such as age, grade, and non-verbal IQ; (2) audibility, as measured by the pure-tone average; (3) auditory processing, as measured by the dichotic digits, and (4) six phonological pro cessing measures from the CTOPP. For all multivariate statistical tests, co rrelation analyses and regressions were conducted with an alpha level set to .05.

PAGE 89

89 Table 3-1. Matching variables (Grade, Age, Gender, and non-verbal intelligence) IQ Gradea Ageb Gender Boys Girls Mean (SD) Min Med Max Mean (SD) Min Max Mean (SD) Min Max Freq % Freq % NH 112.52 (15.844) 88 110 150 3.8 (1.15) 2.1 5.9 111.9 (14.0) 89 134 16 55.2 13 44.8 HI 100.89 (12.918) 84 110 138 3.4 (1.61) 1.0 6.9 110.8 (19.3) 82 152 11 57.9 8 42.1 RD No data 3.96 (1.62) 1.2 6.8 116.6 (20.3) 86 152 18 60.0 12 40.0 Total 107.92 (84-150) 84 107.9 150 49.1 (18.9) 12 92 113.4 (17.9) 82 152 45 57.7 33 42.3 Note: a: in year.months, b: in months, Freq: frequency, %: percentile

PAGE 90

90 Table 3-2. List of tests used. Area of measurement Tests Contents Non-verbal intelligence TONI-3 Non-verbal IQ Oral language Receptive vocabulary PPVT-III Receptive vocabulary Expressive vocabulary EVT Expressive vocabulary Syntactic structure CASL ( Grammatic Knowledge subtest) Grammatic knowledge Phonological processing CTOPP) Phonological awareness Elision, Blending Phonological memory Memory for Digits, Nonword Repetition, Short-term memory Rapid naming RAN-Digit, RAN-Letter Lexical access and retrieval Literacy WRMT-R ( Word Identification ) Word Reading Reading (Untimed) WRMT-R ( Word Attack ) Nonword decoding TOWRE ( Word Efficiency ) Word reading Reading (Timed) TOWRE ( Phonemic decoding efficiency ) Nonword decoding Spelling WRAT-3 ( Spelling) Spelling words Reading comprehension WRMT-R ( Passage Comprehension ) Silent passage comprehension Reading fluency GORT-4 ( Reading accuracy and rate ) Passage reading fluency Abbreviations: TONI (Test of Nonverbal Intelligence), PPVT-III (Peabody Picture Vocabulary Test), EVT (Expressive Vocabul ary Test), CASL (Comprehensi ve Assessment of Spoken Language),CTOPP (Comprehensive Test of Phonological Processing ), WRMT-R (Woodcock Reading Mastery Test Revised), TOWRE (Test of Word Reading Efficiency), WRAT-3 ( Wide Range Achievement Test), GORT-4 (Gray Oral Reading Tests).

PAGE 91

91Table 3-3. List of research hypotheses. Statistical analyses Hypotheses 1. There will be statistically signi ficant differences on a measure of receptive vocabulary (PPVT-3) among students in the NH and HI groups. 2. There will be statistically signi ficant differences on a measure of expressive vocabulary (EVT) among students in the NH and HI groups. 3. There will be statistically signi ficant differences on a measure of grammatical knowledge (CASL) among students in the NH and HI groups. 4. There will be statistically signifi cant differences on the measures of phonemic elision, deletion, rapid naming (letter and digit), and pho nological short-term memory among st udents in the NH, HI, and RD groups (CTOPP). 5. There will be statistically signifi cant differences in untimed word and nonword reading skills (WRMT-R) among students in the NH, HI, and RD groups. 6. There will be statistically signifi cant differences in timed word a nd nonword reading skills (TOWRE) among students in the NH, HI, and RD groups. 7. There will be statistically significant differences in spelling skill (WRA T) among students in the NH, HI, and RD groups. 8. There will be statistically signi ficant differences in passage read ing comprehension (WRMT-R) among students in the NH, HI, and RD groups. MANCOVA 9. There will be statistically signifi cant differences in passage reading accuracy and rate (GORT-4) among students in the NH, HI, and RD groups. 1. There will be significant correlations between hear ing-impaired students de pressed phonology and reading. 2. There will be significant correlations between hear ing-impaired students de pressed phonology and hearing loss. 3. There will be significant correlations between hear ing-impaired students de pressed phonology and oral languages. 4. There will be significant correlations between hearing-impaired students auditory processing and reading. Correlation 5. There will be significant correlations between hear ing-impaired students a uditory processing and phonology. Regression 1. Phonological awareness will sign ificantly predict unique variance in word-level reading after controlling for cognitive ability, hearing acu ity, and grade in both untimed and timed reading tasks.

PAGE 92

92Table 3-3. Continued 2. Phonological short-term memory will significantly pred ict unique variance in wo rd-level reading after controlling for cognitive ability, he aring acuity, and grade in both un timed and timed reading tasks. 3. Rapid naming will significantly predict unique variance in word-level reading after controlling for cognitive ability, hearing acuity, and grade in bot h untimed and timed reading tasks. 4. Auditory processing will significantly predict unique va riance in word-level reading after controlling for cognitive ability, hearing acu ity, and grade in both untimed and timed reading tasks. 5. Phonological awareness will si gnificantly predict unique variance in pa ssage reading fluency after controlling for cognitive ability, hearing acuity, and grad e in both untimed and timed reading tasks. 6. Phonological short-term memory will significantly predic t unique variance in passage reading fluency after controlling for cognitive ability, he aring acuity, and grade in both un timed and timed reading tasks. 7. Rapid naming will significantly predict unique variance in passage reading fluency after controlling for cognitive ability, hearing acu ity, and grade in both untimed and timed reading tasks. 8. Auditory processing will significantly predict unique vari ance in passage reading fl uency after controlling for cognitive ability, hearing acu ity, and grade in both untimed and timed reading tasks. 9. Phonological awareness will si gnificantly predict unique variance in spelling after controlling for cognitive ability, hearing acuity, and grade in bot h untimed and timed reading tasks. 10. Phonological short-term memory will significantly predic t unique variance in spel ling after controlling for cognitive ability, hearing acu ity, and grade in both untimed and timed reading tasks. 11. Rapid naming will significantly predict unique variance in word-level in spelling after controlling for cognitive ability, hearing acu ity, and grade in both untimed and timed reading tasks. 12. Auditory processing will significantly predict unique vari ance in word-level in sp elling after controlling for cognitive ability, hearing acu ity, and grade in both untimed and timed reading tasks.

PAGE 93

93 CHAPTER 4 RESULTS The prim ary goal of this study was to exam ine the effect of congenital mild to moderate SNHL on childrens auditory processing, phonological processing, oral language, and literacy skills. The main focus of this investigation was twofold: (1) to examine differences among normally developi ng, hearing impaired students, and dyslexic readers in oral language, phonology, and read ing skills (group effect) and (2) to investigate interrelationshi ps between reading skills and phonology, auditory processing, and oral language abilities. The following questions were formulated and tested as follows. Are there any significant group effects fo r phonology, oral language, and literacy skills (reading/spelling) be tween the three groups? What interrelationships exist between reading and phonological skills, auditory processing and oral language ability in hard-of-hearing children? Do phonological processing skills make sign ificant contributions to the literacy skills of children with hearing loss? This chapter has four subsections. The first section presents results of the descriptive statistics for all of the variables investigated in this study. The second section deals with the group effect on oral language, phonology, audiologic function, and reading using a set of multivariate analyses of covariance. The third section provides the results of correlation analyses among variables based on the P earson-product correlations. Descriptive statistics and firs t-order bivariate and partial co rrelation coefficients were used to estimate the childrens performan ce in the measures and interrelationships among their performance in different measures. The final section investigates the results of multiple hierarchical regression analyses. The implications of these findings are discussed in Chapter 5.

PAGE 94

94 Descriptive Statistics Prelim inary analyses were conducted to obtain descriptive information on the samples demographic characteristics, audiologic ability, pho nological process skills, oral language skills, and reading skills. Results of the m easures on phonology, language, and reading are displayed in the next secti on where the outcomes of the MANCOVAs are reported. Demographic Data The dem ographic characteristics of the pa rticipants in the three groups, including age grade, non-verbal IQ, and gender are provided in Chapter 3 (See Table 3-1). Audiologic Ability Measures Inf ormation related to the hard-of-hearing childrens audiologic characteristics is displayed in Table 4-1. Demographic informa tion was collated and included age, grade, gender, information about the history of audi ologic invention (age at identification of hearing loss age at initial h earing aid fitting and length of hearing aid fitting), etiology, and basic audiometric results (unaided PTA, SRT, and WRS). A subject number was randomly assigned to each participant to ensure patient confidentiality. Table 4-2 is a descriptive summary of the HI and NH groups performance on the basic audiologic measures. The HI groups unaided PTA on the frequencies of 0,5 Hz, 1KHz, and 2 KHz ranged from 27 to 69 dB HL (mean better ears PTA = 46.9 dB HL). Table 4-3 reports mean PTA of the HI and NH groups for every tested frequency. Figures 4-1 through 4-3 are graphical re presentations of this inform ation. Figure 4-1 and 4-2 are error bar charts for each ear based on the st andard errors of each mean threshold across tested frequencies, showing that, on aver age, the HI group had a symmetrical sloping binaural SNHL, with substantial variance wi thin the group. Figure 43 is a histogram of

PAGE 95

95 better ears pure-tone threshold for the HI group. Table 4-4 displays the means and standard deviations by heari ng group (NH, HI) for the thr ee subtests of the SCAN-A/C and the Dichotic Digits subtest of the CAPD test (Musiek, 1983). Oral Language, Phonology, and Reading Descriptive data of oral language, phonol ogical processing, and reading m easures are provided in the next section dealing with the MANCOVA questions. Inferential Statistics All scores u sed in the statistical analyses were norm-referenced standard scores. Multivariate analyses of covariance (MANCOVA), Pearson-product correlation analyses, and hierarchical block regressi on analyses were used to analyze the data to answer the research questions raised in the previous chapter. The resu lts are organized to provide answers to three questions below: 1. What is the effect of decreased audito ry sensitivity due to congenital SNHL on phonology, oral language, and reading skills? That is, is ther e any significant group effect in each measured area? 2. What is the relationship among hard-o f-hearing readers impaired phonology, reading, oral language, and hearing loss Question Category I: Group Comparison s of Language, Phonology, and Reading To com pare the performance of the HI w ith the two control groups (NH and RD), multivariate analyses of covariance (MANCOV A) with grade (in months) and/or nonverbal intelligence as covariates were used for oral language, phonology, and reading measures. MANCOVA was selected in order to guard against Type I errors due to multiple univariate testing for each area. Three separate MANCOVAs were conducte d: (a) For oral language measures, three dependent variables from the PPVT-III (receptive vocabulary), EVT (expressive

PAGE 96

96 vocabulary), and a subtest of Grammatical Knowledg e from the CASL were analyzed (NH and HI groups only), (b) for the phonologica l tests, six subtests from the CTOPP were analyzed (NH, HI, and RD groups), and (c) thirdly, for the reading measures, standard scores of timed word/nonword reading (TOWRE), untimed word/nonword reading (WRMT-R), reading comprehensi on (WRMT-R), passage reading fluency (GORT-4), and spelling (WRAT) tests were included in the MANCOVA to investigate between-group difference on the literacy skills. MANCOVA is appropriate for use with multip le dependent variables that differ in scales of measurement and fewer Type I erro rs occur with it than with a univariate analysis of variance (Gabriel & Hopkins, 1974). Other than maintaining overall Type I error at a constant level, MANCOVA also allows us to control the systematic group difference caused by covariates Covariates are variables th at are correlated with the dependent variables and are used to adjust for any differences in the scores affected by extraneous confounding factors. By usi ng MANCOVAs, the effects of confounding variables such as non-verbal IQ age, or grade, can be sta tistically removed, helping to ensure that the findings refl ect true differences in phonology, oral language, and reading skills. For the first MANCOVA, childrens non-verb al intelligence and grade were used as covariate variables. For the second and third MANCOVAs, only grade (in months) was used as a covariate variable because the TONI scores were not available in dyslexic participants. Hair et al. (1992) recommends that Wilk's la mbda is the best statistical measure to assess whether an overall significant differe nce is found between groups. All multivariate

PAGE 97

97 statistical analyses were conducted with an alpha level set to .05. Good approximations for significance can be obtained from Wilk's lambda and can be transformed into an Fstatistic. If any significant differences are identified in MANCOVAs, univariate analyses of covariance (ANCOVAs) are justified to i nvestigate the direction and significance of specific dependent measures. ANCOVAs are run on the residuals after predictions of the dependent variable have been made from a set of covariates. In this type of analysis, a predicted dependent variable is computed. Th is predicted score is subtracted from the original dependent variable score. This di fference or residual score is used in the ANCOVA (Meyer, 1993). Effect sizes for significant differenc es between groups, adjusting for covariates, are calculated as the difference between the estimated adjusted means for the two groups, divided by the ro ot-mean-square error and reported as Cohen s d. Cohen (1988) designated an effect size of 0.2 as small, 0.5 as medium, and 0.8 as large. For all ANCOVAs showing significant group effects, pairwise comparisons of groups are conducted. As a precaution against making Type I errors, the p value is adjusted with the Bonferroni method fo r all multiple pairwise comparisons. MANCOVA for oral language measures Perform ance on the oral language tests for the two groups of children (NH and HI groups) was compared to determine if differe nces in group performances were significant for overall measures of expressive/recepti ve vocabulary and grammatical knowledge. A one-way multivariate analysis of covariance (MANCOVA) was used with group membership (NH vs. HI) as the independent va riable, standard scores from the PPVT-III, EVT, and CASL tests as the dependent vari ables, and grade (in months) and non-verbal intelligence as the covariates.

PAGE 98

98 Two assumptions of the MANCOVA were tests: equal covariance and equal variance. MANCOVA assumes that for each group the covariance matrix is similar. Box's M statistic was used and the covari ance assumption was not violated (F[6, 9896.2]) = 4.356, p = .673). MANCOVA also assumes that each dependent variable will have similar variances for all groups. According to the Levene's test, all three measures had equal variances in the two groups (PPVT-III: F[(1,46] = .325, p = .571; EVT: F[1,46] = 2.179, p = .147; and CASL: F[1,46] = 2.850, p =.098). Results of the initial multivariate test using Wilks Lambda ( ) criterion indicated that the groups differed signi ficantly in overall oral langu age measures (F[3,42] = 10.34, p = .000, partial 2 = .425, and observed power = .997). Because significant differences between groups emerged for overall measures, it was necessary to examine specifically which of the tests were influenced across the two groups. That is, each of the oral language variables was examined indivi dually through the use of three one-way univariate ANCOVAs. These follow-up univariat e analyses indicated that differences between groups were significant for all language measures: (a) receptive vocabulary: F[1,44] = 16.123, p = .000, partial 2 = .268, observed power = .975; (b) expressive vocabulary: F[1,44] = 8.694, p = .005, partial 2 = .165, observed power = .822; and (c) grammatical knowledge: F[1,44] = 28.354, p = .000, partial 2 = .392, observed power = .999. Descriptive statistics for oral language measures along with estimated adjusted means and standard errors are reported in Table 4-5. Table 4-6 is the ANOVA table for these ANCOVAs displaying the result of univari ate F-tests for between-group differences.

PAGE 99

99 Table 4-7 is a summary of pa irwise post-hoc comparisons. Fi nally, Figure 4-4 displays clustered box plots of oral language measures of the groups (NH, HI). Pairwise comparison revealed that when controlling for non-ver bal IQ and grade, the hearing impaired group performed signi ficantly below the normal group both on the vocabulary and grammatical knowledge. Specifica lly, the differences of adjusted means of the PPVT-III and Grammatic Knowledge subset of the CASL test was 16.651 and 18.769, respectively, which were larg er than one standard deviation (15.0). The EVT also showed substantial difference of 10.963. Notabl y, as seen in the adjustment mean difference, the effect size was biggest on th e measure of grammatical knowledge (partial 2 = .392). MANCOVA for phonology In order to determ ine whether there was a significant difference between the three study groups (NH, HI, and RD) on phonologica l processing measures, six dependent variables from the CTOPP representing components of phonological awareness, phonological memory, and rapid naming skills were submitted to multivariate analysis of covariance. MANCOVA was conducted with group membership as the independent variable and grade in months as the covariate. Box's M statistic was used to check th e covariance assumption, which was not violated: F[6, 11817.457] = 9896, p = .831), suggesting that the observed covariance matrices of the dependent variables were e qual across groups. The Levene's test was used to test the equal variance assumption. All six measures had equal variances in the three groups ( df1 = 2, df2 = 75): (1) Elision : F = .272, p = .763; (2) Blending : F = 2.102, p = .129; (3) RAN-Digit : F = .433, p = .650; (4) RAN-Letter: F = .005, p = .995; (5) Memory for digit: F = .202, p = .818; and (6) Nonword repetition: F = 1.344, p = .267.

PAGE 100

100 The resulting MANCOVA was significant, yielding a main effect for group, F(12, 138) = 8.450, p <.001, showing a significant difference between the groups overall performance on the phonological processing measures. The observed means, the observed standard deviations, and means adju sted after the statistical removal of the grade effect, standard errors and observed powers for three groups appear in Table 4-8. Because of this significant difference in ove rall measures, a series of follow-up one-way univariate ANCOVAs were conducted. Differences between groups were signif icant for every phonological measure after adjusting for grade in month ( df1 = 2, df2 = 74): (1) Elision (F = 26.803, p = .000, partial 2 = .420, observed power = 1.00); (b) Blending (F = 20.281, p = .000, partial 2 = .354, observed power = 1.000); (c ) RAN for Digit (F = 13.767, p = .000, partial 2 = .271, observed power = .998); (d) RAN for Letter (F = 13.584, p = .005, partial 2 = .269, observed power = .997); (e) Memory for Digits (F = 4.993, p = .009, partial 2 = .119, observed power = .798); and (f) Nonword Repetition (F = 25.78, p = .000, partial 2 = .411, observed power = 1.00). The effect sizes for the Blending and Elision subtests were moderately large ( Elision = .420, Blending = .354). In contrast, the effect sizes for rapid naming skill were smaller than those for phonological awaren ess skills (RAN-Dig it = .271 and RAN-Letter = .269). A summary of six univariate ANC OVAs is shown in Table 4-9. For all phonological processing measur e showing signifi cant between-group differences, a series of post hoc pairwise comparisons for the three groups (i.e., NH vs. HI, HI vs. RD, and NH vs. RD) based on B onferroni adjustment was conducted to investigate how the groups are different acro ss each of the dependent variables. Results

PAGE 101

101 are displayed in Table 4-10. Consistent with the previous literature (Cornwall, 1992; Cronin & Carver, 1998; Snyder & Downey, 1995; Wolf & Bowers, 1999; Wolf et al., 2002), on average, dyslexic childr en in this study were seve rely impaired on all reading measures when compared to the normal controls. Adjusted mean differences (NH minus RD) and t -scores on each measure were as follows: (1) Elision: 4.593, t(57) = 6.917 p = .000; (2) Blending: 2.645, t(57) = 4.401, p = .000; (3) RAN-Digit : 3.043, t(57) = 5.247 p = .000; (4) RAN-Letter: 3.281, t(57) = 5.208, p = .000; (5) Memory for digits: 1.999, t(57) = 2.750, p = .022; and (6) Nonword repetition :3.577, t(57) = 5.478, p = .000. Similarly, HI subjects phonological aw areness and phonological memory skills were significantly lower than in the NH gr oup. Adjusted mean differences (NH minus HI) and t-scores on each measure were as follows: (1) Elision: 3.500, t(46) = 4.888, p = .000; (2) Blending: 3.849, t(46) = 5.949, p = .000; (3) Memory for digits: 1.953, t(46) = 2.494, p = .044; and (4) Nonword repetition: 4.496, t(46) = 6.395, p = .000. Unexpectedly, however, no significant differenc es in the subtests of rapid naming were reported between the NH and HI groups. Adjust ed mean differences (NH minus HI) and t-scores on rapid naming measur es were as follows: (1) RAN-Digit : 1.302, t(46) = 2.083, p = .122 and (2) RAN-Letter : 1.245, t(46) = 1.836, p = .211. Hence, these data have shown that hear ing impaired students lexical access and retrieval skills are well preserved. Additionally, our post hoc comparisons indicated that the HI group performed significantly better than the RD group on the two RAN subtests. While the HI group did not differ from th e RD group on tasks of phonological awareness and phonological memory, when adjusted means (HI minus RD) for the Nonword Repetition and Blending subtests were used, RD subjects scored slightly higher with no

PAGE 102

102 significant difference. Adjusted mean, t-scores, and associated p-values on each measure were as follows: (1) RAN-Digi t: 1.741, t(47) = 2.568, p = .033; (2) RAN-Letter: 2.035, t(47) = 2.803, p = .019; (3) Elision: 1.093, t(47) = 1.427, p = .473 (ns); (4) Blending : 1.204, t(47) = 1.737, p = .260 (ns); (5) Memory for digits: .047, t(47) = .056, p = 1.00 (ns); and (6) Nonword repetition: -.919, t(47) = 1.220, p = .679 (ns). Overall, the HI group showed depressed phonol ogical processing skills only in the areas of phonological awareness and phonologi cal memory. No differences existed between the normal controls and the HI gr oups on the measures on rapid naming. In contrast, the RD subjects in this study showed deficiencies in all three phonological processing components (phonological awar eness, phonological memory, and rapid naming). Descriptive results are displayed gr aphically in Figures 4-5 through 4-7 as clustered boxplots for each phonological component. MANCOVA for literacy measures (reading and spelling) For the reading and spelling m easures a one-way MANCOVA was used with group membership (NH, HI, and RD) as the inde pendent variable and grade in months as the covariate. Three out of eight dependent variables had statistically significant inequalities of variance based on the Levene's te st for equality of variance (p < 0.05): (1) spelling: F[2,75] = 3.782, p = .027; (2) timed word read ing (TOWRE): F[2,75] = 4.275, p = .017; and (3) passage reading rate: F[2,75] = 3.834, p = .026. However, according to Hair et al. (1992), a violation of this assumption has minimal impact if the groups (HI, NH, and RD) are approximately of equal size or if the largest group size divide d by the smallest group size is less than 1.5. The ratio of dyslexic group size (n = 30) to the group of hearing impaired students (n = 19) was near to 1.50 (ratio = 1.579), hence, any violation of this

PAGE 103

103 assumption should have minimal impact. In light of the above findings, it is not surprising that the Box's M-test for multivariate homoscedasticity indicated significant differences as well (F = 1.544, df1 = 72, df2 = 11146.98, p = .002). This test is "notoriously sensitive" (Tab achnick and Fidell, 1996, p. 382), particularly given the large sample size involved. The group sizes in our da ta are not large and relatively equal to each other, so the impact should be mini mal; all analyses of variance have been conducted assuming unique variance between groups. The multivariate analysis results showed a significant difference between the three groups, when MANCOVA was carried out for the entire measures of reading and spelling (Wilks Lambda = 0.228, F[16, 134] = 9.179, partial 2 = .523, and p <.001). This significant differences in the analyses for overall tests justified further investigation of the significance of each dependent measure usi ng univariate tests of ANCOVA. ANCOVAs showed significant group effects for all seven measures: (1) untimed word reading ( F [2,74] = 58.610 p < .001); (2) untimed nonword reading ( F [2,74] = 42.430 p < .001); (3) passage comprehension ( F [2,74] = 35.084, p < .001); (4) spelling ( F [2,74] = 40.899, p < .001); (5) timed word reading timed ( F [2,74] = 68.634, p < .001); (6) timed nonword reading ( F [2,74] = 73.415, p < .001); (7) passage reading rate ( F [2,74] = 77.143 p < .001); and (8) passage reading accuracy ( F [2,74] = 59.777, p < .001). Descriptive statistics such as means, st andard deviations, and estimated adjusted means for all tasks are displayed in Table 4-11. Error bar charts and box plots based on these descriptive data are also seen in Fi gures 4-7 through 4-10. Significant group effects based on eight univariate ANCOVAs for a ll measures and the corresponding ANOVA tables are seen in Table 4-12 and Table 4-13, respectively.

