The Use of Pseudowords for Decoding Instruction

MISSING IMAGE

Material Information

Title:
The Use of Pseudowords for Decoding Instruction
Physical Description:
1 online resource (181 p.)
Language:
english
Creator:
Madsen, Keri Marie
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Special Education, Special Education, School Psychology and Early Childhood Studies
Committee Chair:
LANE,HOLLY BARNES
Committee Co-Chair:
JONES,HAZEL
Committee Members:
SINDELAR,PAUL T
MILLER,DAVID

Subjects

Subjects / Keywords:
decoding -- instruction -- pseudowords -- reading
Special Education, School Psychology and Early Childhood Studies -- Dissertations, Academic -- UF
Genre:
Special Education thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Children who fall behind in reading during the elementary years often remain poor readers, and the effects can be detrimental to future success in school and in life.  A primary deficit experienced by many struggling readers is poor development of skills that lead to decoding and word recognition.  Efforts to promote the development of these foundational reading skills are critical, particularly for at-risk and struggling readers.  Studies have shown that interventions that teach decoding, including word work with manipulatives, are effective at improving reading and spelling outcomes in at-risk or struggling readers.  However, there is still much to learn about what specific aspects of instruction make these interventions effective.   This study serves to extend our understanding of decoding development by exploring the role of pseudowords in decoding instruction for beginning readers.  Specifically, this study examines the effects of incorporating pseudowords during decoding instruction on the decoding accuracy and automaticity performance of kindergarteners who were not yet demonstrating full-alphabetic word recognition.  An experimental pre-test post-test group design was used to compare the decoding outcomes of three groups of students: (a) a treatment group receiving 15 sessions of word work instruction with real words and pseudowords, (b) a comparison group receiving 15 sessions of word work instruction with real words only, and (c) a business-as-usual control group.  Participants were assessed at pre- and post-test using a battery of researcher-developed and standardized assessments of decoding accuracy and automaticity. Analysis of covariance (ANCOVA) was used to test for statistical difference between post-test group means while controlling for pre-test scores.  Results indicate statistically significant differences between groups on researcher-developed measures of real-word accuracy, real-word automaticity, and pseudoword automaticity, with follow-up comparisons indicating both treatment and comparison conditions performed significantly better at post-test than students in the control group.  No significant difference was found between treatment and comparison groups on these measures.  There was also a group by covariate interaction for on the Word Attack subtest of the Woodcock Reading Mastery Tests.  There were no significant differences found for a pseudoword accuracy measure or for the Phonemic Decoding Efficiency subtest of the Test of Word Reading Efficiency.  Overall, the results suggest that incorporating pseudowords during decoding instruction with kindergarteners in the early stages of decoding development is just as effective as using real words alone.  Implications for practice and directions for future research are discussed.
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 Keri Marie Madsen.
Thesis:
Thesis (Ph.D.)--University of Florida, 2014.
Local:
Adviser: LANE,HOLLY BARNES.
Local:
Co-adviser: JONES,HAZEL.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2014
System ID:
UFE0045936:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 THE USE OF PSEUDOWORDS FOR DECODING INSTRUCTION By KERI MARIE MADSEN 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 2014

PAGE 2

2 2014 Keri M. Madsen

PAGE 3

3 This work is dedicated to my mom one of the strongest, most intelligent and hardworking women I will ever know. You have inspired me to push myself beyon d what was even imaginable, and I simply could not have made it this far without you. I also dedicate this to the memory of my dad. It fills my heart t o know how proud you would be if you were here today.

PAGE 4

4 ACKNOWLEDGMENTS First, I wish to thank Dr. Holly Lane for giving me this awesome opportunity and offering me tireless patience and support throughout this long and often difficult journey. Her mentorship was paramount to my growth as a doctoral student as well as an educator, a researcher, and a writer. My success in this doctoral program can be credited largely to h er dedication to her students, and I feel extremely grateful to be fortunate enough to have had Holly as my mentor and the chair of my committee. In addition, I wish to thank the members of my committee, Dr. Hazel Jones, Dr. David Miller, and Dr. Paul Sindelar. I respect each of these individuals immensely for their leadership and scholarship and I appreciate their commitment to providing me with guidance and support throughout my doctoral program. Each offered valuable insight, direction, and feedback that has truly helped me grow as a researcher and a writer. I must also thank the wonderful staff in the School of Special Education, School Psychology, and Early Childhood Studies at the University of Florida. Elizabeth, Lynette, Michell, Shaira, and Vicki have provided continuous assistance and support throughout every aspect of the doctoral pr ogram, ensuring that the entire process moved along flawlessly. These ladies are invaluable and I have grown to appreciate each one of them for their suppor t, hard work, and the friendships we have developed. Several individuals also deserve acknowledgement and gratitude for their direct support and assistance with my dissertation work. First, I would like to thank Kari Cronin for her positive attitude and complete willingness to assist in any way possible. Kari served as a liaison between the school site and me, facilitating the initial contact and arranging all of the details so I could begin my work. Without her support, my study certainly would not have run as smoothly. I also wish to thank the principal, Valde