PAGE 104

104 Next, pairwise comparisons using Bonf erronis adjustment were conducted. Inconsistent with the previous literature (Briscoe et al., 2001; Gibbs, 2004), on average, reading and spelling skills of hearing impair ed children in this study were significantly lower than normal controls on all reading m easures (HI vs. NH) with one exception of the Word Decoding subtest on the TOWRE. Adjusted m ean differences (NH minus HI) and t-scores on each measure were as follows: (1) untimed word reading: 14.386, t(46) = 3.75, p = .001; (2) untimed nonword reading: 10.003, t(46) = 2.65, p = .029; (3) passage comprehension: 12.058, t(46) = 3.40, p = .003; (4) spelling: 13.222, t(46) = 3.72, p = .001; (5) timed word reading: 7.873 t(46) = 2.46, p = .048; (6) timed nonword reading: 6.155, t(46) = 1.90 p = .185; (7) passage readin g rate: 2.664, t(46) = 3.14, p = .007; and (8) passage reading accuracy: 2.883, t(46) = 3.10, p = .007. Descriptively, adjusted mean differences in timed word/nonword reading were smaller than untimed word-level reading, suggesting a contributive role of hearing-impaired readers preserved lexical access skills to fast reading. In addition, as expected, every mean score on the tests of reading were low average for the RD group and the HI group in our sample was significantly better than the RD controls on all measures. Remarkably, in tim ed measures, mean differences were more than 1.5 to 2 standard deviations. Adjusted mean differences (NH minus RD) and t-scores on each measure were as follows: (1) untimed word reading: 21.851, t(47) = 5.71, p = .000 ; (2) untimed nonword reading: 20.091, t(47) = 5.34, p = .000; (3) passage comprehension: 14.018, t(47) = 3.96, p = .001; (4) spelling: 15.013, t(47) = 15.013, p = 4.23; p = .000; (5) timed word reading timed: 24.020, t(47) = 7.53, p = .000; (6) timed nonword reading: 26.832, t(47) = 8.28, p = .000; (7) passage reading rate: 6.394, t(47) =

PAGE 105

105 7.56, p = .000; and (8) passage reading accuracy: 5.782, t(47) = 6.3, p = .000. Table 4-13 is a summary of pairwise pos t-hoc comparisons. Lastly, RD groups reading and spelling scores were significantly lower than normal groups for all measures. Question Category II: Correlations among Measures Our second category of research questions involved the inte rrelationship between hearing im paired participants literacy skills and other expl oratory variables, including oral language, phonological processing, and audito ry processing ability: The first step in this analysis was to explore the pattern of associations in correlation analyses. Correlational analyses are presented fo r the research questions below: 1. Question 1: What is the relations hip between phonology and reading? 2. Question 2: What is the relationshi p between phonology and hearing loss? 3. Question 3: What is the relationshi p between phonology and oral language? 4. Question 4: What is the relationship be tween phonology and aud itory processing? 5. Question 5: What is the relationship between reading, auditory processing? Question 1: Phonology and reading To assess th e independent associations of each of the measures on phonological processing and reading skills, partial correlati on coefficients were calculated with the effects of age, non-verbal intelligence (TON I-3), and hearing loss removed. Hearing loss, as measured by the better ears pure-tone, was also removed from the correlations because the degree of hearing loss exhibited strong negative relations with phonological awareness when age and grade in months were partialed out ( Elision : r = -.631, p = .007; Blending : r = -.728, p = .001; See Table 4-), s hort-term memory span ( Nonword repetition : r = -.608, p = .010). Partial coefficients were computed for phonological processing and reading measures. The partia l correlation matrices for the HI and NH separately are shown in Table 4-15 and Table 4-16. The results will be reported separately for the HI and NH groups.

PAGE 106

106 Hearing-impaired group : As can be seen in Table 4-15, the Elision subtest was more highly correlated with reading measur es than any other phonological processing variables. It was the only measure of phonological awaren ess which was significantly associated with all reading and spelling measures in both word and passage levels. Correlation coefficients between Elision and reading measures ranged from .562 to .751 as seen below: (1) untim ed word reading (r = .582, p < .01); (2) untimed nonword reading (r = .751, p < .001); (3) timed word reading (r = .562, p < .05); (4) timed nonword reading (r = .611, p < .05); (5) spelling (r = .618, p < .01); (6) passage comprehension (r = .610, p < .01); (7) passage r eading rate (r = .605, p < .01); and (8) passage reading accuracy (r = .711, p < .001). Even though the Blending is a measure of phonological awareness, it was not correlated with any word-level reading measures, spelling, and comprehension. It was only significantly associ ated with oral passage reading rate (r = .470, p < .05). The pattern of correlations between rapid naming skills and reading measures was remarkable. Both RAN measures did not s how any significant correlations with untimed tasks of reading at the word or passage level. Correlation coefficients among these variables were as follows: (1) untimed word reading (RAN-D: r = .307, p = .124; RANL: r = .259, p = .166); (2) untimed nonword reading (RAN-D: r = .411, p = .057; RAN-L: r = .301, p = .129); and (3) passage co mprehension (RAN-D: r = .341, p = .098; RAN-L: r = .314, p = .118). Notably, however, significant correlations were found for rapid naming measures and timed reading tests at both word and passage levels: (1) Timed word reading (RAN-D: r = .459, p = .037; RAN-L: r = .559, p = .012); (2) timed nonword reading (RAN-D: r = .560, p = .012; RAN-L: r = .656, p = .003); (3) timed passage

PAGE 107

107 reading rate (RAN-D: r = .723, p = .001; RAN-L: r = .581, p = .009); and (4) timed passage reading accuracy (RAN-D: r = .635, p = .004; RAN-L: r = .571, p = .003). Correlation coefficients ranged from .459 to .656 at word level and from .571 to .723 at the passage level. All correlations in the partial correlation matrix were higher than medium (about .3) or large (above .5 ) correlations (Cohen, 1988). The phonological memory tasks as measured by the Nonword Repetition task was correlated with untimed word reading (r = .594, p = .008). With other literacy measures, phonological STM tasks approached significant le vel (See Table 4-15). Normally hearing controls : As can be seen in Table 4-16, after the effects of age and non-verbal intelligence had been contro lled, phonological awareness variables were highly correlated with reading a nd spelling variables: (1) unti med word reading (elision: r = .711, p < .001; blending: r = .378, p < .05), (2) untimed nonwor d reading (elision: r = .692; p < .001; blending: r = .472, p < .05), (3) spelling (r = .825, p < .001), (4) passage comprehension (elision: r = .520, p <.05), (5) timed word reading (r = .474, p < .001), (6) passage reading rate (elision: r = .443, p < .05), and (7) passag e reading accuracy (r = .671, p < .001). Highest correlations were observed between the Elision and spelling tests. Quite differently from the HI group, si gnificant small to medium intercorrelations were found between rapid naming and untimed word reading tests for the NH subjects (word reading (RAN-D: r = .421, p = .016; RAN-L: r = .414, p = .018); (2) untimed nonword reading (RAN-D: r = .382, p = .027; RAN-L: r = .437, p = .013). Also, similar to the HI group, high correlation coefficients were found in the NH group between rapid naming tasks and timed reading tests for bot h word and passage le vel: (1) timed word reading (RAN-D: r = .559, p = .001; RAN-L: r = .542, p = .002); (2) timed nonword

PAGE 108

108 reading (RAN-D: r = .596, p = .001; RAN-L: r = .629, p = .000); (3) timed passage reading rate (RAN-D: ; r = .382, p = .027; RAN-L: r = .385, p = .026); and (4) timed passage reading accuracy (RAN-D: r = .465, p = .008; RAN-L: r = .540, p = .002). Another notable correlation pattern was found for phonological s hort-term memory measures. Different from the HI group, the tw o untimed word-level reading and spelling measures were highly correlated in the NH gr oup. Partial coefficients are as follows: (1) Memory for Digits : (untimed word reading: r = .657, p = .000; untimed nonword reading: r = .603, p = .001; and spelling: r = .519, p = .003) and (2) Nonword repetition: (untimed word reading: r = .686, p = .000; untimed nonword reading: r = .650, p = .000; and spelling: r = .534, p = .002). Question 2: Phonology and hearing loss The partial correlation matrix adjus ting for the variance accounted for by age and non-verbal IQ for the HI group (n=19) are shown in Table 4-17. Table 4-17 includes the six measures of phonological measures and variables related to heari ng loss such as better ears pure-tone, age at identification, age at initial hearing ai d fitting, and length of hearing aid use. As expected, the better ears PTA was correlated with phonological awareness and short-term memory span as s een in the following partial coefficients: (1) Elision (r = -.473, p = .024); (2) Blending (r = -.597, p = .004); (3) Nonword repetition (r = -.424, p = .040), underscoring the effect of decreased auditory input on metaphonological skills. On the other hand, the RAN tasks (letters and digits) were not correlated with the severity of hearing loss. This finding was also supported by the MANCOVA results, which revealed no effect of hearing loss on rapid naming skills.

PAGE 109

109 Question 3: Phonology and oral language It was also questioned whether lexica l and gramm atical knowledge would be related to phonological processing skills which measures children's ability to encode, store, and retrieve phonological informa tion. The correlation and partial correlation matrices for the HI group are shown in Tabl e 4-18. The results revealed that phonological short-term memory (nonword repetition, memo ry for digits) were strong correlates of both receptive and expressive vocabulary, and grammatical knowledge. Notably, the Memory for digit subtest was highly correlated with all language measures: (1) PPVT-III (r = .533, p = .014), (2) EVT (r = .592, p = .006), and (3) CASL (r = 558, p = .010). The Nonword repetition was also moderately correlated with the receptive and expressive vocabulary, but only approached the significance level: (1) PPVT-III (r = .406, p = .053) and (2) EVT (r = .359, p = .078). In contrast, no significant association was found between the RAN tasks and any of oral language measures. Similarly, phonologi cal awareness tasks did not show any correlations with vocabulary tests. However, a pattern of significant correlations was seen between phonological awareness and grammatical knowledge test revealing coefficients higher than medium level: (1) Elision (r = .511, p = .018) and (2) Blending (r = .443, p = .038). Question 4: Auditory processing, phonology, and reading measures Table 4-19 shows correlations am ong the auditory and phonological measures and reading for the HI group, with partial correlations controlling for age, grade, duration of hearing aid use, and nonverbal intelligence. Auditory processing skills measur ed by the three SCAN subtests ( Filtered Words, Auditory Figure-Ground, and Competing Words) had no significant correlations with any

PAGE 110

110 of phonological awareness and rapi d naming tasks. However, a high degree of significant associations were found between the Competing Words subtest and short-term memory span tasks (i.e., Memory for Digits : r = .653, p = .008; Nonword Repetition : r = .579, p = .024). The Competing Words subtest is a test of dicho tic listening, where the subject hears two words simultaneously, one word presented to each ear. The subject is instructed to repeat the words presented in each ear. For a better performance, the two stimuli presented simultaneously should be e fficiently stored in phonological buffer for later repetition. Thus, the high le vel of association be tween this type of dichotic listening test and short-term memory span was not surprising. Similarly, high correlations were found between the Dichotic Digit test and short-term memory span, an en expected results based on the previous explanation (i.e., Memory for Digits : r = .617, p = .014; Nonword Repetition: r = .548, p = .034). None of the SCAN subtests showed signi ficant correlations with any of reading measures, including word-level reading and passage reading and spelling. However, the Dichotic Digit task correlated with the two nonword reading tasks (untimed nonword: r = .546, p = .035 and timed nonword: r = .525, p = .044). Question 5: Reading and oral language Partial co rrelation coefficients were calcula ted between all of reading measures and oral language scores with the effects of ag e, grade, nonverbal intelligence, and the degree of hearing loss partialed out. Table 4-20 displays the correlations matrix. Passage comprehension was the only reading tests th at correlated with the oral language measures: (1) PPVT-III (r = .474, p = .037), (2) EVT (r = .536, p = .020), and (3) grammatical knowledge (r = .442, p = .049). Interestingly, ev en though the GORT test

PAGE 111

111 measured reading fluency in passage level, no correlations were found between vocabulary and the passage reading fluency. Question Category III: Explorat ory Hierarchical Regressions Although MANCOVA procedures can test fo r the statistical significance of group differences in one or m ore domains, these pr ocedures do not necessa rily provide a clear indication of how hard-ofhearing readers depre ssed phonological skills are predictive of reading skills across wide range of domains. In an effort to explore the relations between the reading measures and the exploratory variables, includi ng phonological and auditory pr ocessing abilities, selected based on the previous partial correlation analys es, a series of multivariate analyses (i.e., hierarchically blocked regressions) was perf ormed separately for each reading measure. According to Whitley (1996), there are thr ee different forms of multiple regression analysis: simultaneous (enter), stepwise, and hi erarchical and each of them has a specific purpose. Hierarchical regr ession should be used for hypothesis testing while simultaneous and stepwise regressions shoul d be used only for simple prediction. In this study, specific hypotheses were test ed regarding the significance of selected exploratory variables, so the use of hierarchical regression model was justified. Findings from multiple regression analyses allowed for examination of whether the childrens performance in reading was significantly asso ciated with the proposed set of exploratory variables, including phonologica l awareness, short-term memory span, rapid naming, and auditory processing ability For regression analyses reported in this section, six phonology measures from the CTOPP test and one auditory processing test (Dichotic Digit ) served as the exploratory variables. All of seven reading measures a nd a spelling measure were used as dependent

PAGE 112

112 variables. Each of these exploratory predictors was entered in to the model individually in Step 2 after the effects of childrens b ackground characteristics were statistically eliminated. Specifically, grade, nonverbal IQ and degree of heari ng loss (better ears PTA) were entered to control for the effect s of these variables on reading performance controlled for. These three factors were entered at Steps 1 in all analyses. These hierarchical regression resu lts are presented in Tabl e 4-22 through Table 4-28. The research questions are as follows: Regressions on word-level reading Separate fixed-order hierar ch ical regression analyses with were carried out separately on four word-level reading achievement measures (i.e., untimed/timed, words/nonwords) to determine the variance contributed to the word-level reading accuracy by phonological awarene ss, rapid naming, short-term memory, and auditory processing skills (32 regressions = 4 reading m easures 8 predictors). As reported in the previous section, the zero-ordered correla tions had shown significant correlations between reading and phonological pr ocessing tests. Especially, the Dichotic Digit test was included in the regression analyses since it was the only central auditory processing measures which was significantly correlate d with phonological measur es (i.e., elision, blending, nonword repetition, and memory for digits) and reading measures (i.e., timed and untimed nonword reading). The results ar e displayed in Table 4-21 through Table 424. Regression on untimed word reading accuracy: The Word Identification subtest on the WRMT-R was the dependent variable. As shown in Table 4-21, when grade, better ears PTA, and nonverbal IQ were entered in Steps 1, they explained 31.9% of the variance in the timed single-word reading scores. The results demonstrate that only the

PAGE 113

113 Elision task significantly explai ned substantial amount of variance (30.3%) in the Word Identification test ( p = .005). Taken as a whole, the three control vari ables (grade, better ears PTA, nonverbal IQ) and the Blending test accounted for 64.1% of variance for the Word Identification showing the importance of phonological awareness in word recognition. The Nonword Repetition and Dichotic Digits subtests only approached significance with the Word Identification test( p = .054 and p = .093, respectively). Both the RAN-Letters and RANDigit tasks did not explain variance in this untimed word reading test (p = .131 and .081, respectively). Regression on untimed nonword reading accuracy: The Word Attack subtest on the WRMT-R was the dependent variable. In Step 1, grade, better ears PTA, and nonverbal IQ explained 11.0% of varian ce in the score. Similar to the Word Identification when the Elision scores were entered in Step 2, it explained almost more than half of additional variance in the scores of the Word Attack subtest in a significant way (53.1 %, p < .0001). When the Nonword Repetition was entered instead of the Elision it also explained 37% of additional variance. So, both the Elision and Nonword Repetition represented a substantial eff ect on the untimed single nonwor d reading scores (Table 422). However, phonological short-term memory ( Memory for Digits ) only approached significance level and it explai ned about 19% of variance (p = .072). Unlike the Blending when the RAN (Digits and Letters) scores were entered in Step 2 individually, none of them was significantly associated with the Word Attack scores. With the Dichotic Digit

PAGE 114

114 entered in Step 2, the overall regression model did not significantly explained the variance in the Word Attack scores (i.e., F[4,14] = 2,300, p = .110). Regression on timed word reading accuracy: The Word Efficiency subtest on the TOWRE was the dependent variable. As Tabl e 4-23 shows, when entered first in the regression function, grade, PTA, and nonverb al IQ accounted for 19.3% of the variance in timed nonword reading. In Step 2, follo wing these controlled variables, each exploratory variable was entere d separately into the regression equations. Similar to the results of regressions on th e untimed reading condition, phon ological awareness skill as measured by the Elision subtest continued to significantly account for unique variance in the dependent variable (28.7%, p = .015). It was noted that the Memory for Digit s ( p = .178), Nonword Repetition ( p = .115), and Dichotic Digit subtests ( p = .306) did not explain the variance in a significant way. It was of further interest to determine if the rapi d naming tasks accounted for significant variance for timed reading activities. Recall th at our correlation analyses indicated that rapid naming skills were strongly associated with timed reading tasks only and no significant partial correlations were observed for untimed reading measures: Correlation coefficients for timed measures were high, ranging from .459 to .656 at word level and from .571 to .723 at passage leve l. This strong correlation between rapid naming and timed reading was confirmed by the regression results. That is, in the second order of entry, the alphanumeric rapid naming measures accounted for significant variance in the Word Efficiency scores. Specifically, the R AN-Letter accounted for an additional 24.5% of the variance of sc ores in the timed word reading ( p = .027) and the RAN-Digit test also explained al most 20% of the variance (19.4%, p = .049).

PAGE 115

115 Regression on timed nonword reading accuracy: The Word Decoding subtest on the TOWRE was the dependent variable. When entered first in the regression function, non-verbal IQ, grade, and PTA accounted only for 9.7% of the variance in the Word Decoding. Similar to the previous re sults, phonological awareness measure ( Elision ) explained a highly significant 37% of uni que variance as shown in Table 4-24. No significance was seen for the cont ribution of phonological memory. Childrens performance on the Dichotic Digit test only approached significance with the Word Decoding ( p = .054). There was also a significant additional contribution from the alphanumeric naming measures at Step 2, explaining a highly significant 39.2% (RANLetter, p = .005) and 31.3% (RAN-Digit, p = .016) of unique varian ce, respectively. This confirmed the fact that there is a significan t additional contribution of alphanumeric RAN to timed reading tasks. Regression on passage reading rate and accuracy Another set of analyses were carried out to test f or th e effects of our exploratory variables on reading skills in passage level. Recall that the RAN tasks did uniquely and significantly account for the vari ances of all timed, but not fo r untimed reading tests. So, we wanted to see if a similar trend would be found as well in passage level reading. As Table 4-25 shows, each of our two RAN variables, when separately entered at the Step 2, accounted for substantial proportions of variance in both passage rate and accuracy scores below the significant level: The RAN-Digit accounted for 37.6% ( p = .002) for rate scores and 31.7% for accuracy scores ( p = .010), respectively. In a similar pattern, it was also observed that the RAN-Letter accounted for 19.7% ( p = .044) for rate and 22.1% ( p = .039) for accuracy.

PAGE 116

116 Like our previous observation, the Elision consistently explained unique variances for both rate and accuracy scores below the significant level: 29.1% for rate ( p = .011) and 42.8% for accuracy ( p = .002), respectively. Childrens performance on the Memory for Digit subtest also significantly explained 18.9% of variance in the passage rate scores ( p = .049), but not for the accuracy score. The Nonword Repetition did not significant account for any of passage-level tests. Finally, the Dichotic Digit test did not show significant association with the performance on the rate measure, but it uni quely and significantly accounted for 26.2% of variance in the rate scores (p = .017). This rate test wa s the only literacy measure for which the variance was significantly accounted for by the Dichotic Digit test. It was noteworthy that when the RAN-Digit was entered in Step 2, the overall regression models explained significantly and uniquely 61.3% and 50.0% of variance, respectively, for passage reading rate and accuracy. Regression on spelling One f inal set of multiple regression analyses were carried out to test for the effects of our exploratory variables on spelling. The Spelling subtest on the WRAT was the dependent variable. In Step 2, chronological grade, better ears PTA, and nonverbal IQ accounted for 18.3% of the variance. The re sults are reported in Table 4-27. Rapid naming measures did not significant pred ict childrens spelling ability. Again, significant contribution of phonologi cal awareness was consistent: The Blending subtest explained statis tically significant 33.8% of variance in spelling ( p = .007). With the controlled variables entered in Step 1 taken together, the overall model accounted for more than half of varian ce in spelling scores (52.1%). Similarly,

PAGE 117

117 phonological short-term memory span as measured by the Memory for Digit explained almost 30% unique variance (p = .015). When the Dichotic Digit subtest was entered in St ep 2, which had significant standardized coefficient ( = .663, p = .016), the overall regression model itself did not significantly reduce the e rror term (F[4,14] = 3.071, p = .052). Role of rapid naming in further regression analyses In our previous regression analyses, ra pid nam ing tests did not account for unique variance in untimed reading tasks in both wo rd and passage level. Instead, it was shown that the RAN tasks are significantly associat ed only with timed reading tests, which demands high efficiency in extraphonological processing such as rapid lexical access and retrieval. Because this finding is relevant to the role of sp eed of processing (RAN) as a non-phonological factor in reading disabilities (Compton, DeFries, & Olson, 2001; Cornwall, 1992; Kirby, Parrila, & Pfeiffer, 2003; Savgrade et al., 2005; Strattman & Hodson, 2005; Wolf & Bowers, 1999; Wolf & O Brien, 2001), it was necessary to ensure that the RAN would still accounted for unique variance of each timed reading task after the effects of phonology-based cognitive sk ills were also controlled for. As a further test, a set of additional hierarchical re gressions were performed by entering two phonological awareness ( Elision/Blending ) and two phonological short-term memory ( Memory for Digit/Nonword Repetition ) in the same block (Step 2) as a group. Each of the RAN tests (digit/letter) was entere d separately in Step 3 for all timed reading measures. Similar to our prev ious regression, grade, non-verb al IQ, and better ears PTA were entered in Step 1. The results of these regressions are displayed in Table 4-28. In accordance to our hypotheses, the results of further analyses demonstrated that once IQ, grade, and better ears PTA were considered first, a significant additional

PAGE 118

118 contribution from the alphanumeric naming meas ures was preserved: In the model where the RAN-Digit was entered in Step 3, it continued to account for unique variances in passage reading rate significantly (23.6%, p = .005), passage reading accuracy (16.0%, p = .022), and timed nonword reading (20.1%, p = .029), respectively. Similarly, the second set of analyses also confirmed that th ere was significant addi tional contribution of the RAN-Letter measure at Step 3, explaining 21. 2% (p = .023) and 35.8% (p = .001) unique variances of timed word and nonword reading accuracy. The RAN-Letter also explained smaller but significant amount of variances in passage reading rate and accuracy (14.9% and 15%, respectively). Summary: 1. The Elision from the CTOPP test, a measure of phonological awarene ss, was the only measure, which consistently and si gnificantly explained variance in all of literacy measures ranging from word and nonword reading, spelling, to passage reading fluency. Unexpectedly, the Blending subtest was not significant for any of tests. The results also showed that alphanumeric rapid naming skills only accounted for unique variance in timed reading achievement in both word a nd passage level. In contrast, no significant contribution to tim ed reading was revealed. 2. When the effects of grade, non-verbal IQ and better ears PTA were eliminated, auditory processing skill as measure by the Dichotic Digit was not significant for any word-level reading and spelling tests. It explained statistically significant 26.2% of variance in passage read ing rate (26.2%). 3. In addition, the significant contribution of phonological short-term memory skill was seen only for two literacy measures. The Memory for Digit subtest explained 28.7% unique variance in spelling scores and the Nonword Repetition accounted for 37.0% unique variance in untimed nonword reading (p = .015 and p = .007, respectively).

PAGE 119

119Table 4-1. Background information and basic audito ry skills of indivi duals in the HI group. WRSd Subj Gende r Agea Gradeb PTAc SRTc R L HAFita Age_Ida HA_ Usea Etiology Hearing aid type 1 F 115 4.3 63 45 88100242488GeneticPhonak Pico Forte 2 M 111 4.3 58 55 100100484857UnknownOticon 380P 3 F 114 4.4 38 40 100100303084GeneticPhonak Pico Forte 4 M 100 2.5 70 65 4667202075UnknownOticon 380P 5 F 112 3.5 68 65 7689272781GeneticOticon Multi-Focus 6 F 106 2.5 38 35 93100424266GeneticPhonak Sono Forte 7 M 86 1.5 31 30 100100242462UnknownUnitron UM 60 8 F 125 3.5 27 20 100100363675UnknownOticon 39 PL 9 F 121 5.6 56 45 90100727248GeneticTelex 366 10 M 91 2.7 65 35 100887780UnknownPhonak Pico Forte 11 M 99 1.9 38 35 9688484825UnknownOticon 380P 12 M 122 2.9 41 35 100100181898GeneticOticon 380P 13 M 84 1.0 43 35 8896292933UnknownPhonak Sono Forte 14 M 152 6.9 41 40 1001003636115UnknownUnitron Icon 15 F 84 1.1 41 35 100100616120GeneticPhonak Pico Forte 16 F 152 6.0 43 30 5268727280UnknownUnitron Icon 17 M 106 3.0 28 10 5672***UnknownUnitron UM 60 18 M 111 3.2 45 45 7280242451UnknownOticon 38 P 19 M 116 3.2 58 50 5276242482UnknownRion HB 75AL M (SD) 11 Boys 8 Girls 110.8 (19.3) 3.4 (1.61) 46.9 (13.5) 39.5 (13.5) 75.2 (29.7) 86.0 (21.9) 35.6 (17.8) 35.6 (17.8) 67.8 (25.1) Note. a :in months, b:in years.months, c :unaided, d: measured in the Most Comfortable Le vel, PTA: Pure-tone threshold, SRT: Speech recognition thresholds, WRS: Word recognition scores, HA_Fit: Age at initial hearing aids fitting, Age_Id: Age at identification, HA_Use: Dura tion of hearing aids use

PAGE 120

120Table 4-2. Descriptive statistic s for the NH and HI groups on PTA, SRT, and WRS scores. NHa,b HIa,c Mean SD Min Max Mean SD Min Max PTA(L)d 5.1 4.7 -1 13.0 51.0 15.4 28 86 PTA(R)d 3.8 1.15 1 21.0 54.9 21.1 27 100 PTA (Better ear)d 3.31 3.9 -1 13.0 46.9 13.5 27 69 SRT(L)e 4.1 3.8 0 15.0 40.8 14.5 10 70 SRT(R)e 5.2 3.4 0 10.0 47.8 22.4 15 100 WRS(L)f 99.7 1.48 92 100 86.0 21.9 10 100 WRS(R)f 99.0 2.96 88 100 75.2 29.6 10 100 Note. a: in dB HL, b: Normally hearing children, c: Hearing-impaired children, d: unaided pure-tone threshold average, e: unaided speech recognition thresholds, f: word recognition scores in the most comfortable level (MCL)

PAGE 121

121Table 4-3. Mean pure-tone thresholds for all tested frequencies (NH and HI groups only). Ear 250 Hz 500 Hz 1000 Hz 2000 Hz 3000 Hz 4000 Hz 6000 Hz 8000 Hz M 37.63 47.11 59.21 63.68 60.79 62.37 68.16 70.53 SD 20.640 21.558 20.225 18.845 69.450 23.943 22.249 20.541 *HI Range 5 ~ 75 5 ~ 90 30 ~ 100 35 ~ 105 25 ~ 90 20 ~ 95 30 ~ 105 30 ~ 105 M 4.31 5.86 4.66 5.69 4.14 6.03 3.10 5.17 SD 5.782 5.012 4.805 6.778 5.680 6.322 5.414 5.745 Right NH Range 0 ~ 20 -5 ~ 15 0 ~ 20 0 ~ 35 -5 ~ 20 -5 ~ 25 -5 ~ 15 -5 ~ 25 M 32.63 39.21 48.42 56.32 56.32 57.89 61.58 65.63 SD 16.361 16.352 19.794 15.532 16.570 21.168 22.977 21.660 *HI Range 5 ~ 65 10 ~ 70 15 ~ 95 35 ~ 95 25 ~ 90 20 ~ 95 30 ~ 105 30 ~ 105 M 3.79 4.66 4.14 3.79 5.34 6.03 2.07 4.41 SD 5.454 5.659 4.446 5.287 4.616 6.461 6.053 6.033 Left NH Range -5 ~ 15 -5 ~ 15 -5 ~ 15 -5 ~ 15 0 ~ 20 -5 ~ 25 -10 ~ 20 -5 ~ 20 Note. *: unaided thresholds, HI (n=19), NH (n=29)

PAGE 122

122 Figure 4-1. Mean PTA thresholds for the left ear (HI Group, N=19) Figure 4-2. Mean PTA thresholds for the Right ear (HI Group, N=19)

PAGE 123

123 Figure 4-3. Histogram of 19 hearing impaired children by better ear pure-tone average.

PAGE 124

124Table 4-4. Descriptive statistics of auditory processing variables. Group SCAN (FW) SCAN (AFG) SCAN (CW) DD Mean 12.10 9.41 10.14 8.54 SD 1.970 2.383 2.812 8.533 Min 6 5 4 67.50 NH Max 15 15 14 98.75 Mean 5.89 3.11 4.11 66.97 SD 4.653 2.331 2.492 18.13 Min 1 1 1 38.75 HI Max 13 7 8 90 Note. SD: standard scores, FW: Filtered Words, AFG: Auditory Figure Ground, CW: Competing Words, DD: dichotic digit in %

PAGE 125

125Table 4-5. Descriptive statistics and estimated adjust ed means of oral language measures (MANCOVA) 95% CI Dependent variables Group M SD Min Max Median EAM SE lower Upper NH 112.00 12.45 81 132 113.0 110.45 2.496 105.42 115.47 PPVT-III HI 91.42 15.31 62 115 93.0 93.80 3.133 87.48 100.10 NH 107.38 10.46 92 134 105.0 106.55 2.238 102.04 111.05 EVT HI 94.32 13.46 63 112 95.0 95.59 2.809 89.92 101.24 NH 117.86 12.68 89 149 118.0 116.51 2.122 112.24 120.78 CASL HI 95.68 10.40 75 115 96.0 97.74 2.663 92.38 103.12 Note. PPVT-III (receptive vocabulary), EVT (expressive vocabulary), CASL (grammatical knowledge), SE (standard error), NH: normally hearing group, HI: hearing-impa ired group, M: mean, SD: standard devia tion, Min: minimum, Max: maximum, EAM: estimated adjusted mean, SE: standard error

PAGE 126

126Table 4-6. ANOVA tables for univariate ANCOVAs for each oral language measure. MS MSE F(1,44) p-value partial 2 Observed power PPVT-III 2736.54 169.733 16.123 .000** .268 .975 EVT 186.241 136.444 8.694 .005* .165 .822 CASL 3477.002 122.627 28.354 .000** .392 .999 Note. MS: mean square, MSE: mean 2 = effect size. *: p < .01, **: p < .001 Table 4-7. Pairwise comparison resu lts based on adjusted group means 95% CI for Difference Dependent Variable A B Adjust means difference (A-B) Standard errors p-value Upper Bound Lower Bound PPVT NH HI 16.651 4.147 .000** 8.294 25.009 EVT NH HI 10.963 3.718 .005* 3.470 18.457 CASL NH HI 18.769 3.525 .000** 11.666 25.873 Note. NH = normally hearing group, HI = hearing impaired group. *: p < .01, **: p < .001 All pairwise comparisons are based on estimated margin al means and all scores are adjusted for multiple comparisons using Bonferronis method. The mean difference is significant at the .05 level.

PAGE 127

127 Figure 4-4. Clustered box plot of oral language measure for the NH and HI groups Note: Dotted line indica tes the mean score. Standard scores (mean=100)

PAGE 128

128Table 4-8. Descriptive sta tistics and estimated adjust ed groups means for phonological processing measures 95% CI group M SD Min Max Adjusted means SE LowerUpper NH 12.102.410 7 16 12.080 .452 11.17912.981 HI 8.632.140 4 13 8.581 .571 7.4439.719 Elision# RD 7.432.528 3 13 7.488 .466 6.5598.416 NH 11.522.400 7 16 11.416 .409 10.60112.232 HI 7.791.584 4 11 7.567 .517 6.5388.597 Blend# RD 8.532.345 3 13 8.771 .422 7.9319.611 NH 11.172.406 7 18 11.064 .395 10.27811.851 HI 10.01.700 7 12 9.762 .498 8.76910.755 RAN-Digit## RD 7.772.112 4 13 8.022 .407 7.2118.832 NH 11.282.389 7 18 11.257 .429 10.40312.111 HI 10.052.094 6 14 10.012 .541 8.93311.091 RAN-Letter## RD 7.932.258 4 14 7.977 .442 7.0978.857 NH 11.072.685 6 17 11.078 .495 10.09212.064 HI 9.112.622 5 14 9.125 .625 7.88110.370 Memory for Digits# RD 9.102.551 4 15 9.079 .510 8.06310.094 NH 12.412.292 8 17 12.433 .444 11.54713.318 HI 7.892.787 3 12 7.936 .561 6.8189.055 Nonword Repetition# RD 8.902.107 5 14 8.855 .458 7.9439.768 Note. NH: normally hearing group, HI: hearing-impaired group, M: mean, SD: standard deviation, Min: minimum, Max: maximum, SE: standard error, #: NH>HI=RD, ##: NH=HI>RD.

PAGE 129

129Table 4-9. Summary of three univariate ANCOVAs on the phonological measures from the CTOPP. MS MSE F(2,74) p-value partial 2 Observed power Group comparisons Elision 155.657 5.807 26.803 .000 .420 1.000 NH > HI = RD Blending 96.433 4.755 20.281 .000 .354 1.000 NH > HI = RD RAN Digits 60.917 4.425 13.767 .000 .271 .998 NH = HI > RD RAN Letters 70.881 5.218 13.584 .000 .269 .997 NH = HI > RD Memory for Digits 34.694 6.949 4.993 .009 .119 .798 NH > HI = RD Nonword Repetition 144.587 5.609 25.780 .000 .411 1.000 NH > HI = RD Note. MS: mean square, MSE: mean square for error, partial 2 =effect size, df1 = 1, df2 = 44

PAGE 130

130Table 4-10. Summary of post hoc pa irwise comparisons of phonologi cal measures (CTOPP subtests). 95% CI for Mean Differencea Dependent Variables A B Mean Difference (A-B) SE t Sig Upper Bound Lower Bound NH HI 3.500 .716 4.888 .000** 1.747 5.252 NH RD 4.593 .664 6.917 .000** 2.965 6.220 Elision HI RD 1.093 .766 1.427 .473 -.784 2.970 NH HI 3.849 .647 5.949 .000** 2.263 5.435 NH RD 2.645 .601 4.400 .000** 1.172 4.118 Blend HI RD -1.204 .693 1.737 .260 -2.902 .4940 NH HI 1.302 .625 2.083 .122 -.2280 2.832 NH RD 3.043 .580 5.247 .000** 1.622 4.464 RAN-Digit HI RD 1.741 .669 2.568 .033* .1020 3.379 NH HI 1.245 .678 1.836 .211 -.416 2.907 NH RD 3.281 .630 5.208 .000** 1.738 4.823 RAN-Letter HI RD 2.035 .726 2.803 .019* .2560 3.814 NH HI 1.953 .783 2.494 .044* .0350 3.870 NH RD 1.999 .727 2.750 .022* .2190 3.780 Memory for Digits HI RD .047 .838 0.056 1.00 -2.006 2.100 NH HI 4.496 .703 6.395 .000** 2.774 6.219 NH RD 3.577 .653 5.478 .000** 1.978 5.177 NWR HI RD -.919 .753 1.220 .679 -2.763 .9260 Note. *: p < .05, **: p < .001. NH = normally hearing group, HI = hearing impaired group, RD = Dyslexic group, All comparisons are based on estimated margin al means using Bonferronis method for multiple comparisons.

PAGE 131

131 Figure 4-5. Clustered boxplot fo r phonological awareness measur es for the NH, HI and RD groups. Note: Dotted line indicates the mean score. Standard scores (mean=10)

PAGE 132

132 Figure 4-6. Clustered boxplot fo r phonological short-te rm memory for the NH, HI, and RD groups. Note: Dotted line indicates the mean score.Standard scores (mean=10)

PAGE 133

133 Figure 4-7. Clustered boxplot for rapid naming for the NH, HI, a nd RD groups. Note: Dotted line indicates the mean score. Standard scores (mean=10)

PAGE 134

134 Table 4-11. Descriptive statistics and estimated adjusted group means for literacy measures. 95 % CI M SD Adjusted means SE lower upper Word NH 115.24 8.947 119.5 2.398 114.728 124.283 HI 108.11 9.451 105.1 2.984 99.174 111.066 RD 83.10 13.469 83.3 2.364 78.558 87.980 Nonword NH 115.62 12.982 115.5 2.360 110.796 120.201 HI 110.16 11.388 105.5 2.937 99.643 111.348 Timed reading (TOWRE) RD 82.40 8.830 85.4 2.327 80.767 90.041 Word NH 119.34 13.036 110.7 2.220 106.235 115.082 HI 106.37 15.475 98.6 2.763 93.096 104.106 RD 82.63 13.405 84.6 2.189 80.221 88.945 Nonword NH 115.38 13.270 113.0 2.226 108.517 117.387 HI 106.42 14.698 99.7 2.770 94.210 105.250 Untimed reading (WRMT) RD 84.93 12.031 84.7 2.195 80.344 89.091 NH 112.93 13.564 115.3 2.000 111.341 119.310 HI 99.89 13.552 107.5 2.489 102.494 112.411 Spelling RD 84.63 8.720 83.4 1.972 79.504 87.361 Comprehension NH 110.62 9.951 115.7 2.030 111.654 119.745 HI 98.89 12.391 109.5 2.527 104.509 114.580 RD 84.43 13.318 82.7 2.002 78.723 86.702 Reading rate NH 14.00 2.104 13.9 .530 12.935 15.048 HI 11.26 3.754 11.3 .660 10.013 12.643 RD 4.97 2.822 4.9 .523 3.892 5.976 Reading accuracy NH 14.55 2.759 14.6 .572 13.407 15.688 HI 11.63 3.804 11.7 .712 10.245 13.084 Passage reading RD 5.90 2.820 5.9 .564 4.758 7.008

PAGE 135

135 Figure 4-8. Clustered boxplot for untimed word-level reading ( Word Identification and Word Attack subtests on the WRMT-R). Note: Dotted lines indicate 1 SD. Standard scores (mean = 1 0 0)

PAGE 136

136 Figure 4-9. Clustered boxplot for untimed word-level reading ( Word Efficiency and Word Decoding subtests on the WRMT-R). Note: Dotted lines indicate 1 SD. Standard scores (mean=100)

PAGE 137

137 Figure 4-10. Clustered boxplot fo r connected-text reading fluenc y (GORT-4 Rate and Accuracy) (Note: Dotted lines indicate 1 SD. Standard scores (mean=100)

PAGE 138

138 Figure 4-11. Clustered boxplot for passage co mprehension (WRMT-R) and spelling (WRAT) (Note: Dotted lines indicate 1 SD). Standard scores (mean=100)

PAGE 139

139Table 4-12. Summary of univa riate ANCOVAs on the readi ng and spelling measures. Dependent Variable MSB MSE F(2,74) p partial 2 Observed Power Group comparison Word 9767.786 166.657 58.610 .000 .613 1.000 NH > HI > RD Untimed Nonword WRMT-R 6850.592 161.457 42.430 .000 .534 1.000 NH > HI > RD Word 7955.022 115.904 68.634 .000 .650 1.000 NH > HI > RD Timed Nonword TOWRE 8774.433 119.518 73.415 .000 .665 1.000 NH = HI > RD Spelling WRAT 5874.333 143.631 40.899 .000 .525 1.000 NH > HI > RD Comp* WRMT-R 5012.224 142.865 35.084 .000 .487 1.000 NH > HI > RD Rate GORT 629.020 8.154 77.143 .000 .676 1.000 NH > HI > RD Passage reading Accuracy GORT 567.806 9.499 59.777 .000 .618 1.000 NH > HI > RD Note. The F tests the effect of group in each ANCOVA. Covariate was grade in months. *: passage comprehension. MSB: mean squares fo r between group, MSE: mean square error

PAGE 140

140 Table 4-13. Summary of post hoc pa irwise comparisons of eight reading and spelling measures. 95% CI for Mean Differencea Dependent Variables A B Mean Difference (A-B) SE t Siga Upper Bound Lower Bound NH HI 14.386 3.833 3.75 .001** 4.997 23.774 NH RD 36.237 3.365 10.77 .000*** 27.994 44.479 WRMTa (untimed word) HI RD 21.851 3.825 5.71 .000*** 12.481 31.221 NH HI 10.003 3.772 2.65 .029* .762 19.243 NH RD 30.094 3.312 9.09 .000*** 21.981 38.207 WRMTa (untimed nonword) HI RD 20.091 3.765 5.34 .000*** 10.869 29.314 NH HI 12.058 3.548 3.40 .003*** 3.366 20.750 NH RD 26.076 3.115 8.37 .000** 18.444 33.707 WRMT (comprehension) HI RD 14.018 3.541 3.96 .001* 5.343 22.693 NH HI 13.222 3.558 3.72 .001* 4.507 21.938 NH RD 28.235 3.124 9.04 .000** 20.583 35.887 WRAT (spelling) HI RD 15.013 3.551 4.23 .000** 6.314 23.711 NH HI 7.873 3.196 2.46 .048* .044 15.703 NH RD 31.893 2.806 11.37 .000*** 25.019 38.767 TOWREb (timed word) HI RD 24.020 3.190 7.53 .000*** 16.206 31.834 NH HI 6.155 3.246 1.90 .185 -1.795 14.106 NH RD 32.987 2.849 11.58 .000*** 26.007 39.967 TOWREb (timed nonword) HI RD 26.832 3.239 8.28 .000*** 18.897 34.767 NH HI 2.664 .848 3.14 .007** .587 4.740 NH RD 9.058 .744 12.17 .000*** 7.235 10.881 GORT (reading rate) HI RD 6.394 .846 7.56 .000*** 4.322 8.467 NH HI 2.883 .915 3.1 .007*** .641 5.124 NH RD 8.664 .803 10.9 .000*** 6.697 10.632 GORT (reading accuracy) HI RD 5.782 .913 6.3 .000*** 3.545 8.019 Note. a: timed test, b: untimed test, *: p < .05, ** p < .01, *** p < .001, Covariate: grade(in months and years)

PAGE 141

141 Table 4-14. Correlations between the degree of hearing loss and phonological measures, partialing out for age and grade. elision blending RANDigit RANLetter Memory for Digit Nonword repetition r -.644 -.721 -.367 -.396 -.287 -.597 HI sig. .007** .002** .162 .129 .281 .015* r -.203 .049 -.174 -.108 .322 .031 PTA (better ear) NH sig. .320 .814 .394 .598 .109 .880 Note. *p < .05, **p < .01 Table 4-15. Partial correlation matrix between phonology and reading (HI group). untimed timed Passage word nonword word nonword spell compre hension rate accuracy r .582 .751 .562 .611 .618 .610 .605 .711 elision sig .009 .000 .012 .006 .005 .006 .007 .001 r .189 .304 .193 .159 .379 .289 .470 .378 blend sig .241 .126 .237 .278 .074 .139 .033 .074 r .307 .411 .459 .560 .418 .341 .723 .635 RANDigits sig .124 .057 .037 .012 .054 .098 .001 .004 r .259 .301 .559 .656 .479 .314 .581 .571 RANLetter sig .166 .129 .012 .003 .030 .118 .009 .010 r .160 .394 .287 .285 .558 .284 .525 .393 Memory for Digits sig .277 .066 .141 .143 .012 .143 .018 .066 r .362 .594 .323 .324 .291 .324 .375 .322 Nonword Repetition sig .084 .008 .111 .110 .137 .111 .076 .112

PAGE 142

142 Table 4-16. Partial correlation matrix between phonology and reading (NH group). untimed timed Passage word nonword word nonword spell Comprehension rate accuracy r .711 .692 .142 .474 .825 .520 .443 .671 elision sig .000 .000 .244 .007 .000 .003 .012 .000 r .378 .472 -.023 .326 .511 .185 .116 .392 blend sig .028 .007 .456 .052 .004 .183 .286 .024 r .421 .382 .559 .596 .311 -.018 .382 .465 RANDigits sig .016 .027 .001 .001 .061 .465 .027 .008 r .414 .437 .542 .629 .399 .006 .385 .540 RANLetter sig .018 .013 .002 .000 .022 .489 .026 .002 r .657 .603 -.106 .305 .519 .502 .185 .377 Memory for Digits sig .000 .001 .302 .065 .003 .005 .182 .029 r .686 .650 .137 .366 .534 .484 .433 .558 Nonword Repetition sig .000 .000 .253 .033 .002 .006 .014 .002

PAGE 143

143Table 4-17. Correlation matrix between better ea rs PTA and phonology, partialing out age (in months). Control Variables Elision Blend RAN-Digit RAN-Letter Memory for Digits Nonword Repetition age_month r -.473 -.597 .120 .085 .116 -.424 Better ear PTA p .024 .004 .318 .368 .323 .040 r -.565 -.528 -.012 .032 .166 -.438 Better ear SRT p .007 .012 .481 .449 .255 .034 r 1.000 .608 .320 .215 .326 .579 Elision p .004 .097 .196 .094 .006 r .608 1.000 .225 .160 .415 .461 Blend p .004 .185 .263 .043 .027 r .320 .225 1.000 .868 .197 .149 RAN-Digit p .097 .185 .000 .216 .278 r .215 .160 .868 1.000 .126 -.004 RAN-Letter p .196 .263 .000 .309 .494 r .326 .415 .197 .126 1.000 .502 Memory for Digits p .094 .043 .216 .309 .017 r .579 .461 .149 -.004 .502 1.000 Nonword Repetition p .006 .027 .278 .494 .017

PAGE 144

144Table 4-18. First-order correlation matrix be tween oral language and phonology measures. Elision Blend RAN-Digit RAN-Letter Memory for Digit Nonword Repetition PPVT r .324 .362 -.063 -.130 .533 .406 p .102 .077 .405 .309 .014 .053 EVT r .252 .224 .112 -.044 .592 .359 p .164 .194 .335 .434 .006 .078 CASL r .511 .443 .210 .162 .558 .494 p .018 .038 .209 .267 .010 .022

PAGE 145

145Table 4-19. Correlation matrices among audito ry processing measures, and phonology and re ading scores, partia ling out for age, grade, duration of hearing ai d use, and nonverbal intelligence. Phonological measures Reading measures PA RAN STM Untimed timed Passage reading elision Blend Digit Letter MD NWR Word NW Word NW Spell Comp Rate Accu .354 .410 -.022 -.012 .143 .409 -.135 -.028 .294 -.100 -.191 -.110 -.270 -.225 SCAN1 (FW) .196 .129 .939 .965 .610 .130 .633 .921 .287 .722 .496 .696 .330 .420 .087 .432 -.131 -.333 .430 .563 -.193 -.155 .075 -.361 -.323 -.118 -.158 -.404 SCAN2 (AFG) .757 .108 .641 .225 .110 .029 .491 .580 .790 .187 .240 .676 .575 .135 .222 .451 -.105 -.250 .653 .579 -.215 -.120 -.059 -.152 -.192 -.272 -.264 -.514 SCAN3 (CW) .426 .091 .711 .368 .008 .024 .442 .671 .835 .589 .493 .327 .341 .050 .714 .659 .391 .069 .617 .548 .306 .546 .234 .525 .479 .322 .439 .261 DD_perc .003 .008 .149 .808 .014 .034 .267 .035 .400 .044 .071 .242 .102 .348 Note. MD: Memory for Digit, NWR: Nonword Repe tition, Comp: Comprehensi on, FW: Filtered Words, AFG: Auditory Figure-Ground, CW: Compe ting Words, DD_perc: Dichotic Digits in percentile, Accu: Accuracy

PAGE 146

146 Table 4-20. Correlation matrix between oral language and reading measures. PPVT EVT CASL .286 .283 .203 Word Identification (WRMT-R) .151 .153 .234 .247 .361 .291 Word Attack (WRMT-R) .187 .093 .147 .474 .536 .442 Passage Comprehension (WRMT-R) .037 .020 .049 .187 .319 .238 Spell (WRAT-3) .252 .123 .197 .368 .127 .363 Word Efficiency (TOWRE) .088 .326 .092 -.025 .128 .129 Word Decoding (TOWRE) .465 .325 .324 .271 .442 .194 GORT-4 Rate .164 .050 .245 .092 .189 .129 GORT-4 Accuracy .372 .250 .323 Note. Age, grade, duration of hearing aid use, and nonverbal intelligence were partialed out.

PAGE 147

147Table 4-21. Hierarchical re gressions of variables separately predicting untimed word reading (the Word Identification subtest on the WRMT-R). B SE B t sig R2 R2 F Step1 Age Better ears PTA Nonverbal IQ .319 Step 2 Elision 4.673 1.395 .646 3.351 .005 .622 .303 11.227 Blending 3.260 2.517 .334 1.295 .216 .392 .073 .216 RAN-Digits 4.061 2.156 .446 1.883 .081 .457 .138 3.547 RAN-Letter 3.055 1.903 .413 1.606 .131 .425 .106 2.578 Memory for Digits 1.662 1.263 .282 1.316 .209 .394 .075 1.732 Nonword Repetition 2.459 1.169 .443 2.103 .054 .482 .164 4.423 Dichotic Digits .380 .211 .445 1.803 .093 .447 .128 3.249 Note. B: estimated beta, SE B: standard error of estimated beta, : estimated standardized beta, R2: multiple correlation, R2: amount of increased multiple correlation coefficient.