PAGE 5

5 Fortner, and all of the kindergarten teachers at my school site for their gracious hospitality and their flexibility during the data collection and intervention of my study. These educators were impressive on many levels, and I am so thankful I had the opportunity to work alongside of them. I must also thank my fellow doctoral students, colleagues, and friends Carrie, Gayle, Julie, Lourdes, Shaunt, and Vivian, each who donated their precious time and effort to help me complete hours of kindergarten assessments Lourdes also conducted fidelity observations and provided meticulous editorial services. These ladies deserve enormous gratitude because, without their selfless acts, I would never have achieved this monumental goal. I must thank the fabulous ladies in my cohort as well Candy, Lourdes, Mary Anne, and Pam for putting up with me and my antics all along the way. These women provided a critical support system to me, and I want them to know how lucky I feel to have shared this amazing experience with t hem. Finally, I must show my appreciation for a few very special people in my life who have provided much needed patience, love, and understanding during my doctoral program. First, I am grateful to Shane for seeing me through this long and diffi cult chapter of my life. Lastly, I want to thank my mom, Paulette and my stepdad, Al, who have sacrificed in enormous ways so I could pursue this goal Without the support of all of these individuals, this achievement would have been insurmountable.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ... 4 LIST OF TABLES ................................ ................................ ................................ ............. 9 LIST OF FIGURES ................................ ................................ ................................ ......... 10 ABSTRACT ................................ ................................ ................................ .................... 11 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ..... 13 Research Problem and Professional Significance ................................ ................... 15 Theoretical Framework ................................ ................................ ............................ 19 Conclusion ................................ ................................ ................................ ............... 22 2 REVIEW OF RELATED LITERATURE ................................ ................................ .... 24 Ehri's Phases of Word Recognition Development ................................ ................... 25 Pre alphabetic Phase ................................ ................................ ........................ 26 Partial alphabetic Phase ................................ ................................ ................... 27 Full alphabetic Phase ................................ ................................ ........................ 28 Consolidated alphabetic Phase ................................ ................................ ........ 29 Automatic Phase ................................ ................................ ............................... 29 The Critical Transition to Skilled Decoding ................................ .............................. 30 Problems Learning to Decode ................................ ................................ ................. 33 Instruction that Promotes Decoding ................................ ................................ ......... 34 Phonemic Awareness Instruction ................................ ................................ ...... 34 Phonics Instruction ................................ ................................ ............................ 37 Intervention to Develop Decoding Skills in Struggling Readers ............................... 39 Phonics with phoneme level instruction ................................ ............................ 40 Multisensory word work ................................ ................................ ..................... 41 Using Pseudowords ................................ ................................ .......................... 48 Instructional Time ................................ ................................ .............................. 49 Group Size ................................ ................................ ................................ ........ 52 Instructor Variables ................................ ................................ ........................... 54 Summary and Conclusions ................................ ................................ ...................... 56 Implications for Research ................................ ................................ ........................ 57 3 METHODS ................................ ................................ ................................ ............... 60 Hypotheses ................................ ................................ ................................ .............. 60 Setting ................................ ................................ ................................ ...................... 61 Participants ................................ ................................ ................................ .............. 62