PAGE 148

148Table 4-22. Hierarchical re gressions of variables separately predicting untimed nonword reading (the Word Attack subtest on the WRMT-R). B SE B t sig R2 R2 F Step1 Age Better ears PTA Nonverbal IQ .110 Step 2 Elision 5.873 1.291 .855 4.550 .000 .641 .531 20.698 Blending 4.008 2.686 .432 1.492 .158 .233 .122 2.227 RAN-Digits 4.495 2.329 .520 1.930 .074 .297 .187 3.726 RAN-Letter 2.636 2.134 .396 1.235 .237 .198 .087 1.525 Memory for Digits 2.505 1.290 .447 1.943 .072 .299 .189 3.774 Nonword Repetition 3.513 1.113 .666 3.156 .007 .480 .370 9.957 Dichotic Digits .539 .209 .665 2.576 .022 .397 .286 6.637 Note. B: estimated beta, SE B: standard error of estimated beta, : estimated standardized beta, R2: multiple correlation, R2: amount of increased multiple correlation coefficient.

PAGE 149

149 Table 4-23. Hierarchical re gressions of variables separately predicting timed word reading (the Word Efficiency on the TOWRE). B SE B t sig R2 R2 F Step1 Age Better ears PTA Nonverbal IQ .193 Step 2 Elision 2.780 .999 .629 2.784 .015 .481 .287 7.748 Blending 1.794 1.705 .301 1.053 .310 .252 .059 1.108 RAN-Digits 2.953 1.397 .531 2.113 .050 .388 .195 4.465 RAN-Letter 2.840 1.148 .629 2.473 .027 .438 .245 6.117 Memory for Digits 1.180 .832 .327 1.418 .178 .294 .101 2.010 Nonword Repetition 1.366 .814 .403 1.678 .115 .328 .135 2.817 Dichotic Digits .159 .150 .305 1.064 .306 .253 .060 1.131 Note. B: estimated beta, SE B: standard error of estimated beta, : estimated standardized beta, R2: multiple correlation, R2: amount of increased multiple correlation coefficient.

PAGE 150

150 Table 4-24. Hierarchical re gressions of variables separately predicting timed nonword reading (the Word Decoding on the TOWRE). B SE B t sig R2 R2 F Step1 Age Better ears PTA Nonverbal IQ .097 Step 2 Elision 3.809 1.217 .716 3.130 .007 .469 .372 9.795 Blending 1.990 2.194 .277 .907 .147 .147 .050 .823 RAN-Digits 4.510 1.653 .673 2.728 .016 .411 .313 7.445 RAN-Letter 4.328 1.319 .796 3.280 .005 .489 .392 10.761 Memory for Digits 1.484 1.063 .342 1.397 .184 .207 .110 1.951 Nonword Repetition 1.723 1.039 .422 1.658 .120 .245 .148 2.748 Dichotic Digits .364 .173 .579 2.105 .054 .414 .217 4.431 Note. B: estimated beta, SE B: standard error of estimated beta, : estimated standardized beta, R2: multiple correlation, R2: amount of increased multiple correlation coefficient.

PAGE 151

151 Table 4-25. Hierarchical re gressions of variables separately predicting timed passage r eading rate on the GORT-4. B SE B t sig R2 R2 F Step1 Age Better ears PTA Nonverbal IQ .238 Step 2 Elision 1.110 .378 .633 2.938 .011 .528 .291 8.631 Blending 1.198 .604 .505 1.982 .607 .405 .167 3.929 RAN-Digits 1.628 .441 .737 3.690 .002 .613 .376 13.614 RAN-Letter 1.011 .458 .564 2.210 .044 .435 .197 4.884 Memory for Digits .640 .298 .447 2.149 .050 .427 .189 4.616 Nonword Repetition .506 .317 .375 1.596 .133 .355 .117 2.547 Dichotic Digits .132 .049 .636 2.705 .017 .499 .262 7.317 Note. B: estimated beta, SE B: standard error of estimated beta, : estimated standardized beta, R2: multiple correlation, R2: amount of increased multiple correlation coefficient.

PAGE 152

152 Table 4-26. Hierarchical re gressions of variables separately predicting timed passage r eading accuracy on the GORT-4. B SE B t sig R2 R2 F Step1 Age Better ears PTA Nonverbal IQ .183 Step 2 Elision 1.366 .348 .768 3.929 .002 .611 .428 15.435 Blending 1.047 .660 .436 1.586 .135 .307 .124 2.514 RAN-Digits 1.514 .509 .676 2.977 .010 .500 .317 8.860 RAN-Letter 1.085 .476 .596 2.279 .039 .404 .221 5.195 Memory for Digits .520 .332 .359 1.565 .140 .305 .122 2.449 Nonword Repetition .482 .337 .353 1.427 .175 .287 .104 2.038 Dichotic Digits .180 .056 .515 1.931 .074 .355 .172 3.727 Note. B: estimated beta, SE B: standard error of estimated beta, : estimated standardized beta, R2: multiple correlation, R2: amount of increased multiple correlation coefficient.

PAGE 153

153 Table 4-27. Hierarchical re gressions of variables separately predicting scores on the Spelling on the WRAT. B SE B t sig R2 R2 F Step1 Age Better ears PTA Nonverbal IQ .183 Step 2 Elision 4.322 1.375 .683 3.143 .007 .521 .338 9.877 Blending 4.091 2.310 .478 1.771 .098 .332 .150 3.137 RAN-Digits 3.816 2.079 .479 1.836 .088 .341 .159 3.369 RAN-Letter 3.377 1.770 .522 1.908 .077 .351 .169 3.642 Memory for Digits 2.849 1.034 .551 2.755 .015 .470 .287 7.592 Nonword Repetition 1.790 1.195 .368 1.498 .156 .296 .113 2.245 Dichotic Digits .495 .181 .663 2.735 .016 .467 .285 7.480 Note. B: estimated beta, SE B: standard error of estimated beta, : estimated standardized beta, R2: multiple correlation, R2: amount of increased multiple correlation coefficient.

PAGE 154

154 Table 4-28. Summary of hier archical regressions on timed reading measures, controlled for phonological awareness and short-term memory scores (showing the results from Step 3 only). Criterion variables Entered explorator y variable (RAN-Digit) in Step 3 B SE B t Sig. R2 R2 Reading rate (GORT-4) 1.522 .425 .689 3.583 .005 .816 .236 Reading accuracy (GORT-4) 1.270 .471 568 2.695 .022 .779 .160 Words (TOWRE) 2.449 1.592 .440 1.538 .155 .592 .096 Non-words (TOWRE) 4.259 1.666 .636 2.556 .029 .692 .201 Criterion variables Entered exploratory variab le (RAN-Digit) in Step 3 B SE B t Sig. R2 R2 Reading rate (GORT-4) .927 .396 .517 2.340 .041 .728 .149 Reading accuracy (GORT-4) .945 .370 .520 2.553 .029 .770 .150 Words (TOWRE) 2.786 1.035 .617 2.691 .023 .708 .212 Non-words (TOWRE) 4.368 .895 .803 4.882 .001 .850 .358 Note. B: estimated beta, SE B: standard error of estimated beta, : estimated standardized beta, R2: multiple correlation, R2: amount of increased multiple correlation coefficients.

PAGE 155

155 CHAPTER 5 DISCUSSION Introduction This present study sought to explore four m a jor questions concerning the specific link between mild to moderate SNHL an d reading and spelling abilities. (1) Are there significant effects of SNHL on ha rd-of-hearing childrens skills of spoken language, phonological processi ng, and reading/spelling? (2) What are the patterns of correlations among reading, phonological ability, and spoken language? Which phonological variables are associ ated with which types of reading tasks in the HI group? (3) Are phonological processing skills important in reading achieve ment of hearing impaired readers? Specifically, do they predict unique variance in reading and spelling? Oral Language Skills Findings The first purpose of the present study was to re p licate the results of previous studies on the spoken language ability of the HI children compared to children with normal hearing. There are only a few scientific studies of the language sk ills of children with mild to moderate SNHL (Davis, Shepard, Stelmachowicz, & Gorga, 1981; Gilbertson & Kamhi, 1 995; Davis, Elfenbein, Schum, & Bentler, 1986; Blair, Peterson, & Vi ehweg, 1985; Blamey, Sarant, Paatsch, Barry, Bow, Wales, 2001; Gilbertson and Kamhi, 1995; Hansson et al., 2004; Stelmachowicz et al., 2004). Most previous studies, conducted on children with a range of lingui stic skills, revealed that a significant proportion of ch ildren with mild to moderate SNHL have depressed vocabulary and morphosyntactic performance.

PAGE 156

156 Group Comparison The investigators goal was to determ ine whet her mild to moderate SNHL has significant effects on hard-of-hearing childre ns vocabulary and grammatical knowledge and, if so, to what degree, and in what areas when controlling for non-verbal IQ and grade. As indicated in Delage and Tu ller (2007), SNHL is associated with not only lowered hearing thresholds, but also distortion of speech signals and degraded language input, factors that are assumed to affect age-appropriate development of linguistic knowledge. Considering the fact that normal speech recognition development is conti ngent upon successful auditory input during a constrained critical period in the first few years of life (e .g. McConkey Robbins, Burton Koch, Osberger, Zimmerman-Phillips & Kishon-Rabin, 2004; Sharma, Dorman & Spahr, 2002), one would expect that many children with mild to moderate HL are at risk for delayed language development. The average age for initial hearing aids fitting of the HI subjects included in this study was close to 4 years (mean: 44.4 months). Researchers have indicated that although language input can be noticeably improved through hearing aids, mild to moderate level of SNHL in children is detected and remediated at rela tively older ages than profound hearing loss, with the average age for remediation between 4 and 5 (Hansson, Forsberg, Lfqvist, Mki-Torkko, & Sahln, 2004; Stelmachowicz, Pittman, Hoover, & Lewis, 2004; Tuller & Jakubowicz, 2004). Thus, due to perceptual distortion of phonologica l input during the critical la nguage acquisition period, if not aided early, hard-of-hearing chil dren are likely to experience inaccurate speech perception and develop unclear phonological representations of words, and de layed of phonological sensitivity (awareness). Since phonological awareness is required for efficient phonemic segmentation, which is necessary for distinct phonological repres entations of newly learned words, hearing loss may ultimately cause language impairments (Chiat, 2001).

PAGE 157

157 Consistent with previous studies the findings described in this study also indicate that the group of aided, school-aged children with mild to moderate SNHL have significantly lower spoken language skill in all tests areas than would be expected for children with normal hearing. Adjusted means for the HI groups were more than 1 SD below the mean of the NH children for both receptive vocabulary and grammatical know ledge tests: 16.651 for the PPVT-III and 18.769 for the CASL (See Table 4-5). The largest gr oup difference was noted for morphosyntactic knowledge task, as evidenced by a strongest effect size of .392. Although the HI group's performance on the PPVT-III, EVT, and CASL test s was not impaired below the mean score of 100 (See Figure 4-4), further inspections of raw scores in each test showed that 31% of the HI children had standard scores of 1 SD below th e mean and only 26% of them had raw scores above the mean. Findings of oral language skills in this study are different from those reported by Briscoe et al. (2001), who suggested that the vocabulary an d receptive grammar of HI children were not significantly lower than their control group matched on age. One possible reason for this discrepant result is associated w ith the statistical treatment of th e data. In Briscoe et al. (2001), multiple univariate ANOVAs were used to inves tigate the group effect instead of multivariate analyses which can control the effects of confounding covarying factors (i.e., age, grade, or IQ). Secondly, Briscoe and colleagues (2001) sample was comprised primarily of children with mild hearing loss; only 3 children had moderate HL. Fu rther, Briscoe et al. reported that their HI subjects can be divided into two subgroups ba sed on the severity of phonological impairments: unimpaired and impaired groups. The latter s ubgroup showed significan tly poorer vocabulary skills than the group with unimp aired phonological skills. Close sc rutiny of their data showed that the impaired groups scores on vocabulary were at or belo w 1 SD below the normative mean

PAGE 158

158 scores: 86.6 (receptive) and 84.6 (e xpressive), respectively. Interestingly, these two subgroups did differ significantly in the severity of hearing loss, that is, the impaired groups hearing loss was more severe across all frequencies. Based on the graphic display of the hearing thresholds (Briscoe et al., 2002:337), the unimpaired groups PTA based on 0.5, 1, and 2 KHz was around 25 dB HL. This strongly suggests that the unimpaired groups or al language skills were not affected by this mild level of hearing loss. Base d on these observations, it can be inferred that the phonologically unimpaired groups normal level of la nguage skills masked the negative effect of hearing loss on spoken language ab ility causing overall mean scores of the HI group to be exaggerated. That is, significant differences migh t have been captured if the sample of the HI group was more representative of the population w ith mild to moderate SNHL. Lastly, existing significant difference could have been observed when confounding va riables such as age, grade, or non-verbal IQ were statistically controlled. Interrelationships between Phonological Processing and Language Skills The investigator also tried to determ ine wh ether linguistic knowledge was related to HI subjects phonological processing skills that include the ability to encode, store, and retrieve phonological information. Phonological short-term memory and vocabulary Previous res earch has consistently indicat ed a close association between childrens phonological short-term memory (STM) and thei r vocabulary knowledge (Baddeley, Gathercole, and Papagno, 1998; Gathercole & Baddeley, 1990; Gathercole, Hitch, Service, & Martin, 1997; Jarrold, Baddeley, Hewes, Leeke, & Phillips, 2004). Baddeley et al. (1998) argued that the central purpose of verbal STM to facilitate the efficiency of the long-te rm learning of vocabulary. For example, learning an unfamiliar word requires ordering the constituent phonemes of the word along with its accurate phonological represen tation for storage in memory. They suggested

PAGE 159

159 that phonological material is fi rst represented in phonological ST M, which makes it available for long-term learning. As reviewed in chapter 2, the authors described the phonological loop as a language learning device and ar gued that the ability to reta in small amounts of phonological information in STM evolved as a human charact eristic that has a selection advantage for language acquisition. Gupta (2003) provided a similar explanation wh en he suggested that during learning of unfamiliar phonological sequences for new word ac quisition, incoming phonemes are stored in a specific short-term memory system, called sequence memory. This stored information permits the replay or verbal rehearsa l of the new phonological sequences and increases the probability that phonological representation in the long-term memory will be stable. Findings from the current study support this word learning hypothesis. In fact, the subjects STM tested by a digit span and a nonword re petition test, was str ongly correlated with vocabulary (See Table 4-18). Especially, the Memory for digits was significantly associated with all vocabulary and morphosyntactic with partia l coefficients ranging from .533 to .592. The Nonword repetition was highly correlated with morphosyntax measure (partial coefficient = .494, p = .022) and partial coefficients for receptive an d expressive vocabulary measures approached the significant level (r = .406, p = .053 with the PPVT-III and r = .359, p = .078 with the EVT, respectively). In Baddeley and colleagues(1998) st udy of memory in 6-year old children, firstorder correlation coefficients between two measur es of verbal STM (auditory digit span and nonword repetition) and vocabul ary knowledge were .44 and .56. Si milarly, in Gathercole and colleagues (1997) study, digit sp an scores correlated significan tly with all tested vocabulary measures, with significant corr elations ranging from .38 to .44 ( p < .01).

PAGE 160

160 Grammatical knowledge and phonological processing In a study of short-term memory in childre n with specific language impairment (SLI), Gathercole and Baddeley (1990) developed a th eoretical framework linking deficits in phonological STM (i.e., nonword repetition) to impaired language acquisition. The authors pointed out that poor performance on nonword re petition is a marker of specific language impairment. However, as pointed out by Leonard (1998), vocabulary is typically less impaired than syntax in children with SLI (Leonar d, 1998). This raises the question of whether phonological STM is implicated in acquisition of syntactic knowledge as well as vocabulary. Considerably less work has addressed the ro le of phonological STM in the acquisition of syntax, even though there is in creasing support for the notion th at the same psychological mechanisms underpin vocabulary and syntax acq uisition (Marchman & Bates, 1994). In their 1998 study, Baddeley et al. indicated that in typically developi ng children, nonword repetition, a measure of phonological l oop capacity, was related not only to vocabulary learning but also to acquisition of syntax. In this respec t, the HI groups performance on the Grammatic Knowledge subtest in this study is of particular in terest since it was strongly associated all of the phonological awareness and phonological memory te sts (range of partia l coefficients: .443 ~ .558). How would we explain this strong asso ciation between phonological STM, phonological awareness, and morphosyntactic knowledge in th e HI group? Gathercole (2006) offers three hypotheses to explain the li nk between weak phonological STM and poor morphosyntax. The first hypothesis relates to a storage capacity proposing that incoming speech stream needs to be stored in a temporary buffer while the morphosyntactic analysis is carri ed out. To process the grammatical morpheme, each phoneme should be pe rceived well and the inflection marker must be separated from the content part. Next, its grammatical function must be realized based on the

PAGE 161

161 utterance situation. All of this must be completed before the next item in an utterance enters the system. So, perceived material may be lost owing to restrictions in working memory or because previous material is still being processed. That is, an inflected ve rb will be fully processed some times and at other times it will not. Therefore, if stored phonological material is rapidly decayed, subsequent operations will become inefficient. The second hypothesis is relate d to the contribution of phonological awareness to morphosyntactic awareness. It proposes that poor morphosyntactic skills will follow if a child persists in segmenting or identifying incomi ng speech at the supra-phonemic level such as syllable. For instance, past tense markers in monosyllabic words such as walked, hopped, and laughed may not be recognized if the chil d is only processing inco ming words at syllable or onset-rime borders, not being able to separately recognize the word-final grammatical morphemes. Thus, underdeveloped phonological awar eness can cause the skills for extracting a morphosyntactic rule, such as past tense fo rmation to be delayed. That is, childrens phonological sensitivity (awareness) will help to successfully divi de a word into content and bound morphemes (phonemic segmentation), enhanci ng their ability to perceive and manipulate morphological units, too. The high association found in this study between phonological awareness and knowledge of gramma tical morphemes (r = .511 for the Elision and r = .443 for the Blending ) well supports this theo retical standpoint. The third and final hypothesis addresses the possible challenges faced by some children to perceive English grammatical morphemes that have low-perceptual salience (Rom & Leonard, 1990), which might result in weaker or low-quality representa tion of grammatical markers such as -ed (past tense) and -s (plu ral or third person singular marker). Recent data from children with cochlear implants suggests that the patter n of language development is strongly influenced

PAGE 162

162 by the perceptual prominence of relevant morphol ogical markers (Svirsky, Stallings, Lento, Ying, and Leonard, 2000). As indicated above, lower-quality phonological repr esentations resulting from inaccurate speech perception may cause slowed morphosyntactic computation. Although the speech of children with mild to moderate hear ing loss is intelligible, research indicated that aided speech perception would be still problematic For instance, Elfenbein et als (1994) study reported that misarticulation of fricatives and affricates was common, particularly for children with 3-frequency pure-tone average thresholds greater than 45 dB HL. Thus, as hypothesized in Nobury et al. (2001), even mild hearing impairment might result in morphemes of low perceptual salience be ing missed and therefore delaying grammatical development. Phonological Processing Findings We know that an intact phonological system provides an important foundation for learning to read in normally hearing children (N athan, Stackhouse, Goulandris, & Snowling, 2004; Rayner, Foorman, Perfetti, Pesetsky, & Seidenber g, 2001). This present research was aimed at investigating what happens to the hard-of-hearing ch ildrens phonological syst em when auditory stimulus signals are distorted. Sp ecifically, this study was designed to investigate which part of the phonological system is affected and which part is preserved (i.e., the whole phonological system, or sub-processes such as some aspects of the metaphonological skills, phonological memory, or lexical access/retrieval skills?) Previous studies have reported that, overall, phonological processi ng skills of children with mild to moderate SNHL are significantly lowe r than those children without hearing loss. Surprisingly, however, little information is known re garding (1) the degree to which these skills are impaired and (2) the range of phonological skills affected.

PAGE 163

163 Furthermore, no previous studies have co mpared HI childrens phonological skills with those of dyslexic children at the varying range of phonological processing skills. Since dyslexic children are known to have persistent deficits in phonological proces sing with no auditory perceptual limitations related to cochlear damage, comparing these two groups should help to elucidate the nature and degree of deficit in HI childrens phonological processing skills. The fundamental hypothesis underlying this study was that hearing loss would affect HI childrens auditory perceptual accuracy, leading to phonological proc essing abilities. In general, phonemic manipulations become very difficult if the relevant segments are not distinctly specified in the representation of the words (e .g. Elbro, 1996, 1998; Elbro, Borstrm & Petersen, 1998; Foy & Mann, 2001; Fowler, 1991; Griffith & Snowling, 2002; Swan & Goswami, 1997a; Wesseling & Reitsma, 2001). Group data in this study revealed statis tically significant differences between the HI and NH groups on phonological awarene ss and phonological STM skills. HI children had lower scores on tasks of: Elision ( p < .0001), Blending ( p < .001), Memory for digits ( p < .05), and Nonword repetition ( p < .0001). Adjusted mean differences (NH minus HL) were more than 1 SD for the Elision (3.50), Blending (3.849), and Nonword repetition (4.496) measures. Associat ed with this depression, the HI groups skills on phonological awareness and STM were not better than those of the RD group. The results of this present investigation are on ly partially in line with previous studies. Most of research indicated that mild to m oderate hearing SNHL would adversely affect phonological processing ability. For in stance, in Briscoe et al. (2001) children with SNHL were significantly lower than normal age controls on phonological awareness (as measured by onset and rhyme matching), phonological discrimination, and nonword repetition. Interestingly, they found that hearing-impaired childrens digit recall was not lower than that of normal controls. In

PAGE 164

164 this study, however, the HI groups phonological STM skill was significantly lower than the NH group on both tasks. But, the Nonword repetition task showed bigger grou p mean difference than the Memory for Digit task. One way of explaining this result is that the of degree of familiarity of the phonological representation for digit names (e.g., one, two, ) should place fewer demands on long-term memory than the nonword re petition task (Bruck, 1992; Bruck & Treiman, 1990). Since immediate nonword repe tition is highly dependent on verbal memory function, it should be more sensitive than digit span in detecting any existing memory deficit. In contrast, this present study found no significant differences between the HI and NH groups on the alphanumeric RAN tasks. At the sa me time, the HI group showed significantly better rapid naming skills than the RD children. This remarkable finding suggests that hearingimpaired students lexical access and retrieval skills are preserve d in spite of their impaired phonological awareness and shor t-term memory skills. This finding was supported further by correlati ons which revealed th at degree of hearing loss was only associated with phonological awareness and short-term memory skills, not with the RAN measures. Specifically, when chronological age (in months) was controlled, better ears PTA and SRT showed moderate to high negative correla tions with the Elision Blending and Nonword repetition. This tight association suggests that hearing and speech perception would impact the HI childrens phonologically-based skills. Gibbs (2004) also reported similar negative co rrelation levels between hearing thresholds and phonological awareness for rhyme awareness and initial phoneme awareness. As noted above, no significant relation was noted between th e RAN tasks and the scores of better ears PTA or SRT. This lack of significant associati on indicates that periphera l hearing loss does not affect the efficiency of fast lexical access and retrieval or speed of processing.

PAGE 165

165 Hearing Loss and Phonological Representation It is now we ll established that early reading problems are associated with difficulties in phonological domain of language (Mann, 1998; Stanovich, 1993; Wagner & Torgesen, 1987) and most researchers view weak phonological awareness skills as contributing directly to dyslexics deficiency in word recognition skill (Blachman, 1989; Bowey, Cain, & Ryan, 1992). However, despite the strong rela tionship between phonological skills and reading, the nature of the phonological deficits that underlie poor reading remains to be elucidated. Elbro and Jensen (2005) proposed that dys lexic readers problems with phonemic awareness are due to can be explained by their poorly specified phonologi cal representations. In line with Elbro and Jensen (2005)s proposal, other researchers have suggested that the relationship between impaired phon ological awareness and reading di fficulty be considered from the perspective of the quality (o r distinctness as in Elbro, 19 98) of phonological representations of words in the mental lexicon. This hypothesis, referred to as the phonological representation hypothesis, states that a lack of distinct ness or segmental specificity in phonological representations cause phonological processing difficulties (e.g. Elbro, 1996; Fowler, 1991; Hansen and Bowey, 1994; Metsala, 1997; Snowling et al ., 1986; Swan and Goswami, 1997). Fowler (1991) has proposed that words in each individuals mental lexicon vary according to the precision of the phonological specificati on of the underlying representations. So, words with less detailed representations are those that lack full segmental organization into a sequence of discrete phonemic elements. As a consequenc e, segmental manipulation of the phonetic form of the word may be hampered as a crystallized phonetic code is not available. Snowling and Hulme (1989), Hulme and Snowling (1992), and McDougall, Hulme, Monk, and Ellis (1994) have also suggested that an awareness of the phonological segments of words may not be as critical to success on phonological awareness tasks as the accuracy of the underlying

PAGE 166

166 phonological representations of the words. In accordance with Fowl er, they argue that if the integrity of a persons phonologica l representations is compromi sed, it follows that the person should have difficulty performing segmental ope rations on those representations. Similarly, Manis et al. (1997) noted that if children cannot perceive clear distinctions between phonemes, it will be hard for them to develop representations that can be easily accessed to further processing. Several lines of evidence also have suggest ed that poor phonological awareness relates to poor speech perception. For example, consonant pe rception in newborns is associated with language competence at 3 to 5 years, is pred ictive of reading achievement at 8 years (Scarborough, 1990). Children and adults with read ing difficulties have difficulties perceiving speech in noise (Brady, Poggie, & Rapala, 1989; Brady, Shankweiler, & Mann, 1983), and children who are poor readers requi re a longer segment of a gated word in order to perceive it correctly (Metsala, 1997). In sum, problems with accessing phonological representations, in turn, lead to difficulty in segmenting and manipulating phonemes (phonological awareness). Thus, knowing that children with congenital SNHL may experien ce degraded auditory input that interferes with their speech perception (Abraham, 1993) and even small speech perception problems may be very important for the ability to form fully specified phonological repr esentation (Manis et al., 1997), it is expected that hypothesized that hard-of-children with prelinguis tic auditory disadvantage should thus have lexical representations that are less distinct or accura te, leading to difficulties in phonological awareness. The HI groups weak p honological awareness skills found in this study support this theoretical position. Phonological Representation and Short-term Memory Skill The results of this study replicated previous studies showing that hear ing loss can also lead to inefficient storage capacity. However, while the measures on nonword repetition and digit

PAGE 167

167 span are good markers of depre ssed phonological STM, it is not clear why hearing loss should affect the capacity of verbal memory. Research has documented that phonologi cal STM is also constrained by the quality of stored material. Service, Maur y, and Luotoniemi (2007) explored the idea that good verbal STM depends on the quality of representations of phonological sequences that are encoded into the phonological store. Generally speaking, this idea is justified because the efficiency of phonological segments in temporary storage should be maximized if the materi al to be stored is as distinct or clear as possible. As Nairne (1990) and Neath & Nairne (1995) suggested, phonological representations of good quality are less prone to bei ng overwritten and should not quickly decayed. Of course, such better-quality traces will serve as a good resource for more efficient and accurate phonological awareness. The present findings of reduced STM in children with SNHL appear to support the theories that predict a relationship betw een the quality of phonological repr esentations and STM. That is, individuals with hearing loss w hose auditory-phonetic inputs are not clearly represented in STM have smaller spans. As they can hold less ma terial of good quality in working memory, they would also not be as good at manipulating st ored phonemes and passing information from the phonological store to the articulatory control process as would children with normal hearing do. Articulatory control process is an active process component in Baddeleys memory model that can keep the contents of the phonological stor e active and counteract time-based decay of phonological traces by a kind of i nner speech (Baddeley, 2003). Intact Rapid Naming In the present study, the rapid nam ing skills of HI children with mild to moderate SNHL were preserved in spite of their impaired phonological awareness and phonological memory abilities. While rapid naming predicts reading skills, the nature of the RAN as a measure

PAGE 168

168 phonological skill is still unclear. Wagner and Torg esen (1987) argued that rapid naming is a phonologically-based skill. Wagner, Torgesen, and Rashotte (1999) proposed that young readers first retrieve sounds associated with letters or letter pairs, th en retrieve the pronunciations of word segments, and finally retrieve the whole word. From this perspective, rapid naming should pred ict reading skill because this task measures the efficiency of retrieving phonological codes associated with phonemes, word parts, or entire words (Shankweiler & Crain, 1986; Share, 1995; Torgesen & Burgess, 1998). To them, RAN tasks are an index of the speed with whic h phonological information can be accessed from memory and, thus, are best describe d as tapping into an aspect of phonological processing. However, other researchers have argued that rapid naming is not exclusive a phonological skill and that it requires several other skills, including executive functioning (Denckla & Cutting, 1999), the ability to detect and represent ort hographic redundancy (Bower s & Wolf, 1993; Wolf & Bowers, 1999), global processing efficiency (Ka il, Hall, & Caskey, 1999), and attention skill (Neuhaus, Foorman, Francis, & Carlson, 2001). In a similar context, Wolf and Bowers (1999) ha ve proposed that proces ses related to serial naming speed represent a second core-deficit in children with RD. This alternative model, known as the double-deficit hypothesis, presumes th at phonological processing and rapid automatized naming (RAN) deficits are separable sources of reading dysfuncti on. As discussed in chapter 2, three separate subtypes of RD individuals (i.e. phonological-onl y, rate-only, and double deficit) can be predicted based on the varying effects of deficits in phonologi cal processing and RAN. Among these, the phonological-defi cit-only subtype is characterized as having significant deficits in phonological processing (PA and STM) with otherwise intact naming speed processes.

PAGE 169

169 This present study revealed that the HI gr oups performance on the rapid naming tasks was well preserved and hearing loss differentially aff ected their phonologicallybased skills only (PA and STM). Hence, their profile is, at least to some degree, characte ristic of the single (phonological-only) deficit group described by Wolf and Bower (1999). The RD group who showed depression in both rapid naming and phonologically based skills (PA and STM) is characteristic of double-deficit group. Literacy Findings Based on the findings discussed in the previ ous section, it was hypothe sized that if poorer hearing in the HI child ren influenced their phonological proc essing perform ance, then their performance on a range of reading and spelling tasks will also be affected. This position is consistent with the well-recognized phonologica l core deficit hypothesis. It was also hypothesized that the correla tions between phonological pro cessing ability and reading performances would be significant. Group Comparisons Although the HI childrens performances on literacy measures were within normal range on all tasks using standardized te st scores as the crit erion, statistically si gnificant quantitative differences between the NH and HI groups were found. That is, the reading and spelling skills of the hearing-impaired children studied lagged behi nd those of their age-matched normal-hearing peers and were significantly better than those of their RD peers. Specifically, the HI groups reading and spelling skills were significantly lower than normally developing group on all word-level and passage-level m easures with the exception of the nonword reading subtest (Decoding Efficiency on the TOWRE). When compared to the NI controls, both the HI and NH subjects had signifi cantly better scores that their RD peers on all tested measures. The RD group overall showed marked deficits in both phonological awareness

PAGE 170

170 and rapid naming (i.e., double-defi cits) while the HI group showed deficits in phonological awareness (i.e., phonological deficit only). That is, the data clearly showed that two core deficits (i.e., phonological deficit and rapid naming deficit) were selectively depressed in the HI and RD groups literacy skills. These results lend s upport to the double-deficit hypothesis, which proposes that reading disability can be cause d either by phonological processing deficits, by RAN deficits, or (in the most se vere cases) by a combination of both deficits (Wolf and Bower, 1999). However, these findings were mode rately different from the previ ous studies (Briscoe et al., 2001, Gibbs, 2004, Halliday & Bishop, 2006). Halliday & Bishop (2006) used two timed word reading tests from the TOWRE a nd they reported no significant difference between the HI and NH groups on both tests. Briscoe et al. (2001) used three readi ng tests, including two untimed word and nonword reading tests a nd one passage comprehension test and reported that their HI childrens reading was not different from the agematched controls and was slightly better than that of an SLI group in spite of their depre ss phonological processing sk ills. Based on this observation, the authors conclude d that hard-of-hearing children s reading development would not necessarily require the s upport from phonological processes. If deficient PA skills in HI kids do not im pact their reading, this finding would pose a serious challenge to the universality of phonological core de ficit theory. To date, many experimental studies with speci al populations have provided fi ndings that support the case for the universal necessary role of phonological pr ocessing capacity in reading acquisition. For example, a series of studies with deaf reader s provides strong evidence for the necessity of phonological coding in memory for skilled re ading. Hanson, Goodell, & Perfetti (1991) documented that congenitally d eaf individuals who had become fluent readers had somehow

PAGE 171

171 acquired a phonological coding stra tegy despite never having he ard speech. Fowler, Doherty, & Boynton (1995) also documented the contribution of phonological memory to reading acquisition in young adults with Down syndrome. These subjects, performance in phonological memory accounted for a substantial portion of the within-group va riability in reading achievement. Furthermore, neuroimaging studies s upport a universal phon ologically-based neurocognitive differences in all dyslexics. Res earchers posited that if dyslexia is caused by phonological deficit, this deficit should be observe d in neurobiological measures of the brain of dyslexics across cultures and language. Recently, Paulesu, Demonet, Fazio, McCrory, Chanoine, Brunswick, Cappa, Cossu, Habib, Fr ith, and Frith (2008) analyzed three dyslexic groups who spoke either Italian, French, or English on tasks of single word reading and phonological shortterm memory. Positron emission tomography (P ET) scans during reading showed the same reduced activity in a region of the left hemis phere in all dyslexics groups, with the maximum peak in the middle temporal gyrus and additional p eaks in the inferior and superior temporal gyri and middle occipital gyrus. They concluded that dyslexia is a result of a universal phonological processing deficit that is eviden t across different orthographies. In a similar vein, this stu dy showed that the HI groups partly impaired phonological processing skill has a strong association with lite racy, confirming its contributive and necessary role in developing skilled reading. Thus, as sugg ested by previous studies, the central role of phonological processing skills should not be diminished in the popul ation of children with mild to moderate SNHL. In the following section, the results from the correlational analyses and hierarchical regressi on are discussed.

PAGE 172

172 Correlation and Regression Findings No previous studies focused on the re lationship between reading and phonological process ing ability in children with mild to mode rate SNHL, when the effect of hearing loss was statistically partialed out. Four key findings emerged from the correlational and regression data. First, as similar before, the correlations between the HI children s phonological awareness ( Elision ) and all reading and spelling measures was remark ably consistent once covariations with age, nonverbal IQ, and degree of hearing loss we re controlled (partial coefficients ranged from .582 to .751). These correlations occurred regardless of the test conditions (i.e., timed, untimed) and level of tasks (i.e., word/passa ge) (Bowers, 1993; Bradley & Bryant, 1985; Cornwall, 1992; Cronin & Carver, 1998; Mu ter, Hulme, Snowling,& Taylor, 1997). This strong relationship between phonologica l processing skill and reading skill was supported with hierarchical regression analyses. Regression analyses on scores of untimed tasks indicated that the Elision subtest accounted for around 33% of the variance of word reading and 50% of variance of nonword readin g measure. More than 60% of variance was reduced for each skill by including the Elision subtest and background control variables (age, nonverbal, and better ears PTA). Similarly, on timed word-level reading tasks, an av erage of 33% of the variance were associated with the Elision subtest. The Elision test also significantly accounted for the variance in both passage readi ng rate (53%) and accuracy (43%). Interestingly, consistent with the findings reported in the CTOPP manual (Wagner et al., 1999), the Blending subtest did not correlated with any li teracy measures. Findings from this study support that although Elision and Blending tests are phonological awareness measures, they tap into different cognitive abilities. Blending may be less challenging than elision because it requires only one operation, the synthesis of se parate sounds into a word while holding these sounds in STM. On the other hand, the task of elision requires segmenting sound in words,

PAGE 173

173 omitting the target sound g, and creating a new word while holding all sounds in STM. Hence, elision appears to be a more ri gorous test of phonologi cal awareness. In their study of children with dyslexia, Katzir, Kim, Wolf, OBrien, Kennedy, Lovett, & Morris (2006) found that the dyslexic groups blending ability was co rrelated with only on e reading test ( Word Attack subtest on the WRMT-R) out of five phonological awarene ss and reading measures. In contrast, elision was significantly correlated with all five measures. These cumulative findings imply that not all phonological tasks are created equal and that greater precision in our conceptu alization of phonological tasks is required. Secondly, the relationships between rapid naming skill and reading were examined. Data of the HI group in this study underscored the indepe ndent characteristic of the RAN task. That is, it was remarkable that the RAN measures did not show any significan t correlations with untimed tasks of reading at both the word or passage level, but large correlations were found with timed reading tests at both word and passage levels. These data were supported in the regression analyses. These cumulative results support previous findings suggesting that the rapid letter naming task assesses different underlying constructs than those assessed by phonological measures (Misra, Katzir, Wolf, & Poldr ack, 2004; Wolf & Bowers, 1999; Wolf et al., 2002). Thus, in a task that primarily places no demands on speed ed reading, phonologically-based skills are important. For a task requiring fast lexical access (timed measures), the rapid naming skills play more of a role. In other words, RAN primarily affects performance on reading tasks that require speeded/fluent response, and phonological awareness and short-term memory primarily affect performance on reading tasks that em phasize phonological pr ocessing skill.

PAGE 174

174 In a study of 476 children with reading disorder, Compton, DeFr ies, & Olson (2001) similarly reported that the double-deficit group most resembled the rate-deficit group on measures that require fluent/speeded word r eading skill and reading comprehension, whereas the double-deficit group tended to pe rform similarly to the phonologi cal-deficit group on measures emphasizing phonological processing skills. Katzir et al. (2006) also repo rted that phonological awareness (elision and blending ) contributed more variance to phonologically-based reading measure and RAN contributed more to high freque ncy word reading and wo rd-reading efficiency, that are more related to speed of processing. In support of this obser vation, no significant correlations were found between the tw o alphanumeric rapid naming tasks and all other phonologically-based measures in the HI group (PA and STM) (See Table 4-17). For further analyses, the investig ator tried to evaluate explicitly the additive nature of the RAN tasks prediction of timed reading measur es. A final set of regression analyses was conducted, where phonological awarene ss and short-term memory vari ables were entered in Step 2 and each of the two RAN measures (letter and di git) was separately entered in Step 3 showed that RAN contributed independent va riance to timed reading measures even after controlling for childrens background variables, phonological awareness, and phonological short-term memory variables. This additive prediction by the RAN seems to confirm the findings from Wolf and Bowers (1999), who viewed rapid naming as an extraphonological construct and view rapid naming deficits as an additive source of r eading difficulty over phonological awareness. As yet, it is not clear which cognitive processes underlie RAN and account for its relationship with reading. Even though Wolf and Bowers theo retical framework can neatly explain the heterogeneity of clinical data complex interrelationships between RAN, phonological awareness, and reading itself ma ke causal inferences very difficult.

PAGE 175

175 In their large-sized study which assessed 1,010 childrens reading performance, Powell, Stainthorp, Stuart, Garwood, & Quin lan (2007) also reported that RAN deficits occurred in the absence of phonological awarene ss deficits. In their structur al equation modeling, solutions where RAN was subsumed within a phonological pro cessing factor did not provide a good fit to the data, suggesting that processes outsid e phonology may drive RAN performance and its association with reading. Convi ncingly, children with single RAN deficits performed more slowly in speed of processing than did closely matched controls with normal RAN skill. Kail, Hall, & Caskey (1999) also proposed th at speed of processing underlies the RAN skills. The reason for the lack of clarity about the causal natu re of the link between RAN and reading is that most of previous studies have b een correlational. Using th is approach, they could not rule out the possibility that, rather than be ing causal in nature, the relationship between RAN and reading may be driven by a third unknown fa ctor. This view was proposed by Kail and colleagues (Kail, 1991; Kail & Hall 1994; Kail et al., 1999), who ar gued that, rather than being constricted to the reading system or even to ling uistic processes, RAN performance reflects more generalized processing speed. According to this approach, early childhood is characterized by a general and gradual increase in global processing speed. Thus the relationship between RAN and reading should be underst ood by the fact that both are in fluenced by the same underlying factor, namely, processing speed. The present st udys results from the co rrelation and regression analyses which revealed important role of rapid naming in timed reading seem to fit into this approach. Thirdly, relationships between reading and or al language skills we re investigated. As expected, only grammatical knowledge and vocabulary skills were corr elated with passage comprehension, suggesting that linguistic capac ity dealing with vocabulary or grammatical

PAGE 176

176 knowledge does not significantly contribute to word-lev el reading tasks. This is likely to be the case because it appears that word-level reading does not necessarily require syntactic or semantic computation. That is, at the word level, poor phonological difficulties are likely to hinder childrens word recognition ability, but at the se ntence level, children may have more difficulties integrating the information and extract meaning due to grammatical and syntactical difficulties and/or their restricted semantic knowledge (Hartas, 2005; Perfetti 1985; Perfetti, 1992). Lastly, the predictive role of auditory proces sing skills for reading ability was investigated. Most of auditory processing tasks scores failed to show significant correlations with reading and spelling measure and did not predict reading in the HI group. Specifically, no SCAN subtests were correlated with reading measures. Only the Dichotic Digit test showed significant correlations with two nonword reading tests. Base d on the lack of association, it appears that auditory tests of this nature provide little predictive value for asse ssing reading skills, at least in children who have normal levels of intellectual abilities. Summary Contem porary research has reveal ed a great deal about the fact ors that interfere with the process of learning to read (Snow, Burns, Gri ffin, 1998). Perhaps the greatest contribution over the past two decades is accumulating evidence of a phonological processing deficit as the core problem leading to poor reading (Stanovich & Siegel, 1994). Marschark (1997) explained hardof-hearing childrens reading difficulty as resu lting from their impair ed phonological channel. Such findings support the theory about the role of impaired auditory perception in reading disability and provided the impetus for the co mparison between two populations with impaired phonological processing skills and depressed readi ng ability: (1) dyslexics with a good peripheral hearing system, but whose phonological processi ng is impaired, and (2) a non-dyslexic hearing-

PAGE 177

177 impaired population, whose peripheral hearing syst em is impaired. This dilemma was the basis of this present research. This study was designed with the premise that that literacy developm ent requires the same acquisition of phonological processing skills, whether a childs hear ing is impaired or normal. First, the principle aim of the study was to examine the effect of mild to moderate SNHL on childrens performances on a range of spoke n language, phonological processing skills, and literacy (reading and spelling). Performances of two controls groups of children (NH and RD), matched for age, grade, and non-verbal intelligence were used to investigate the HI groups weaknesses and strengths in phonolog ical processing ability and its relationship to reading skills. Consistent with previous findi ngs, the HI groups receptive and expressive vocabulary and morphosyntactic skills were significantly delayed when compared to those of the NH group. Similarly, the HI groups performance on the pho nological processing tasks was significantly lower than their normal controls. Howeve r, their phonological pr ocessing skills were selectively depressed such that only phonologically-based tasks (i.e ., phonological awareness and phonological short-term memory) were seen significan tly lower than the normal controls and the RAN skills, a measure of general speed of processing or efficiency of lexical access, was seen intact. In line with Katzir and colleag ues (2006) explanation, this fi nding suggests that auditory perceptual distortion due to SNHL only affect s phonologically-based pro cessing abilities and not extraphonological processing abilities (RAN). Secondly, turning to performance on the literacy tasks, this study sustains the view that reading is the most difficult academic challenge for the hearing impaired (Marschark & Harris, 1996). It was found that the HI groups literacy skills were significantly lower than their

PAGE 178

178 normally hearing controls in most of reading and spelling measures. Therefore, it was clearly noted that phonological processing deficit related to early perceptual inaccuracy is very likely to affect later literacy and related ta sks of children with mild to moderate SNHL. This critical role of phonological processing component was also c onfirmed by the correl ational and regression analyses which showed that phonological pro cessing skills are impor tant correlates and predictors of hearing-impaired childrens readin g and spelling as well. This finding contradicts previous studies which argued that reading ski lls of hearing-impaired would not necessarily require the support from phonologica l processing ability. As note d previously, RAN measures were not associated with any of untimed reading tasks, but had significant correlations only with timed reading tasks. This extraphonological property of the RAN was clearly demonstrated in the regression analyses, where RAN measures additively contributed independent variance to timed reading measures afte r phonologically-based variables were controlled for. One final issue for comment is the impact of degree of hearing loss on language, phonological processing and literacy. Many previous studies of readi ng in children with mild to moderate SNHL have not reported a significant eff ect of hearing loss on reading (Blair et al., 1985; Davis et al., 1995; Elfenbein et al., 1986; Gilbertson et al., 1995). In a similar vein, degree of hearing loss did not show any significant correl ations with reading and spelling measures after age and non-verbal IQ were controlled for. On e possible explanation for this insignificance is that, with the advent of bette r hearing aid technol ogy, the impact of de gree of hearing loss perceptual might have become latent in pr edicting reading achiev ement, only affecting phonological processing component. It is possible that the latent relationship between hearing loss and reading could be mediated by other c onfounding factors such as age at identification

PAGE 179

179 and/or initial hearing aid fitting, clinical efficacy of hearing aids educational environment, and so on. Clinical Implications First, consistent with prior research, the current study revealed that hearing-im paired childrens depressed phonological processing abilities are signifi cantly related to reading and spelling skills. The clinical implications are evident. Language, phonologi cal processing skills, and reading ability should be t horoughly examined even in children with mild to moderate HL. Secondly, in clinical settings, a common view is that some children with mild or moderate impaired hearing loss who are trained orally do s how systematic phonological deficits similar to those of normal hearing children with phonological disorders (Abraham, 1993). It is generally accepted that deaf and hard-of-hearing children learn to read following the same sequence of skill development that hearing children do (Chall, 1996; Hanson, 1989; King & Quigley, 1985; Paul, 1998, 2001). Therefore, it can be assumed th at hard-of-hearing individuals, like hearing individuals, will also benefit from the developm ent of phonological processing skills as part of their beginning reading instruc tion. Nielsen and Luetke-Stahlma n (2002) maintained that if hearing-impaired children are to read the sound-based printed word s and phrases of English, they must develop and improve phonologi cal sensitivity (awareness). Thus, previous results which suggested that phonological ability is not necessarily required in ch ildren with mild to moderate HL might lead many educators to assume that HI childrens weak phonolog ical awareness would not be a barrier to successfu l reading development and igno re the importance of phonological awareness training. There is research evidence to suggest that the ability to use phonological information while reading is a distinguishing vari able when comparing accomplished deaf and hard-of-hearing readers to average deaf and hard-of-hearing readers (Conrad, 1964; Engle, Cantor, & Turner,

PAGE 180

180 1989; Hanson, 1982; Hanson & Fowler, 1987; Hanson & Lichtenstein, 1990; Hanson, Goodell, & Perfetti, 1991; Leybaert, 1993; Musselman, 2000; Perfetti & Sandak, 2000). These findings imply that effective programs and strategies for teaching hearing-impaired children these skills may be the key to obtaining higher levels of reading achievement for this population. Unfortunately, survey studies of instructional me thods being used by teachers of hard-of-hearing children reported that the overwhelming majority of teachers do not incorporate direct teaching of phonological awareness in thei r reading instruction (Coley & Bockmiller, 1980; Hasenstab & McKenzie, 1981; LaSasso, 1978, 1987; LaSasso & M obley, 1997). Thus, currently, it is not surprising that hearing impaired students are at ri sk for acquiring sufficient decoding skills due to their limited access to and instruction focuse d on the phonological aspects of the English language (e.g., phonemic awareness and phonics). In this vein, the findings of this research would have implications on the direction of inte rvention planning and the strategies of teaching reading to members of the studied populations. Furthermore, speech pathologists who work with hard-of-children need to be aw are of clinical methods of e xplicit phonological instructions beyond traditional interven tion techniques used for children with articulation or phonological disorder. Limitations While this study was the first of its kind to provide a com prehensive profile of the phonological and literacy profiles HI childrens with mild to mode rate SN hearing loss, a few limitations in the studys met hodology must be acknowledged. First, the relatively small sample size of th e HI group (N=19) makes it difficult to draw conclusions regarding the eff ect of degree of hearing loss beyond this study. In most psychometric studies of hearing im paired children, it is well assumed that there will be a problem associated with participation. Children with mild hearing impairments are less likely to

PAGE 181

181 participate in extensive testing than children w ith more extensive hearing losses because their problems are not perceived as being as serious a nd parents of these children may not feel the immediate need to see results of their childrens performance in the areas of language and literacy. Secondly, the administration of more than ten psychometric tests combined with hearing tests might have been too demanding for some of the children, reducing the validity of tests. Lastly, in a recent comprehension review of readi ng research for students w ho are deaf or hard of hearing, Schirmer and McGough (2005) indicated that intervention research evaluating the effectiveness of phonological instruction for students is extrem ely limited. Unfortunately, no clinical studies exist regarding the efficacy of direct phonological instruction for children with moderate levels of HL. This lim ited number of studies certainl y warrants attention. As a crosssectional study, this research could not address th e developmental trajectory of reading skills in children with mild to moderate SNHL. Future longitudinal studies are needed to unravel the long-term effects of hearing loss on academic achievement.