PAGE 7

7 Measures ................................ ................................ ................................ ................. 65 Screening Measure ................................ ................................ ........................... 65 Pre test/Post test Measures ................................ ................................ .............. 67 Assessment Procedure ................................ ................................ ............................ 70 Intervention ................................ ................................ ................................ .............. 71 Treatment Condition ................................ ................................ .......................... 72 Comparison Condition ................................ ................................ ....................... 76 Control Condition ................................ ................................ ............................... 77 Instructional Fidelity ................................ ................................ ................................ 78 Analysis ................................ ................................ ................................ ................... 78 4 RESULTS ................................ ................................ ................................ ................ 82 Pretest Findings ................................ ................................ ................................ ....... 83 Analysis of Hypothesized Intervention Effect ................................ ........................... 85 Results for Decoding Accuracy ................................ ................................ ......... 86 Real Word Accuracy ................................ ................................ ......................... 86 Pseudoword Accuracy ................................ ................................ ....................... 88 Woodcock Reading Mastery Tests Word Attack ................................ ............ 89 Results for Decoding Automaticity ................................ ................................ ........... 91 Real word automaticity ................................ ................................ ...................... 91 Pseudoword Automaticity ................................ ................................ .................. 93 Test of Word Reading Efficiency Phonemic Deco ding Efficiency .................. 95 Summary ................................ ................................ ................................ ................. 96 5 DISCUSSION ................................ ................................ ................................ .......... 98 Summary of Findings ................................ ................................ ............................... 99 The Effects of Decoding Instruction with Pseudowords on Real Word Accuracy ................................ ................................ ................................ ...... 100 The Effects of Decoding Instruction with Pseudowords on Pseudoword Accuracy ................................ ................................ ................................ ...... 101 The Effects of Decoding Instruction with Pseudowords on Real Word Automaticity ................................ ................................ ................................ 102 The Effects of Decoding Instruction with Pseudowords on Pseudowor d Automaticity ................................ ................................ ................................ 102 Discussion and Interpretation ................................ ................................ ................ 103 Limitations ................................ ................................ ................................ .............. 108 Implications for Practice ................................ ................................ ......................... 109 Implications for Future Research ................................ ................................ ........... 111 Conclusions ................................ ................................ ................................ ........... 115 APPENDIX A IRB APPROVAL DOCUMENATION ................................ ................................ ...... 117 B REAL WORD ACCURACY PROTOCOLS AND STIMULUS PLATES ................. 118

PAGE 8

8 C PSEUDOWORD ACCURACY PROTOCOLS AND STIMILUS PLATES .............. 132 D REAL WORD AUTOMATICITY PROTOCOLS AND ST IMULUS PLATES ........... 146 E PSEUDOWORD AUTOMATICITY PROTOCOLS AND STIMULUS PLATES ...... 154 F WORDS FOR TREATMENT LESSONS ................................ ............................... 162 G WORDS FOR COMPARISON LESSONS ................................ ............................. 164 H FIDELITY INSTRUMENT FOR TREATMENT CONDITION ................................ .. 166 I FIDELITY INSTRUMENT FOR COMPARISON CONDITION ............................... 168 LIST OF REFERENCES ................................ ................................ .............................. 170 BIOGRAPHICAL SKETCH ................................ ................................ ........................... 180

PAGE 9

9 LIST OF TABLES Table page 3 1 Demographic information for sample by group ................................ ................... 64 4 1 Descriptive statistics for pre tests by group ................................ ......................... 84 4 2 One way ANOVA for pre test means ................................ ................................ .. 85 4 3 Summary of analysis of covariance for R Acc ................................ ..................... 87 4 4 Pairwise comparisons for R Acc ................................ ................................ ......... 8 8 4 5 Adjusted group means for R Acc ................................ ................................ ........ 88 4 6 Summary of analysis of covariance for P Acc ................................ ..................... 89 4 7 Adjus ted group means for P Acc ................................ ................................ ......... 89 4 8 Summary of analysis of covariance for WRMT WA ................................ ............ 90 4 9 Summary of analysis of covariance for R Auto ................................ ................... 92 4 10 Pairwise comparisons for R Auto ................................ ................................ ........ 93 4 11 Adjusted group means for R Auto ................................ ................................ ....... 93 4 12 Summary of analysis of covariance for P Auto ................................ ................... 94 4 13 Pairwise comparisons for P Auto ................................ ................................ ........ 94 4 14 Adjusted group means for P Auto ................................ ................................ ....... 95 4 15 Summary of analysis of covar iance for TOWRE PDE ................................ ........ 96 4 16 Adjusted group means for TOWRE_PDE ................................ ........................... 96

PAGE 10

10 LIST OF FIGURES Figure page 4 1 Interaction between group and pre test covariate for WRMT WA. ...................... 90