PAGE 182

182 APPENDIX A INFORMED CONSENT LETTER FO R PARENTS AND GUARDIANS Informed C onsent Letter for Parents or Guardians Jan, 2007 Dear Parent/Guardian, I am a doctoral student of Communication Scie nces and Disorders at the University of Florida, conducting research under the supervisi on of Dr. Linda Lombardino. We are interested in examining the effect of hearing loss on oral language, phonological processing, and reading skills of 7to 12-year-old children. In spite of increasing number of chil dren with hearing loss, little is known about thei r reading performance. Especially, ve ry few research efforts have been directed to literacy development of children with less-than-profound h earing loss, so studies strictly focusing on the reading ability in this population are needed. The main purpose of this study is to a ssess the oral language, phonology, and reading skills of children with or without sensorineural hearing loss and compare them. For the study, we need; 1) 30 children with normal hearing/reading ability 2) 30 children with mild or moderate sensorineural hearing loss Children with conductive hearing loss would NOT be able to participate in the study and only childr en whose primary communication mode is spoken language are qualified. Students will be mostly enga ged in linguistic activities that have them listen to the sounds, read and write words and se ntences, or give appropriate answers to the questions of the research er. Assessment is expected to take 90 to 100 minutes and will be done using a set of standardized oral language and reading tests. Only when we could not finish the assessment, your child will be asked to work with us once more on the other day. Either a trained student or me will be present during all sessi ons to help your child to do the tests. There is no expected risk of physical, psychological, or economic harm to your child This assessment project will directly benefit your childs literacy skills regardless of their hearing level. That is, through the whole assess ment procedures, your childs profile of oral language and reading development will be inves tigated. Based on this, you and the classroom teachers are also expected to benefit from this study by being informed of the childs current level of oral language and literacy performance, which can be used to decide the steps needed to improve reading and oral language instruction. A copy of the results of the assessment will be available upon your request. With your permission, we will audio record your childs responses during assessment and obtain your childs hearing loss level from school files if your child has a hearing loss. Tapes Jungjun Park Doctoral Student, Dept. of Communication Sciences and Disorders, Univ. of Florida, Gainesville, FL 32611-7420 Tel.: (352) 392-2113 ext. 293 E-mail: pajj_gsc@hotmail.com

PAGE 183

183 will be used only to validate our assessment resu lts and the contents will not be transcribed. It will be available only to the assistan t student, my supervisor, and me. I also have attached a questionnaire aski ng for some information regarding the developmental characteri stics of your child and family background. Including the information from this questionnaire, all the da ta will be numerically coded and will not be marked with your childs name. A ll individual records including th e audiotapes will be destroyed once the study has ended. Your childs identity will strictly be kept confidential to the extent provided by law. No real names, initials, or ot her identifying information will be used during spoken or written presentation of study results. Participation or non-part icipation in this study will not affect your childs grades or placement in programs at all. As mentioned, children with normal hearing ability are also needed for this study. If your child would participate in th e study as a subject with normal hearing regardless of his/her reading proficiency, his/her hearing will be screened at the University of Florida Speech Hearing Clinic (UFSHC) by a graduate clin ician of Doctoral of Audiology program of the University of Florida. Hearing will be sc reened based on puretone audiometry and tympanometry and there is no cost for this evaluation. For this audiologic assessment, I will make an appointment with you beforehand. You and your child have the right to withdraw consent for your childs participation at any time without consequence. No compensation will be offered. If you have any questions about this research protocol, please contact me, Jungjun Park at (352) 392-2113, ext 293 or (352) 328-7671( pajjgsc@gmail.com ) or Dr. Lombardino at (352) 392-2113, ext 285. Questions or concerns about your ch ilds rights as a research part icipant may be directed to the UFIRB Office, University of Florida, B ox 112250 Gainesville, FL., 32611, (352) 392-0433. If you should decide to grant perm ission, please co mplete the attached questionnaire form and return it with this consent letter to the classr oom teacher. You can also keep a copy of this informed consent letter for your information. Thank you for the consideration. Sincerely, Park, Jungjun

PAGE 184

184 I have read the procedure described above and I have received a copy of this form. I understand that my decision is entirely voluntary and agree to allow my child ___________________, to participate in the study to in vestigate oral and written language skills. I also authorize the principal investigator, Ju ngjun Park, to secure and use information in the questionnaire form for his research. Signatures: Parent/Guardian: ___________________________ Date:__________ Principal Investigator: _________________________ Date:__________ Supervisor/Investigato r: ________________________ Date:__________ Please sign below and return this letter to the classroom teacher

PAGE 185

185 Child Assent Script The following is a script that will be used prior to each session to ensure that the student knows of his/her involvement and that he/she may choose not to participate if he/she does not want to. Investigator: We are going to do some activities with sounds and words. During these activities you will be looking at listening to, or reading different sounds and words. If you dont want to do this, that is okay. Do you want to do the activities with me? If the student indicates yes the investigator will begin the assessment session. If the student indicates no the investigator will say: Thats okay. Maybe I will come back later to see if you want to do these activities When the investigator returns, the same script will be used. Date: ________________________________ Signature: ________________________________

PAGE 186

186 APPENDIX B QUESTIONNAIRE FORM 1. Childs Legal Name: ____________________________ 2. Date of Birth: ________________ Age: _______ years and ______ months 3. Address: _______________________________________________________________________ Street City/State Zip Code 4. Mailing Address (if different): ____________________________________________________ 5. Person Completing This Form: ___________________ _____________ Name Relationship 6. Phone (Home):(_______) ___________________ (Work): (_______) ___________________ (Cell): ( _______) ___________________ 7. E-mail address (if available) ______________@_____________________________ IDENTIFYING INFORM A TION

PAGE 187

187 1. EDUCATION Name of Current School: ____________________________________________________ Grade: ____________ Teachers Name: _____________________________ Describe the childs progress in school: _________________________________________ ___________________________________________________________________________ Does your child have any problems in reading b ooks or classroom materi als? If yes, please describe in your own words your childs difficulty. Give examples, if possible. ___________________________________________________________________________ ___________________________________________________________________________ 2. SPEECH/LANGUAGE/READING Does your child have any speech and/ or language problems? Yes No If yes, please describe in more detail. ___________________________________________________________________________ ___________________________________________________________________________ Does your child have any reading problem? Yes No Please describe in your own words your childs difficulty. Give ex amples, if possible. ___________________________________________________________________________ ___________________________________________________________________________ Because of this, has your child received speech -language therapy or any remedial reading program at clinic or school? Yes No If yes, please indicate the following: When/where:_________________________________________________________ Dates:_______________________________________________________________ EDUCATIONAL HISTORY & GENERAL HEALTH

PAGE 188

188 Length and frequency of sessions: _______________________________________ Primary focus of therapy: ______________________________________________ How do you feel about your childs reading competence? Excellent Good Fair Slightly Poor Very Poor 3. GENERAL HEALTH Do you feel your child has normal vision? Yes No If not, has the childs visi on been examined? Yes No Describe any illness, accidents, injuries, operations, and/or hospitalizations and include the age of the child: ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________

PAGE 189

189 1. Mothers Name: ___________________ Age: _______ 2. Fathers Name: ____________________ Age: _______ 3. Guardians Name: _________________ Age: _______ 4. Education level of the primary income earn er in the home (Please mark in the box.) Completed elementary school Completed junior high school Received general education degree Completed high school Completed 1 or more years of technical/vocational school Completed technical/vocational school Completed 1 or more years of university/college Bachelors degree Completed 1 or more years of graduate school Masters degree Course work completed for Ph.D., but no dissertation; law degree without bar; medical degree without internship completed Ph.D.; law degree with bar; medica l degree with internship completed 4. Annual Income: $ __________________ (This information is especially very important for successful comparison of the childrens performance and will be kept comple tely confidential in any circumstances.) FAMILY BACKGROUND

PAGE 190

190 Please check your childs hearing status and use an appropriate section for this audiologic history information. If your child has a hearing loss please answer only to the following questions. For children with normal hearing different questions are provided at page 6 of this questionnaire form. Hearing loss 1. When and how did you first notice your childs hearing problem? _______________________________________________________________________ _______________________________________________________________________ 2. When did you first consult an Ear, Nose, and Throat physician or an audiologist since your initial notice? _______________________________________________________________________ 3. If you remember, please indicate the clinic name. _______________________________________________________________________ 4. What was the initial treatment after the physician or the audiologist identified your childs hearing loss? _______________________________________________________________________ _______________________________________________________________________ 5. Do you know the result(s) of initial audi ologic diagnosis? This information may be available from a copy of init ial audiologic evaluation. If you know any one of the followings, please indicate based on exact information. If not sure, just leave it blank. Cause(s) of hearing loss (e.g., cong enital, genetic, developmental) ____________________________________________________________________ Degree of hearing loss ( in dB units; if known) : Right ear ________(dB HL) Left ear ________(dB HL) Type of hearing loss (if known) : sensorineural conductive mixed AUDIOLOGIC HISTORY (For Hearing impaired Child)

PAGE 191

191 6. Does your child currently wear hearing aid(s)? If yes, please describe: Make _____________________ Model ________________________ Both ears_______ Right ear only _______ Left ear only _______ Does the child wear it all time ? Yes No If not, when does your child not wear it? _____________________________ 7. When did your child first start to wear hearing aid(s)? _____________________ Based on this information, for how many years has your child been wearing a hearing aid? __________ years and ___________months 8. Is there a history of hearing loss in the family? Yes No Not sure If yes, which family member(s) have a hearing loss? ________________________ What caused their hearing loss? _________________________________________ _____________________________________________________________________ Speech and Language 1. Do you have any trouble understanding your childs speech? Yes No 2. If yes, briefly provide the difficulties. _______________________________________________________________________ ______________________________________________________________________ 3. Does your child use both oral and sign language at home? Yes No 4. Which one of the followings is your child s major communication mode at home? oral language Sign language 5. Does any parent of your child use sign language fluently? Yes No 6. Which language mode does your childs school use mostly? oral language Sign language

PAGE 192

192 A. History of Ear Infections 1. My child has suffered from ear infections. Yes No 2. My child had more than 3 ear infections between birth and 12 months of age. Yes No 3. My child has had at least one ear infection which lasted more than 3 months. Yes No 4. What was the treatment for the ear infections at that time? __________________________________________________________________ B. Hearing 1. My child has never failed hearing screening(s) in school. Yes No 2. My child has at least once failed a hearing screening in school. Yes No Date of screening (if available): ____________________ 3. Has your child ever been referred to pr ofessionals for hearing evaluation? Yes No 4. If yes, what was the result? ___________________________________________________________________ ___________________________________________________________________ AUDIOLOGIC HISTORY (normal hearing)

PAGE 193

193 APPENDIX C SCORING SHEETS Date: _____/ _____/ 2007 Subject #____ Language a nd Literacy Hearing level _____Normal ______ mild _____ moderate ______ severe ______ RD PTA Left: Right: Tympanometry Oral language area Raw Standard %-ile Age-equi(month) Grade-equi PPVT-III ( ) EVT ( ) CASL ( ) Phonology area raw Stand %-ile Age equi Grade equi Sum standard Composite standard Comp %-ile Elision ( ) Blending ( ) Rapid Digit Naming ( ) Rapid Letter Naming ( ) Memory for Digit ( ) Nonword Repetition ( ) Rapid Object Naming ( ) Digit Ordering WM Raw: Digit Backward WM Raw: Name DOB & age(year/month) ___ /___/19___ (___years /___months) Gender School Grade Testing Location Tester Others

PAGE 194

194 Literacy Age-Norm Grade-Norm area raw stan % AE GE stan % Word ID (WRMT) ( ) Word attck(WRMT) ( ) Pass Comp(WRMT) ( ) Spelling (WRAT) Age-Norm Grade-Norm R-profic R-profici raw stan % Sum stan % stan % Sum stan % AE GE Word effi(TOWRE) word deco(TOWRE) raw stand % Age-equi Grade-equi Passage Reading Rate(GORT-4) Passage Reading Accuracy (GORT-4) GORT fluency TONI ( )

PAGE 195

195 Audiologic Tests Puretone Audiometry (Unaided) 250 Hz 500 Hz 1 K 2K PTA 3K 4K 6K 8K SRT WRS(%) Right Left *SRT: Threshold level for speech (word) recognition (dB HL) *WRS: Word recognition score based the MCL (m ost comfortable hearin g level) % score Type of Hearing loss: Normal : Only Right Ear Only Left Ear Both Ear Normal If not: Left Ear : ____________________________________ Sensorineural hearing loss Right Ear : ____________________________________ Sensorineural hearing loss Tympanometry (Middle ear status): Normal If not, Right Ear: _______________________ Left Ear: ____________________________ SCAN-C/A TEST (Unaided; Using 50 dB HL for normal group and the MCL for HI group) Raw Score Subtests Right Left Sum(L/R) Standard %-ile Filtered Words /20 /20 /40 Figure-ground /40 /40 /80 Competing Words (Dichotic words) /30 /30 /60 Sum of Standard scores Mean of 3 Standard scores Dichotic Digits (Two Digits only) : (Unaided using 50 dB above the PT A for the normal group and the MCL for HI group) Raw Score Subtests Left Right Sum (L/R) Double Digits /40 /40 /80 % Name DOB & age(year/month) ___ /___/19___ (___years /___months) Gender Testing Date

PAGE 196

196 LIST OF REFERENCES Abraham S. (1993). Differential Treatment of Phonological Disability in Children With Impaired Hearing Who Were Trained Orally. American Journal of SpeechLanguage Pathology, 2, 23-30. Abu-Rabia, S. (2002). Phonemic awarene ss and middle-ear disease among Bedouin Arabs in Israel. Reading Psychology, 23, 289-296(8) Adams, M. (1990). Beginning to read: Thinking and learning about print Cambridge, MA: MIT Press. Allen, T. E. (1986). Patterns of academic achie vement among hearing impaired students: 1974 and 1983. In A. N. Schildroth & M. A. Karchmer (Eds.), Deaf Children in America (pp. 161-206) San Diego, CA: College Hill Press. Allen, T. E. (1994). What are the deaf and hard-of-hearing students leaving high school and entering postsecondary education? Unblished ma nuscript. Washington, DC: Gaullaudet University. Allen, T. E., & Schoem, S. R. (1997). Educa ting deaf and hard-of-hearing youth: What works best? Paper presented at the Combined Otolaryngological Spring Meetings of the American Academy of Otolaryngology. Scottsdale, AZ. Anthony, J. L. & Francis, D. (2005). Development of phonological awareness. Current Directions in Psychological Science, 14, 255-259. Archer, A. L., Gleason, M. M., & Vachon, V. L. (2003). Decoding and fluency: Foundation skills for str uggling older readers. Learning Disability Quarterly 26, 89-101. Baddeley, A. D. & Hitch, G. J., (1974). Working memory. In Bower, G.H. (ed.), Recent Advances in Learning and Motivation (Volume 8), New York: Academic Press. Baddeley, A. D. (1986). Working Memory London: Oxford University Press. Baddeley, A. D., Gathercole, S. E., & Papagno, C. (1998). The phonological loop as a language learning device. Psychological Review, 105, 158-173. Baddeley, A.D. (2003) Double dissociat ion: Not magic but still useful. Cortex, 39, 129131. Beitchman, J. H. & Young, A. R. (1997). Learning disorders with a special emphasis on reading disorders: A review of the past 10 years. Journal of the American Academy of Child and Adolescent Psychiatry, 36, 1020 1032.

PAGE 197

197 Bess, F. H., Dodd-Murphy, J., & Parker, R. A. (1998). Children with minimal sensorineural hearing loss: Prevalence, educational performance, and functional status. Ear and Hearing, 19, 339-354. Biddle, K. R. (1996). Timing deficits in impa ired readers: An investigation of visual naming speed and verbal fluency. Unpublished doctoral dissertation, Tufts University, Boston, MA. Blachman B. (1989). Phonological awareness and word recognition: Assessment and intervention. In A. Kamhi & H. Catts (Eds.), Reading disabilities: A developmental language perspective (pp. 133-158). Boston, MA: College-Hill Press. Blachman, B. A. (1991). Early intervention fo r childrens reading problems: Clinical applications of the resear ch in phonological awareness. Topics in Language Disorders, 12, 51-65. Blair, J., Peterson, M., & Viehweg, S. ( 1985). The effects of mild hearing loss on academic performance of young school-age children. Volta Review, 87, 87. Blair, J., Peterson, M., Viehweg, S. (1985). The effects of mild sens orineural hearing loss on academic performance of young school-age children. Volta Review, 87, 87. Blamey, P. J., Sarant, J., Paatsch, L. E., Barr y, J. G., Bow, C. P., Wales, R. J. (2001). Relationships among speech perception, production, language, hearing loss and age in children with impaired hearing. Journal of Speech, Language and Hearing Research, 44, 264. Bowers, P. G. & Wolf, M. (1993). Theoreti cal links among naming speed, precise timing mechanism, and orthogra phic skill in dyslexia. Reading and Writing: An Interdisciplinary Journal, 5, 69-85. Bowers, P. G. (1989). Naming speed a nd phonological awareness: Independent contributors to reading di sabilities. In S. McCormick & J. Zutell (Eds.), Cognitive and social perspectives for literacy resear ch and instruction: 38th Yearbook of the National Reading Conference (pp. 165-173). Chicago: National Reading Conference. Bowers, P. G. (1993). Text reading and rer eading: Determinants of fluency beyond word recognition. Journal of Reading Behavior, 25, 133-53. Bowers, P. G. (1995). Re-examining selected re ading research from the viewpoint of the double-deficit hypothesis. Paper presente d at the Annual Meeting of the Society for Research in Child Development, Indianapolis, IN. Bowers, P. G., & Swanson, L. B. (1991). Naming speed deficits in reading disability: Multiple measures of a singular process. Journal of Experimental Child Psychology, 51, 195-219.

PAGE 198

198 Bowers, P. G., Golden, J. O., Kennedy, A., & Young, A. (1994). Limits upon orthographic knowledge due to processe s indexed by naming speed. In V. W. Berninger (Ed.), The varieties of orthographic knowledge: Theoretical and developmental issues (pp. 173). Dordrecht: Kluwer Academic Publishers. Bowers, P. G., Steffy, R., & Tate, E. (1988) Comparison of the effects of IQ control methods on memory and naming speed predictors of reading disabilities. Reading Research Querterly, 23, 304-309. Bowey, J. A., Cain, M. T., & Ryan, S. M. (1992). A reading-level design study of phonological skills underlying f ourth-grade children's wo rd reading difficulties. Child Development, 63, 999-1011. Bradley, L. & Bryant, P.E. (1983). Categoriz ing sounds and learning to read: A Causal connection. Nature, 30, 419-421. Bradley, L., and Bryant, P.E. (1985). Rhyme and Reason in Reading and Spelling. Ann Arbor: University of Michigan Press. Brady, S. (1991). The role of work ing memory in reading disability In Phonological Processes in Literacy: A Tri bute to Isabelle Y. Liberman Brady, S., Shankweiler, D. (Eds.), Hillsdale, NJ: Lawrence Erlbaum Associates. Brady, S. (1997). Ability to encode phonol ogical representations: An underlying difficulty of poor readers. In Foundations of Reading Acquisition and Dyslexia: Implications for Early Intervention, Blachman, B., et al. (Eds), Mahwah, NJ: Lawrence Erlbaum Associates. Brady, S., Poggie, E., and Rapala, M. M. (1989). Speech repetition abilities in children who differ in reading skill. Language and Speech, 32, 109-22. Brady, S., Shankweiler, D., and Mann, V. A. (1983). Speech perception and memory coding in relation to reading ability. Journal of Experiment al Child Psychology, 35, 345-367. Brandes, P. J., & Ehinger, D. M. (1981). The effects of early middle ear pathology on auditory perception and academic achievement. Journal of Speech and Hearing Disorder, 46, 301-307. Briscoe, J., Bishop, D. V. M., & Norbur y, C. F. (2001). Phonological processing, language, and literacy: A comparison of children with mild-to-moderate sensorineural hearing loss and thos e with specific language impairment. Journal of Child Psychology and Psychiatry, 42, 329-340. Briscoe, J., Gathercole, S. E., & Marlow N. (2001). Everyday memory and cognitive ability in children born prematurely. Journal of Child Psychology and Psychiatry, 42, 749-754.

PAGE 199

199 Brown, P. M. & Brewer, L. C. (1996). Cogniti ve processes of deaf and hearing skilled and less skilled reader. Journal of Deaf Studies and Deaf Education, 1, 263-270. Brown,L, Sherbenou, R. J., Johnsen, S. K. (1997). Test of nonverbal intelligence-3 (TONI-3). Pearson Education, Inc. Bruck, M. (1992). Persistence of dysl exics' phonological awareness deficits. Developmental Psychology, 28, 874-886. Bruck, M., & Treiman, R. (1990). Phonologi cal awareness and spelling in normal children and dyslexics: The case of initial consonant clusters. Journal of Experimental Child Psychology, 50, 156. Bryant, P. E. (1990). Phonological development and reading. In P. D. Pumfrey, & C. D. Elliot (Eds.), Children's difficulties in re ading, spelling and writing Challenges and responses (pp. 63-82). London The Falmer Press. Bryant, P. E., Bradley, L., MacLean, M., & Crossland, J. (1989). Nursery rhymes, phonological skills and reading. Journal of Child Language, 16, 407-428. Byrne, B., & Fielding-Barnsley, R. (1989). Phonemic awareness and letter knowledge in the child's acquisition of the alphabetic principle. Journal of Educational Psychology, 81, 313-321. Byrne, B., & Fielding-Barnsley, R. (1990). Acqu iring the alphabetic principle: A case for teaching recognition of phoneme identity. Journal of Educational Psychology, 82 805-812. Campbell, R., & Wright, H. (1989). Immediate memory n the orally trained deaf: Effects of lipreadability in the recall of written syllables. British Journal of Psychology, 80, 299-312. Carrow-Woolfolk, E. (1999). Comprehensive Assessment of Spoken Language Circle Pines, MN: American Guidance Service. Chall, J. (1996). Learning to Read: The Great Debate (1967). New York: McGraw Hill. Chall, J. S. (1996). Stages of reading development. Fort Worth, TX: Harcourt Brace. Chiat, S. (2001). Mapping theories of de velopmental language impairment: Premises, predictions and evidence. Language and Cognitive Processes 16, 113-42. Coley, J., & Bockmiller, P. (1980). Teaching reading to the deaf: An examination of teacher preparedness and practices. American Annals of the Deaf, 125, 909. Compton, D. L., DeFries, J. C, & Olson, R. K. (2001). Are RANand phonological awareness-deficits additive in chil dren with reading disabilities? Dyslexia, 7, 125149.

PAGE 200

200 Conrad, R. (1977). The reading ab ility of deaf school-leavers. British Journal of Educational Psychology, 47, 138 -48. Cornwall, A. (1992). The relationship of phonological awareness, rapid naming, and verbal memory to severe r eading and spelling disability. Journal of Learning Disabilities, 25, 532-8. Cornwall, A. (1992). The relationship of phonological aware-ness, rapid naming, and verbal memory to severe r eading and spelling disability. Journal of Learning Disabilities, 8, 532-538. Cronin, V. and Carver, P., 1998. Phonological sensitivity, rapid naming, and beginning reading. Applied Psycholinguistics 19 447. Davis, J. M., Effenbein, J., Schum, R., & Be ntler, R. A. (1986). Effects of mild and moderate hearing impairments on langua ge, educational, and psychosocial behaviours of children, Journal of Speech and Hearing Disorder, 51, 53. Davis, J. M., Shepard, N. T., Stelmachowicz, P. G., & Gorga, M. P. (1981). Characteristics of hearing-impaired ch ildren in the public schools. Part II. Psychoeducational data. Journal of Speech and Hearing Disorders, 46, 130-137. Delage, H., & Tuller, L. (2007). Language development and mild-to-moderate hearing loss: does language normalize with age? Journal of Speech, Language, and Hearing Research, 50, 1300-1313. Denckla, M. B. & Cutting, L. E. (1999). Hist ory and significance of rapid automatized naming. Annals of Dyslexia, 49, 29. Denckla, M., & Rudel, R. (1974). Rapid automatized naming of picture objects, colors, letters, and numbers by normal children. Cortex, 10 186-202. Denckla, M., & Rudel, R. (1976). Rapid automatized naming (R.A.N.): Dyslexia differentiated from other learning disabilities. Neuropsychologia, 14 471-479. Dodd, B., & Hermelin, B. (1977). Phonologica l coding by the preli nguistically deaf. Perception and Psychophysics, 21, 413-417. Duara, R., Kushch, A., Gross-Gleen, K. Barker, W. W., Jallad, B., Pascal, S., Loewenstein, D. A., Sheldon, J., Rabin, M. Levin, B., Lubs, H. (1991). Neuroanatomic differences between dys lexic and normal readers on magnetic resonance imaging scans. Archives of Neurology, 48, 410-416. Dunn, L. M., & Dunn, L. M. (1997). Peabody Picture Vocabulary Test-Third Edition: Manual. Circle Pines, MN: American Guidance Services. Dunn, L. M., & Dunn, L. M. (1997). Peabody Picture Vocabulary TestIII. Circle Pines, MN: American Guidance Service.

PAGE 201

201 Ehri, L. C. & Wilce, L. S. (1987). Does lear ning to spell help beginners learn to read words? Reading Research Quarterly, 12, 47-65. Ehri, L. C. (1989). The development of sp elling knowledge and its role in reading acquisition and reading disability. Journal of Learning Disabilities, 22, 349-364. Ehri, L., Nunes, S., Willows, D., Schu ster, B., Yaghoub-Zadeh, Z., & Shanahan, T. (2001). Phonemic awareness instruction he lps children learn to read: Evidence from the National Reading Panels meta-analysis. Reading Research Quarterly, 36, 250. Eimas, P. (1974). Auditory and linguistic pro cessing of cues for plac e of articulation by infants. Perception & Psychophysics, 16, 513-521. Elbro, C. (1996). Early linguistic ability and reading development: A review and a hypothesis about underlying differences in distin ctness of phonological representations of lexical items. Reading and Writing: An Interdisciplinary Journal, 8, 453-85. Elbro, C. (1998). When reading is "readn" or somthn. Distinct ness of phonological representations of lexi cal items in normal and disabled readers. Scandinavian Journal of Psychology. 39 149-53. Elbro, C., & Jensen, M. N. (2005). Qualit y of phonological representations, verbal learning, and phoneme awareness in dyslexi c and normal readers. Scandinavian Journal of Psychology, 46, 375-384. Elbro, C., Borstrm, I., & Petersen, D. K. (1998). Predicting dyslexia from kindergarten: The importance of distinctness of phonologi cal representations of lexical items. Reading Reseaerch Quarterly, 33 36-57. Elbro, C., Nielsen, I., & Peters en, D. K. (1998). Predicting dyslexia from kindergarten: the importance of distinctness of phonological representations of lexical items. Annals of Dyslexia, 44, 205-226. Elfenbein, J. L., Hardin-Jones, M. A., Davis, J. M. (1994). Oral Co mmunication Skills of Children Who Are Hard of Hearing Journal of Speech and Hearing Research, 37, 216 226. Engle, R. W., Cantor, J., & Turner, M. L. ( 1989). Modality effects: Do they fall on deaf ears? Quarterly Journal of Experimental Psychology, 41A, 273-292. Felton, R. H. & Wood, F. (1992). Cognitive deficits in reading deficits in reading disability and attention deficit disorder. Journal of Learning Disabilities, 22, 3-13. Felton, R. H., Wood, F., Brown, I. S., & Campbell, S. K. (1987). Separate verbal memory and naming deficits in attention defi cit disorder and reading disability. Brain and Language, 31, 171-184.

PAGE 202

202 Fletcher, J. M., Shaywitz, S. E., Shankweiler, D. P., Katz, L., Liberman, I. Y., Stuebing, K. K., Francis, D. J., Fowler, A., & Sh aywitz, B. A. (1994). Cognitive profiles of reading disability: Comparisons of disc repancy and low achievement definitions. Journal of Educational Psychology, 85, 1-18. Foorman, B. R., Francis, D. J ., Beeler, T., Winikates, D., & Fletcher, J. M. (1997). Early interventions for children with reading problems: Study designs and preliminary findings. Learning Disabilities: A Multi-Disciplinary Journal 8, 63 Fowler, A. E. (1991). How early phonologica l development might set the stage for phoneme awareness. In Phonological Processes in Lite racy: A Tribute to Isabelle Y. Liberman, Brady, S., Shankweiler, D. (E ds.), Hillsdale, NJ: Lawrence Erlbaum Associates. Fowler, A. E., Doherty, B. J., & Boynton, L. (1995). The basis of reading skill in young adults with Down syndrome. In L. Nadel & D. Rosenthal (Eds.), Down syndrome living and learning in the community (pp. 182). New York: Wiley. Foy, J. G. and Mann, V. (2001) Does streng th of phonological representations predict phonological awareness? Applied Psycholinguistics, 22, 301-325. Foy, J. G., & Mann, V. (2001). Does streng th of phonological representations predict phonological awareness? Applied Psycholinguistics, 22, 301-325. Friel-Patti, S., & Finitzo, T. (1990). Language learning in a prospective study of otitis media with effusion in the first two years of life. Journal of Speech Hearing Research, 33, 188-94. Gathercole, S. E. (2006). Nonword repetiti on and word learning: The nature of the relationship [Keynote]. Applied Psycholinguistics, 27 513. Gathercole, S. E., & Baddeley, A. D. (1990). Phonological memory deficits in language disordered children: Is there a causal connection? Journal of Memory and Language, 29, 336-360. Gathercole, S. E., Hitch, G. J., Service, E. & Martin, A. J. (1997). Phonological shortterm memory and new word learning in children. Developmental Psychology 33 966-979. Gibbs, S. (2004). The skills in reading shown by young children with permanent and moderate hearing impairment. Educational Research, 46, 17-27. Gilbertson, M., & Kamhi, A.G. (1995). Novel word learning in children with hearing impairment. Journal of Speech and Hearing Research, 38, 630-642. Gillon, G. T. (2004). Phonological Awareness: from research to practice New York: Guilford Press.

PAGE 203

203 Goldin-Meadow, S., & Mayberry, R. I. (2001). How do profoundly deaf children learn to read? Learning Disabilities Research & Practice, 16, 222-229. Goswami, U. (2000). Phonological and Lexical Processes. In Barr, R., Kamil, M., Mosenthal, P., and Pearson, D. (2000) Handbook of Reading Research, 3, 251268. Goswami, U. (2000). Phonological representati ons, reading development and dyslexia: Towards a cross-linguistic theoretical framework. Dyslexia, 6, 124 -132. Goswami, U. (2001). Developmental dyslexia. In N. J. Smelser & P. B. Baltes (Eds.), International encyclopedia of th e social & behavioral sciences (pp. 3918-3921). Amsterdam: Elsevier Sciences. Goswami, U. (2003). Why theories about de velopmental dyslexia require developmental designs. Trends in Cognitive Sciences, 7, 534-54. Goswami, U., & Bryant, P (1990). Phonological skills and learning to read. Hove, East Sussex: Lawrence Erlbaum Associates Ltd. Gottardo, A., Stanovich, K. & Siegel, L. (1996). The relationships between phonological sensitivity, syntactic processing and ve rbal working memory in the reading performance of third-grade children. Journal of Experiment al Child Psychology, 63, 563-582. Griffiths, Y. M., & Snowling, M. J. (2002). Predictors of exception word and nonword reading in dyslexic children: the severity hypothesis. Journal of Educational Psychology, 94, 34-43. Grigorenko, E. L. (2001). Developmental dyslexia: An update on genes, brains, and environments. Journal of Child Psychology and Psychiatry, 42 91-125. Grushkin, D. A. (1998). Why shouldnt Sam r ead? Toward a new paradigm for literacy and the deaf. Journal of Deaf Studies and Deaf Education. 3, 179-204. Gupta, P. (2003). Examining the relationshi p between word learning, nonword repetition, and immediate serial recall in adults. Quarterly Journal of Experimental Psychology, 56, 1213-1236. Halliday, L. F. & Bishop, D. V. (2006) Audito ry frequency discrimination in children with dyslexia, Journal of Research in Reading, 29, 213. Halliday, L. F., & Bishop, D. V. (2005). Freque ncy discrimination and literacy skills in children with mild to moderate sensorineural hearing loss. Journal of Speech, Language, and Hearing Research, 48, 1187-1203. Hansen, J., & Bowey, J. (1994). Phonological analysis skills, verbal working memory, and reading ability in second-grade children. Child Development, 65 938-50.

PAGE 204

204 Hanson, V. L. (1982). Short-term recall by deaf signers of American Sign Language: Implications for order recall. Journal of Experimental Psychology: Learning, Memory, and Cognition, 8 572. Hanson, V. L. (1989). Phonology and reading: Ev idence from profoundly deaf readers. In D. Shankweiler & I. Lieberman (Eds .), Phonology and reading disability: Solving the reading puzzle (pp. 69-89). Ann Arbor, MI: University of Michigan Press. Hanson, V. L., & Fowler, C. A. (1987). Phonolog ical coding in word reading: Evidence from hearing and deaf readers. Memory and Cognition, 15, 199-207. Hanson, V. L., & Lichtenstein, E. (1990). S hort-term memory coding by deaf signers: The primary language coding hypothesis reconsidered. Cognitive Psychology, 22, 211-224. Hanson, V. L., & McGarr, N. S. (1989). Rhyme generation by deaf adults. Journal of Speech and Hearing Research, 32, 2-11. Hanson, V. L., Goodell, E. W., & Perfetti, C. A. (1991). Tongue twister effects in the silent reading of hearing and deaf college students. Journal of Memory and Language, 30, 319. Hansson, K., Forsberg, J., Lofqvist, A., Ma ki-Torkko, E., & Sahlen, B. (2004). Working memory and novel word learning in ch ildren with hearing impairment and children with specific language impairment. International Journal of Language and Communication Disorders, 39, 401 Harris, M., & Beech, J. R. (1998). Implic it phonological awareness and early reading development in prelingually deaf children Journal of Deaf Studies and Deaf Education, 3, 205-216. Hartas, D. (2005). Language and Communication Difficulties, London: Continuum. Hasenstab, M. S., & McKenzie, C. D. (1981). A survey of reading programs used with hearing-impaired students. The Volta Review, 83, 383. Hayiou-Thomas, M. E., Bishop, D. V. M. & Pl unkett, K. (2004). Simulating SLI: General cognitive processing stressors can pr oduce a specific linguistic profile. Journal of Speech, Language, and Hearing Research 47, 1347. Hemmer, V.H. & Ratner, N.B. (1994). Communicative development in twins with discordant histories of recurrent otitis media. Journal of Communication Disorders, 27, 91-106. Hickok, G., & Poeppel, D. (2000). Toward s a functional neuroanatomy of speech perception. Trends in Cognitive Sciences, 4, 131.

PAGE 205

205 Hoffmeister, R. J. (2000). A piece of the puzzl e: ASL and reading comprehension in deaf children. In C. Chamberlain, J. P. Morford, & R. I. Mayberry (Eds.), Language Acquisition by Eye (pp. 165-89). Mahwah, NJ: Earlbaum. Hoien, T., Lundberg, L., Stanovich, K. E., & Bjaalid, L. K. (1995). Components of phonological awareness. Reading & Writing: An Inte rdisciplinary journal, 7, 171188. Holt, J. (1994). Classroom attributes and achie vement test scores for deaf and hard of hearing students. American Annals of the Deaf, 139, 430-437. Hulme, C., & Snowling, M. ( 1992). Phonological deficits in dys lexia: a reappraisal of the verbal deficit hypothesis. In N. Singh & I. Beale (Eds.), Current perspectives in learning disabilities (pp. 270-331). New York: Springer-Verlag. Hurford, D. P., Darrow, L. J., Edwards, T. L ., Howerton, C. J., Mote, C. R., Schauf, J. D., & Coffey, P. (1993). An examination of phonemic processing abil ities in children during their first-grade year. Journal of Learning Disabilities, 26, 167-177. Hynd, G. W., Semrud-Clikemand, M., Lorys, A. R., Novey, E. S., & Eliopulos, D. (1990). Brain morphology in developmenta l dyslexia and attention deficityhyperactivity disorder (ADHD): Morphometric analysis of MRI. Archives of Neurology, 47, 919-926. Jarrold, C., Baddeley, A. D., & Phillips, C. E. (2002). Verbal short-term memory in Down syndrome: A problem of memory, audition, or speech? Journal of Speech,Language, and Hearing Research, 45, 531. Jenkins, R., & Bowen, L. (1994). Facilitating development of preliterate childrens phonological abilities. Topics in Language Disorders, 14, 26-39. Justas, S., & Wilkinson, G. S. (1984). Wide Range Achievement TestRevised. Wilmington, DE: Justas Associates. Kail, R. (1991). Developmental change in speed of processing during childhood and adolescence. Psychological Bulletin, 109 490. Kail, R., & Hall, L. K. (1994). Procesisng speed, naming speed, and reading. Developmental Psychology, 30, 949. Kail, R., Hall, L. K., & Caskey, B. J. (1999) Processing speed, exposure to print, and naming speed. Applied Psycholinguistics, 20, 303. Katz, R. (1986). Phonological deficiencies in children with reading disability: Evidence from an object-naming task. Cognition, 22, 225-257. Katzir, T., Wolf, M., OBrien, Kennedy, B., Lovett, M., & Morris, R. (2006). Reading Fluency: The Whole is more than the Parts. Annals of Dyslexia. 56, 51-82.

PAGE 206

206 Keith, R. W. (1986). SCAN: A Screening Test for A uditory Processing Disorders San Antonio, TX: The Psychological Corporation, Harcourt, Brace, Jovanovich, Inc. Keith, R. W., Rudy, J., Donahue, P. A., & Katbamna, B. (1989). Comparison of SCAN results with other audito ry and language measures in a clinical population. Ear and Hearing, 10, 382. King, C.M., & Quigley, S.P. (1985). Reading and deafness. San Diego, CA: College-Hill Press. Kramer, J. H., Knee, K., & Delis, D. C. (2000) Verbal memory impairments in dyslexia. Archieves of Clinical Neuropsychology. 15, 83-93. Kushch, A., Gross-Glenn, K., Jallad, B., Lubs H., Rabin, M., Feldman, E., & Duara, R. (1993). Temporal lobe surface area meas urements on MRI in normal and dyslexic readers. Neuropsychologia, 31, 811-821. Landerl, K., Wimmer, H., & Fr ith, U. (1997). The impact of orthographic consistency on dyslexia: a German-English comparison. Cognition, 63, 315-334. Larsen, J. P., Hoien, T., Lundberg, I., & Od egaard, H. (1990). MRI evaluation of the size and symmetry of the planum temporale in adolescents with developmental dyslexia. Brain Lanugage, 39, 289-301. LaSasso, C. (1978). A national survey of materi als and procedures used to teach reading to hearing-impaired children. American Annals of the Deaf, 123 22. LaSasso, C. (1987). Survey of reading instruct ion for hearing impaired students in the United States. The Volta Review, 89 85. LaSasso, C., & Mobley, R. T. (1997). National Su rvey of reading instruction for deaf or hard-of-hearing students in the U.S. The Volta Review, 99, 31. Laybaert, J., & Charlier, B. (1996). Visual sp eech in the head: The e ffect of Cued-Speech on rhyming, remembering, and spelling. Journal of Deaf Studies and Deaf Education, 1, 234-248. Leonard, C. M., Eckert, M. A ., Lombardino, L. J., Oakland, T., Kranzler, J., Mkohr, C. M., King, W. M., & Kreeman, A. (2001) Anatomical risk factors for phonological dyslexia. Cerebral Cortex, 11, 148-157. Leonard, C. M., Puranik, C., Kuldau, J. M., & Lombardino, L. J. (1998). Normal variation in the frequency and location of human auditory cortex landmarks: Heschl's gyrus: Where is it? Cerebral Cortex, 8 397-406.

PAGE 207

207 Leonard, C. M., Voeller, K. K., Lombar dino, L. J., Morris, M. K., Hynd, G. W., Alexander, A. W., Andersen, H. G. Garofalakis, M., Honeyman, J. C., Mao, J., Agee, F. & Staab, E. V. (1993). Anomalous cerebral structure in dyslexia revealed with magnetic resonance imaging. Archives of Neurology, 50, 461-469. Leonard, L. (1989). Language learnability and specific language impa irment in children. Applied Psycholinguistics, 10, 179. Leonard, L. (1995). Functional categories in the grammars of children with specific language impairment. Journal of Speech and Hearing Research, 38, 1270. Leonard, L. (1998). Specific Language Impairment. MIT Press, Cambridge, MA. Levelt, W. J. M. (1989). Speaking: from intent ion to articulation. Cambridge, Massachusetts: The MIT Press. Leybaert, J. (1993). Reading in the deaf : The roles of phonological codes. In M. Marschark & M. D. Clark (Eds.), Psychological perspect ives on deafness (pp. 269-309). Hillsdale, NJ: Lawrence Erlbaum Associates. Leybaert, J. (1998). Phonologica l representations in deaf children: the importance of early lingu istic experience. Scandinavian Journal of Psychology, 39, 169-173. Leybaert, J., & Alegria, J. (1990). Cued sp eech and the acquisition of reading by deaf children. The Cued Speech Journal, 4, 24-38. Liberman, I. Y., Shankweiler, D., & Liberm an, A. M. (1989). The alphabetic principle and learning to read. In D. Sha nkweiler & I. Y. Lieberman (Eds.), Phonology and reading disability: Solv ing the reading puzzle (pp. 1-33). Ann Arbor: University of Michigan Press. Lichtenstein, E. H. (1998). The relationshi ps between reading pr ocesses and English skills of deaf college Students. Journal of Deaf Studies and Deaf Education, 3, 80-134. Lieberman, I. Y., & Shankweiler, D. (1985). Phonology and problems of learning to read the write. Remedial and Special Education, 6 8-17. Lindsay, P. H., Shapiro, A., Musselman, C ., & Wilson, A. (1988). Predicting language development in deaf children using subscales of the Leiter International Performance Scale. Canadian Journal of Psychology, 42, 144-162. Lovett, M. W., Steinbach, K. A., & Frijters, J. C. (2000). Remediating the core deficits of developmental read ing disability: a doubledeficit perspective. Journal of Learning Disabilities, 33, 334-358. Lyon, G. R. (1998). Why reading is not a natural process. Educational Leadership, 55, 419.

PAGE 208

208 Lyon, G. R., Shaywitz, S. E. & Shaywitz, B. A. (2003). A definition of dyslexia. Ann Dyslexia, 53, pp. 1-14. Lyxell, B., Anderson, U., Borg, E., & Ohlss on, I. S. (2003). Working-memory capacity and phonological processing in deafened a dults and individuals with a severe hearing impairment. International Journal of Audiology, 42, Suppl 1:S86-89. Lyxell, B., Rnnberg, J. & Samuelsson, S. (1994). Internal speech functioning and speechreading in deafened and normal hearing adults. Scandinavian Audiology, 23, 179-185. Majerus, S., Amand, P., Boniver, V, Demanez, J. P., Demanez, L., & Linden, M. V. (2005). A quantitative and qualitative a ssessment of verbal short-term memory and phonological processing in 8-year-olds with a history of repetitive otitis media. Journal of Communication Disorders, 9, 1-26. Manis, F. R., Doi, L. M., & Bhadha, B. (2000). Naming speed, phonological awareness, and orthographic knowledge in second graders Journal of Learning Disabilities, 33, 325. Manis, F. R., McBride-Chang, C., Seidenberg, M. S., Keating, P., Doi, L. M., Munson, B., & Petersen, A. (1997). Are speech pe rception deficits associated with developmental dyslexia? Journal of Experiment al Child Psychology, 66, 211-235. Mann, V. A. (1993). Phoneme awareness and future reading ability. Journal of Learning Disabilities, 26, 259-269. Mann, V. A. (1998) Language problems: A key to early reading problems. In B. Wong (Ed.) Learning about Learning disabilities, 2nd Ed (pp. 163-202). Chicago. Marchman, V. A. & Bates, E. (1994). C ontinuity in lexical and morphological development: A test of the critical mass hypothesis. Journal of Child Language, 21, 339-366. Marschark, M. & Harris, M. (1996). Success and failure in learning to read: the special case of deaf children. In C. Cornoldi & J. Oakhill (Eds.), Reading comprehension difficulties: Processes and intervention (pp 279 300). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Marschark, M. (1997). Raising and educating a deaf child New York: Oxford University Press. Marschark, M. (2003). Cognitive functioning in deaf adults and children. In M. Marschark & P. E. Spencer (Eds.), Oxford handbook of deaf studies, language, and education (pp. 464). New York: Oxford University Press. Marschark, M., & Everhart, V. S. (1999). Problem solving by deaf and hearing children: Twenty questions. Deafness and Education International 1, 63.

PAGE 209

209 Marschark, M., & Lukomski, J. (2001). Unde rstanding language and learning in deaf children. In M. D. Clark, M. Ma rschark, & M. Karchmer (Eds.), Cognition, context, and deafness (pp. 71). Washington, DC: Gallaudet University Press. McConkey Robbins, A., Burton Koch, D., Osberger, M.J., Zimmerman-Phillips, S., Kishon-Rabin, L. (2004). Effect of age at cochlear implantation on auditory skill development in infants and toddlers. Archives of Otolaryngology, Head, and Neck Surgery, 130, 570-574. McDougall, S., Hulme, C., Ellis, A., & Monk, A. (1994). Learning to read: The role of short-term memory and phonological skills. Journal of Experimental Child Psychology, 58, 112. Merzenich, M. M., Jenkins, W. M., Johnson, P ., Schreiner, C., Miller, S., & Tallal, P. (1996). Temporal processing deficits of language-learning impaired children ameliorated by training. Science, 271, 77-80. Metsala, J. L. (1997). Spoken word recognition in reading in readi ng disabled children. Journal of Educational Psychology, 89, 159-169. Metsala, J. L. (1999). Young childrens phonol ogical awareness a nd nonword repetition as a function of vocabulary development Journal of Educational Psychology, 91, 3-19. Metsala, J. L., & Walley, A. C. (1998). S poken vocabulary growth and the segmental restructuring of lexical representations: precursor s to phonemic awareness and early reading ability. In J. L. Metsala & L. C. Ehri (Eds.), Word Recognition in Beginning Literacy, 89-120. Hillsdale, NJ: Erlbaum. Misra M., Katzir, T., Wolf, M., & Poldrac k, R. A. (2004). Neural systems of rapid automatized naming in skilled readers: Unraveling the RAN-reading relationship. Scientific Studies of Reading, 8, 241-256. Misra, M., Katzir, T., Wolf, M. & Poldr ack, R.A. (2004). Neural systems for rapid automatized naming (RAN) in skilled readers: Unraveling the RAN-reading relationship. Scientific Studies in Reading. 8, 241-256. Mody, M. (2003). Rapid auditory processing deficits in dyslexia: A commentary on two differing views. Journal of Phonetics, 31, 529-539. Mody, M., Studdert-Kennedy, M., & Brady, S. ( 1997). Speech perception deficits in poor readers: Auditory processing or phonological coding? Journal of Experimental Child Psychology, 64, 199 -231. Mody, R., Schwartz, G., Gravel, J.S., & Ruben, R. J. (1999). Speech perception and verbal memory in children with and without histories of otitis media. Journal of Speech, Language, and Hearing Research, 42, 1069-1079.

PAGE 210

210 Moeller, M. P. (2000). Early intervention a nd language development in children who are deaf and hard of hearing. Pediatrics, 106, 43-51. Moeller, M. P., & Luetke-Stahlman, B. (1990) Parents use of signing exact English: A descriptive analysis. Journal of Speech and Hearing Disorders, 55, 327-38. Moncrieff, D. W. (2002). Auditory Pr ocessing Disorders and Dyslexic Children. http://www.audiologyonline.com/articles/ pf_article_detail.asp? article_id=369. Moore, B. C .J. (1995). Perceptual Consequences of Cochlear Damage, Oxford University Press, Oxford. Morals, J., Alegria, J., & Content, A. (1987). The relationships between segmental analysis and alphabetic lite racy: An interactive view. Cahiers de Psychologie Cognitive, 7, 415-38. Morford, J. P., & Mayberry, R. I. (2000). A reexamination of early exposure and its implication for language acquisition by eye. In C. Chamberlain, J. P. Morford, & R. I. Mayberry (Eds.), Language Acquisition by Eye (pp. 165-89). Mahwah, NJ: Earlbaum. Morris, R. D., Stuebing, K. K, Fletcher, J. M ., Shaywitz, S. E., Lyon, G. R., Shankweiler, D. P. (1998). Subtypes of reading disability: Variability around a Phonological Core. Journal of Educational Psychology, 90, 347. Musiek, F. E., & Chermak, G. D. (1994). Th ree commonly asked quest ions about central auditory processing disorders: Assessment. American Journal of Audiology, 3 23. Musiek, F. E., & Reeves, A. G. (1990). As ymmetries of the aud itory areas of the cerebrum. Journal of the American Academy of Audiology, 1, 240-245. Musselman, C. (2000). How do children who can't hear learn to read an alphabetic script? A review of the literature on reading and deafness. Journal of Deaf Studies and Deaf Education 5, 9. Muter, V., Hulme, C., Snowling, M., & Ta ylor, S. (1997), Segmentation, not rhyming, predicts early progress in learning to read. Journal of Experimental Child Psychology, 65, 370. Nairne, J. S. (1990). A feature model of immediate memory. Memory & Cognition, 18 251. Nathan, L., Stackhouse, J., Goulandris, N., & Snowling, M. (2004). The development of early literacy skills among ch ildren with speech difficulties: A test of the critical age hypothesis. Journal of Speech, Language, and Hearing Research, 47 377391.

PAGE 211

211 Neath, I., & Nairne, J. S. (1995). Word -Length effects in immediate memory: Overwriting trace decay theory. Psychonomic Bulletin & Review, 2, 429-441. Nemeth, S. (1992). The role of phonology and context in word r ecognition in hearing impaired readers. Unpublished masters thesis. Montreal, Quebec, Canada: McGill University. Neuhaus, G. F., Foorman, B. R., Francis, D. J., & Carlson, C. (2001). Measures of information processing in Rapid Automati zed Naming (RAN) and their relation to reading Journal of Experimental Child Psychology, 78 359-373. Nicholson, R. I., & Fawcett, A. J. (1990) A New Framework for Dyslexia Research? Cognition, 35, 159-182. Nielsen, D. C., & Luetke-Stahlman, B. ( 2002). Phonological Awareness: One key to the reading proficiency of deaf children. American Annals of the Deaf, 147 11-19. Nittrouer, S. (1996). Discriminability and per ceptual weighting of some acoustic cues to speech perception by 3-year-olds Journal Speech and Hearing Research, 39 278-297. Nittrouer, S. (1996). The relation between speech perception and phonemic awareness: Evidence from low-SES children and children with chronic OM. Journal of Speech and Hearing Research. 39, 1059-1070. Nittrouer, S., & Burton, L. T. (2005). The ro le of early language experience in the development of speech perception and phonological processing abilities: Evidence from 5-year-olds with histories of otitis media with effusion and low socioeconomic status. Journal of Communication Disorders, 38, 29-63. Norbury, C. F., Bishop, D. V. M., & Briscoe, J. (2001). Production of English finite verb morphology: A comparison of SLI and m ild-moderate hearing impairment. Journal of Speech, Language, and Hearing Research, 44, 165. Nozza, R. J., Bluestone, C. D., Kardatzke, D., Bachman, R. (1994). Identification of middle ear effusion by aural acoustic admittance and otoscopy. Ear and Hearing, 15, 310-23. OConnor, R. E., Jenkins, J. R., Leiceste r, N., & Slocum, T. A. (1993). Teaching phonological awareness to young children with learning disabilities. Exceptional Children, 59 532-546. Oakhill, J. & Kyle, F. (2000). The Relati on between Phonological and Working Memory. Journal of Experiment al Child Psychology, 75, 152. O'Brien, B., Wolf, M., Morris, R., & Lovett, M. (2004). An Investigation of subtypes of developmental dyslexia using taxometric classification. Annual Meeting Brain, Neurosciences, and Educ ation SIG Paper Session.

PAGE 212

212 Olson, R. Forsberg, H., Wise, B., & Rack, J. (1994). Measurement of word recognition, orthographic, and phonological sk ills. In G. R. Lyon (Ed.), Frames of reference for the assessment of learnintg disab ilities: New views on measurement issues (pp. 243-277). Baltimore: Brookes. Ostrander, C. (1998). The relationship between phonological coding and reading achievement in deaf children: Is cued speech a special case? Unpublished manuscript. Padden, C., & Ramsey, C. (2000). Americans Si gn Language and readin g ability in deaf children. In C. Chamberlain, J. P. Morford, & R. I. Mayberry (Eds.), Language Acquisition by Eye (pp. 165-89). Mahwah, NJ: Earlbaum. Paul, P. V. (1998). Literacy and deafness. Boston: Allyn and Bacon. Paul, P. V. (2001). Language and deafness. San Diego, CA: Singular. Paulesu, E., Dmonet, J., Fazio, F., McCror y, E., Chanoine, V., Brunswick, N., Cappa, S., Cossu, G., Habib, M., Frith, C., Frith, U. (2001). Dyslexia : cultural diversity and biological unity, Science 291 2165-2167. Penhune, V. B., Zatorre, R. J., MacDonald, J. D., & Evens, A. C. (1996). Interhemispheric anatomical differences in human primary auditory cortex; probabilistic mapping and vol ume measurement from magnetic resonance scans. Cerebral Cortex, 6, 661-672. Perfetti, C. A. (1985). Reading ability. New York: Oxford University Press. Perfetti, C. A.(1992). The representation probl em in reading acquisition. In P. Gough, L. Ehri, & R. Treiman (Eds.), Reading acquisition (pp. 107-143). San Diego: Acadmeic Press. Perfetti, C. A., & Sandak, R. (2000). R eading optimally builds on spoken language: Implications for deaf literacy. Journal of Deaf Studies and Deaf Education, 5, 3150. Peters, S. A., Grievink, E. H., van Bon, W. H., & Schilder, A. G. (1994). The effects of early bilateral otitis media with effusion on educational attainment: a prospective cohort study. Journal of Learning Disabilities. 27, 111-21. Plapinger, D. S., & Sikora, D. M. (1995). Th e Use of Standardized Test Batteries in Assessing the Skill Development of Children With Mild-to-Moderate Sensorineural Hearing Loss. Language, Speech, and Hearing Services in Schools, 26, 39-44. Porpodas, C. D. (1999). Patterns of phonologi cal and memory processing in beginning readers and spellers of Greek. Journal of Learning Disabilities, 32, 406-416.

PAGE 213

213 Powell, D ., Stainthorp, R ., Stuart, M, Garwood, H & Quinlan, P. (2007). An experimental comparison between rival theories of ra pid automatized naming performance and its relationship to reading. Journal of Experimental Child Psychology, 98 46-68. Ramus, F. (2001a). Dyslexia Talk of two theories. Nature, 412, 393-395. Ramus, F. (2001b). Outstanding ques tions about phonologi cal processing. Dyslexia, 7, 197-216. Ramus, F. (2002). The neural basis of r eading acquisition. In M. S. Gazzaniga (Ed.), The Cognitive Neurosciences (3rd ed., pp. 815-824). Cambridge, MA: MIT Press. Ramus, F. (2003). Developmental dyslexia: specific phonological deficit or general sensorimotor dysfunction? Current Opinion in Neurobiology, 13, 212-218. Ramus, F., Pidgeon, E., & Frith, U. (2003). Th e relationship between motor control and phonology in dyslexic children. Journal of Child Psychology and Psychiatry, 44 (5), 712-722. Ramus, F., Rosen, S., Dakin, S. C., Day, B. L., Castellote, J. M., White, S., & Frith, U. (2003). Theories of developmental dyslexi a: Insights from a multiple case study of dyslexic adults. Brain, 126, 1-25. Rayner, K., Foorman, B. R., Perfetti, C. A., Pesetsky, D. & Seidenberg, M. S. (2001). How psychological science informs the teaching of reading Psychological Science in the Public Interest, 2 31-74. Rice, M., & Wexler, K. (1996). Toward Tense as a clinical marker of specific language impairment in English-speaking children, Journal of Speech and Hearing Research 39, 1239-1257. Roberts, J. E., Burchinal, M. R., & Zeisel, S. A. (2002). Otitis m edia in early childhood in relation to children's school-age language and academic skills. Pediatrics, 110, 696-706. Roberts, J. E., Hunter, L., Gravel, J., Rose nfeld, R., Berman, S., Haggard, M., Hall, J., Lannon, C., Moore, D., Lynne, V., & Wallace, I. (2004). Otitis media, hearing loss, and language learning: cont roversies and current research. Journal of Developmental & Behavioral Pediatrics, 25, 110 -122. Roberts, J. E., Rosenfeld, R. M., & Zeisel S. A. (2004). Otitis media and speech and language: a meta-analysis of prospective studies. Pediatrics, 113, 238-48. Rom, A., & Leonard, L. (1990). Interpreting deficits in grammatical morphology in specifically language-impaired children: Preliminary evidence from Hebrew. Clinical Linguistics and Phonetics, 4, 93-105.

PAGE 214

214 Rosen, S. (2003) Auditory processing in dyslexia and specific language impairment: Is there a deficit? What is its na ture? Does it e xplain anything? Journal of Phonetics, 31, 509-527. Ruben, R. J. (1972). The ear. In H. Barnett & A. Einhorn (Eds.) Pediatrics (15th ed., pp. 1881-92). New York: Appleton, Century, Crofts. Rutter, M., & Yule, W. (1975). The con cept of specific r eading retardation. Journal of Child Psychology and Psychiatry 16,181-197. Scanlon, D. M. & Vellutino, F. R. (1997). A co mparison of the instructional backgrounds and cognitive profiles of poor, average, and good readers who were initially identified as at risk for reading failure. Scientific Studies of Reading, 1, 191-215. Schirmer, B. R. (1994). Language and Literacy Development in Children Who are Deaf. Merrill: Merrill/ Macmillan Publishing Company. Schirmer, B. R., & McGough, S. M. (2005). T eaching reading to children who are deaf: Do the conclusions of the Na tional Reading Panel apply? Review of Educational Research, 75 83-117. Service, E., Maury, S., Luotoniemi, E. (2005). Forgetting and redintegration of consonants and vowels in pseudoword lists. Memory, 13, 340. Shankweiler, D., & Crain, S. (1986) Langua ge mechanisms and reading disorder: A modular approach. Cognition, 24, 139-168. Share, D. L. (1995). Phonological recoding a nd self-teaching: Sine qua non of reading acquisition. Cognition, 55, 151. Share, D. L. (1999). Phonological recoding and orthographic learning: A direct test of the self-teaching hypothesis. Journal of Experiment al Child Psychology, 72, 95. Share, D. L., Chalmers, D., Silva P. A., & St ewart, I. A. (1986). Reading disability and middle ear disease. Archives of Disease in Childhood, 61, 400-401, Sharma, A., Dorman, M., & Spahr, A. (2002). A sensitive period for the development of the central auditory system in children with cochlear implants: Implications for age of implantation, Ear & Hearing 23 532. Shaywitz, S. (1998). "Curre nt concepts: Dyslexia." The New England Journal of Medicine, 338, 307-312. Shaywitz, S. (2003). Overcoming dyslexia: A new and co mplete science-based program for reading problems at any level. New York: Knopf.

PAGE 215

215 Shaywitz, S. E., Fletcher, J. M., Holahan, J. M., Shneider, A. E., Marchione, K. E., Stuebing, K. K., et al. (1999). Persis tence of dyslexia: The Connecticut Longitudinal Study at adolescence. Pediatrics, 104, 1351-1359. Shaywitz, S., & Shaywitz, B. (2005). Dysl exia (Specific Reading Disability). Biological Psychiatry, 57, 1301. Shaywitz, S., Fletcher, J., Holahan, J., Shneid er, A., Marchione, K., Stuebing, K., et al. (1999). Persistence of dyslexia: Th e Connecticut Longitudinal Study at adolescence. Pediatrics, 104(6), 1351. Siok, W. T., & Fletcher, P. (2001). The ro le of phonological aw areness and visualorthographic skills in Chinese reading acquisition. Developmental Psychology, 37, 886-899. Snow, C. E., Burns, M. S., & Griffin, P. (1998). Preventing Reading Difficulties in Young Children. Washington, DC: National Academy Press. Snowling, M. & Hulme, C. (1989). A long itudinal case study of developmental phonological dyslexia. Cognitive Neuropsychology 6 379. Snowling, M. J. (1998). Reading de velopment and its difficulties. Educational and Child Psychology, 15, 44-58. Snowling, M. J. (2001). From language to reading and dyslexia. Dyslexia, 7, 37-46. Snowling, M. J., Goulandris, N., Bowlby, M ., & Howell, P. (1986). Segmentation and speech perception in relati on to reading skill: a developemental analysis. Journal of Experimental Child Psychology, 41 489-507. Snowling, M., Bishop, D. V. M., & Stothar d, S. E. (2000). Is pre-school language impairment a risk factor for dyslexia in adolescence? Journal of Child Psychology and Psychiatry, 41, 587-600. Snowling, M., Nation, K., Moxham, P., Gallagher, A., & Frith, U. (1997). Phonological processing skills of dyslexic students in higher education: A preliminary report. Journal of Research in Reading, 29, 31-41. Spector, J. (1995). Phonemic awareness training: Application of prin ciples of direct instruction. Reading and Writing Quarterly, 11 37-51. Stanovich, K. E. (1980). Toward an inter active-compensatory model of individual differences in the development of reading fluency. Reading Research Quarterly, 16, 32-71 Stanovich, K. E. (1994). Annota tion: Does dyslexia exist? Journal of Child Psychology and Psychiatry, 35, 579.

PAGE 216

216 Stanovich, K. E. (1998). Twenty-five years of research on the readi ng process: The grand synthesis and what it means for our field. In T. Shanahan & F. Rodriguez-Brown (Eds.), Forty-Seventh Yearbook of th e National Reading Conference (pp. 44-58). Chicago: NRC. Stanovich, K. E., & Siegal, L. S. (1994). Phe notypic performance profile of children with reading disabilities: A regresson-based test of the phonological-core variabledifference model. Journal of Educational Psychology, 86, 24-53. Stanovich, K.E. (1981). Relationships betw een word decoding speed, general nameretrieval ability, and reading pr ogress in first-grade children Journal of Educational Psychology, 73 809-815. Stanovich, K.E. (1986). Matthew Effects in Re ading: Some consequences of individual Differences in the acquisition of literacy. Reading Research Quarterly, 21, 360407. Stanovich, K.E. (1988). Explai ning the differences between the dyslexic and the Gardenvariety poor reader: The phonological-co re variable difference modes. Journal of Learning Disabilities, 21, 590-612. Stanovich, K.E. (1992). Speculations on the causes and consequen ces of individual Differences in early reading acquisition. In P. Gough, L. Ehri, and R. Treiman (Eds.), Reading acquisition (pp. 307-342). Hillsdale, NJ: Lawrence Erlbaum Associates. Stanovich, K.E. (1993). Does reading make you smarter? Literacy and the development of verbal intelligence. In H. Reese (Ed.), Advances in Child Development and Behavior (Vol. 24, pp. 133-180). San Diego, CA: Academic Press. Stanovich, K.E., and West, R.F. (1989). E xposure to print and or thographic processing. Reading Research Quarterly, 24, 402-433. Stelmachowicz, P.G., Pittman, A.L., Hoover, B.M., & Lewis, D.E. (2004). Novel-word learning in children with normal hearing and hearing loss. Ear and Hearing 25, 47-56. Sterne, A., & Goswami, U. (1998). Deaf ch ildren can make rhymes: evidence from a simple rhyme task. Surdite et acces a la langue ecrite. 137-155. Strong, M., & Prinz, P. (2000). Is American Sign Language skill related to English literacy? In C. Chamberlain, J. P. Morford, & R. I. Mayberry (Eds.), Language Acquisition by Eye ( pp. 165-89). Mahwah, NJ: Earlbaum. Svirsky, M. A., Robbins, A. M., Kirk, K. I ., Pisoni, D. B., & Miyamoto, R. T. (2000). Language development in profoundly deaf children with cochlear implants. Psychological Science, 11, 153-158.

PAGE 217

217 Svirsky, M. A., Stallings, L. M., Lento, C. L., Ying, E., & Leonard, L. B. (2002). Grammatical morphologic development in pa ediatric cochlear implantation users may be affected by the perceptual prominence of the relevant markers. Annals of Otology Rhinology and Laryngology, 189(Suppl.), 109. Swan, D., & Goswami, U. (1997a). Phonologica l awareness deficits in developmental dyslexia and the phonological re presentations hypothesis. Journal of Experimental Child Psychology, 66, 18 -41. Swan, D., & Goswami, U. (1997b). Picture-na ming deficits in developmental dyslexia: The phonological represen tations hypothesis. Brain and Language, 56, 334-353. Tallal, P. (1980). Auditory temporal perception, phonics, an d reading disabilities in children. Brain and Language, 9, 182. Tallal, P., Miller, S. L., Bedi, G., Byma, G ., Wang, X., Nagarajan, S. S., Schreiner, C., Jenkins, W. M., & Merzerich, M. M. (1996). Language comprehension in language-learning impaired children impr oved with acoustically modified speech. Science, 271, 81-84. Tallal, P., Miller, S., & Fitch, R. (1993). Ne urological basis of speech: A case for the prominence of temporal processing? In P. Tallal, A. Galaburda, R. Llinas, & C. von Euler (Eds.), Annals of the New York Academy of Sciences: Vol. 682. Temporal information processing in th e nervous system (pp. 27-37). New York: New York Academy of Sciences. Tijms, J. (2004). Verbal memory a nd phonological processing in dyslexia. Journal of Research in Reading, 27, 300-310. Torgesen, J. K. (1997). The prevention a nd remediation of reading difficulties: Evaluating what we know from research. Journal of Academic Language Therapy, 1, 11. Torgesen, J. K. (2000). Individual differences in response to early interventions in reading: The lingering problem of treatment resisters. Learning Disabilities Research and Practice, 51 (1), 55. Torgesen, J. K., & Bryant, B. (1994). Test of Phonological Awareness Austin, TX: PROEd. Torgesen, J. K., & Burgess, S. R. (1998). Consistency of readi ng-related phonological processes throughout early childhood: Evid ence from longitudinal-correlational and instructional studies. In J. Metsala & L. Ehri (Eds.). Word Recognition in Beginning Reading. Hillsdale, NJ: Lawrence Erlbaum Associates. Torgesen, J. K., & Mathes, P. G. (2000). A basic guide to understanding, assessing, and teaching phonological awareness. Austin, TX: Pro-Ed.

PAGE 218

218 Torgesen, J. K., & Wagner, R. K. (1998). Alte rnative diagnostic a pproaches for specific developmental reading disabilities. Learning Disabilities Research and Practice, 13, 220-232. Torgesen, J. K., Rashotte, C. A., & Wagne r, R. K. (1994). Longitudinal studies of phonological processi ng and reading. Journal of Learning Disabilities, 27, 276 286. Torgesen, J. K., Wagner, R. K., & Rashotte, C. A. (1997a). The prevention and remediation of severe read ing disabilities: Keeping the end in mind. Scientific Studies of Reading, 1, 217-234. Torgesen, J. K., Wagner, R. K., Rashotte, C. A., Burgess, S., & Hecht, S. (1997b). Contributions of phonological awareness a nd rapid automatic naming ability to the growth of word-reading skills in secondto fifth-grade children. Scientific Studies of Reading, 1, 161-185. Torgesen, J.K., Wagner, R. K., & Rashotte, C.A. (1999). Test of Word Reading Efficiency .Austin, TX: PRO-ED Publishing, Inc. Treiman, R., & Bourassa, D. (2000). Children's written and oral spelling. Applied Psycholinguistics, 21, 183-204. Treiman, R., & Bourassa, D. (2000). The development of spelling skill. Topics in Language Disorders, 20, 1-18. Treiman, R., & Hirsh-Pasek, K. (1983). Silent reading: Insights from second-generation deaf readers. Cognitive Psychology, 15, 39-65. Trezek, J., & Malmgren, K. W. (2005). The Efficacy of Utilizing a Phonics Treatment Package with Middle School Deaf and Hard-of-Hearing Students. Journal of Deaf Studies and Deaf Education, 10, 256 271. Vellutino, F. R. (1979). Dyslexia: Theory and research. Cambridge, MA: MIT Press. Vellutino, F. R., Fletcher, J.M., Snowling, M. J., & Scanlon, D. M. (2004). Specific reading disability (dyslexia) : What have we learned in the past four decades? Journal of Child Psychology and Psychiatry, 45, 2. Vellutino, F. R., Scanlon, D. M., Sipay, E. R., Small, S. G., Pratt, A., Chen, R. S., & Denckla, M. B. (1996). Cognitive profiles of difficult to remediate and readily remediated poor readers: Early interven tion as a vehicle for distinguishing between cognitive and experiential deficits as basic causes of specific reading disability. Journal of Educati onal Psychology, 88, 601-638. Wagner, R. K., Torgesen, J. K., & Rashotte, C. A. (1994). Development of readingrelated phonological processing abilities: New evidence of bidirectional causality from a latent variable longitudinal study. Developmental Psychology, 30, 73-87.

PAGE 219

219 Wagner, R. K., Torgesen, J.K., & Rashotte, C.A. (1999 ). Comprehensive Test of Phonological Processes Austin, TX: PRO-ED Publishing, Inc. Wagner, R. K., Torgeson, J. K., Laughon, P. L., Sommons, K., & Rashotte, C. A. (1993). Development of young readers' phonol ogical processing abilities. Journal of Educational Psychology, 85, 83 -103. Wagner, R., & Torgesen, J. (1987). The natu re of phonological proces sing and its causal role in the acquisition of reading skills. Psychological Bulletin, 101 192-212. Wake, M., & Poulakis, Z. (2004). Slight and m ild hearing loss in prim ary school children. Journal of Paediatric Child Health, 40, 11-13. Wake, M., Hughes, E., Poulakis, Z., Collins, C., & Rickards, W. (2004). Outcomes of children with mild-profound congenital hearing impairment at 7-8 years: a population study. Ear and Hearing, 25, 1-8. Walley, A. (1993). The role of vocabulary de velopment in childrens spoken word recognition and segmentation ability. Developmental Review, 13, 286-350. Wandel, J. (1989). Use of Internal Speech by Hearing and Hearing-Impaired Students in Oral, Total Communication, and Cued Speech Programs. Unpublished Doctoral Dissertation, Columbia University, New York. Wesseling, R. F., & Reitsma, P. (1998). Phonemically aware in a hop, skip, and a jump. In P. Reitsma & L. Verhoeven (Eds.) Problems and interventions in literacy development (pp. 81-94). Dordrecht: Kluwer. Wiederholt, J. L., & Bryant, B. R. (2001). Gray Oral ReadingTests, Fourth Edition Austin, TX: Pro-Ed. Williams, K. T. (1997). Expressive Vocabulary Test Circle Pines, MN: American Guidance Service. Leonard, L. (1998). Children with specific language impairment Cambridge, MA: MIT Press. Wimmer, H. (1993). Characteristics of de velopmental dyslexia in a regular writing system. Applied Psycholinguistics, 14, 1-33. Wolf, M. & Bowers, P. G. (2000). The question of naming-speed deficits in developmental reading disability: An introduction to the double-deficit hypothesis. Journal of Learning Disabilities, 33, 322-324. Wolf, M. (1997). A provisional, integrat ive account of phonological naming speed deficits in dyslexia: Implications for di agnosis and interventi on. In B. Blachman (Ed.). Foundations of reading acquisition and dyslexia (pp. 67-93). Mahwah, NJ: Lawrence Erlbaum Associates.

PAGE 220

220 Wolf, M., & Bowers, P. G. (1999). The doubledeficit hypothesis for the developmental dyslexia. Journal of Educational Psychology, 91, 124. Wolf, M., Bowers, P. G., & Biddle, K. (2000). Naming-speed processes, timing, and reading: A conceptual review. Journal of Learning Disabilities, 33, 387-407. Wolf, M., Goldberg ORourke, A., Gidney, C., Lovett, M., Cirino, P., & Morris, R. (2002). The second deficit: An investigat ion of the independence of phonological and naming-speed deficits in developmental dyslexia Reading and Writing, 15 43-72. Woodcock, R. W. (1987). Woodcock reading mastery testRevised. Circle Pines, MN:American Guidance Service. Woodcock, R. W., & Johnson, M. B. (1989). Woodcock Johnson Psycho-Educational Battery-Revised. Allen, TX: DLM Teaching Resources. Woodcock, R.W. (1987). Woodcock Reading Mastery TestsRevised American Guidance Service, Circle Pines, MN. Wren, S. (2005). The double-deficit hypot hesis for decoding fluency. Personal manuscript. Yopp, H. K. (1988). The validity and reliabi lity of phonemic awareness tests. Reading Research Quarterly, 23, 159-177. Yopp, H. K. (1992). Developing phone mic awareness in young children. Reading Teacher, 45, 696-703. Yoshinaga-Itano, C. (1999). Benefits of early intervention for children with hearing loss. Otolaryngology and Clinic of North America, 32, 1089-1102. Yoshinga-Itano, C., Sedey, A. L., & Coulter, D. (1998). Language of earlyand lateridentified children with hearing loss. Pediatrics, 102, 1161-1171. Ziegler, J. C. (2006). Do differences in brai n activation challenge universal theories of dyslexia? Brain and Language, 98, 341.

PAGE 221

221 BIOGRAPHICAL SKETCH Jungjun Park was born and raised in Seoul, Korea. Jungjun earn ed his Bachelor of Arts degree in 1990 and a Master of Arts degree in Linguistics in 1994 from the Seoul National University. After his doctoral coursework in Li nguistics at the Seoul Nati onal University in 1997 followed by four years of teaching experience, Jungjun began his doctoral work at the Department of Communication Scie nces and Disorders at the Univ ersity of Florida in August 2001 with a concentration in r eading disorder and child langu age development. In August 2007, Jungjun will join the faculty of the Department of Communication Sciences and Disorders at the Baylor University in Waco, Texa s as an assistant professor.