PAGE 11

11 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 THE USE OF PSEUDOWORDS FOR DECODING INSTRUCTION By Keri Marie Madsen May 2014 Chair: Holly B. Lane Major: Special Education Children who fall behind in reading during the elementary years often remai n poor readers, and the eff ects can be detrimental to future success in school and in life. A primary deficit experienced by many struggling readers is poor development of skills that lead to decoding and word recognition Efforts to promote the development of these foundational reading skills are critical particularly for at-risk and struggling readers. Studies have shown that interventions that teach decoding, including word work with manipulatives, are effective at improving reading and spelling outcomes in at risk or struggling readers. However, ther e is still much to learn about what specific aspects of instruction make these interventions effective. This study serves to extend our under standing of decoding development by exploring the role of pseudowords in decoding instruction for beginning reade rs. Specifically, this study examines the effects of incorporating pseudowords during decoding instruction on the decoding accuracy and automaticity performance of kindergarteners who were not yet demonstrating full -alphabetic word recognition. An experimental pre-test post-test group design was used to compare the decoding outcomes of three groups of students: (a) a treatment group receiving 15 sessions of

PAGE 12

12 word work instruction with real words and pseudowords, (b) a comparison group receiving 15 sessions of word work instruction with real words only, and (c) a businessas-usual control group. Participants were assessed at preand post-test using a battery of researcher-developed and standardized assessments of decoding accuracy and automaticity. Analysis of covariance (ANCOVA) was used t o test for statistical difference between post -test group means while controlling for pre -test scores. Results indicate statistically significant differences between groups on researcher-developed measures of real-word accuracy, real-word automaticity, and pseudoword automaticity, with f ollowup comparisons indicating both treatment and comparison conditions performed significantly better at post-test than students in the control group. No significant difference was found between t reatment and comparison groups on these measures. There was also a group by covariate interaction for on the Word Attack subtest of the Woodcock Reading Mastery Test s. There were no significant differences found for a pseudoword accuracy meas ure or for the Phonemic Decoding Efficiency subtest of the Test of Word Reading Efficiency. Overall, the results suggest that incorporating pseudowords during decoding instruction with kindergarteners in the early stages of decoding development is just as effective as using real words alone. Implications for practice and directions for future research are discussed.

PAGE 13

13 CHAPTER 1 INTRODUCTION Reading is among the most essential skills to develop during the early school years. Without the ability to read, students have limited access to the content of every other academic subject Unfortunately, children who fall behind in reading during the elementary years often remain poor readers throughout school (Juel, 1988; Snow, Burns & Griffin, 1998; Stanovich, 1986), with as many as 70% of older students struggling with reading (Biancarosa & Snow, 2004) Once these reading trajectories are established, they are difficult to change (Francis, Shaywitz, Stuebing, Shaywitz, & Fletcher, 1996; Good, Baker, & Peyton, 2009). Moreover, reading failure is likely to lead to negative consequences such as grade retention, dropout, limited employment opportunities, and difficulties with basic life activities (Lyon, 2001). Clearly, the longterm effects of early reading difficulties can be devastating. Efforts to promote the development of foundational reading skills are critical for beginning and at -risk or struggling readers. To be skillful at reading, one must be able to extract meaning from written language; however, understanding print is dependent upon the abilities to first recognize the symbols of print and convert these symbols into meaningful units of oral language (Gough & Tunmer, 1986). When a reader has learned to perform these foundational tasks, he or she is able to decode print. Decoding ability is an early and strong predictor of future success in reading (Juel, 1988); yet, the primary deficit identified in many struggling readers is poor development of skills that lead to decoding and word recognition (Chall, 1983 ; Henry, 1993; Torgesen, 1999). Fortunately, cumulative research spanning across the past three

PAGE 14

14 decades has provided educators with a better understanding of effective instruction to promote these foundational literacy skills. Reports generated by the National Research Council (NRC, Snow et al., 1998) and the National Reading Panel (NRP, 2000) have examined a sizable body of evidence on effective early reading instruction and made recommendations for practice. The NRP report focused on five core areas of reading: phonemic awareness, phonics, fluency with connected text, vocabulary and language development, and comprehension of text. While instruction in each area is important for success in reading, phonemic awareness and phonics instruction contribute most to the development of decoding, and this is especially true for students who are struggling to learn how to read (Adams, 1990; Snow et al. 1998; NRP 2000 ). Decoding instruction may be approached in a variety of ways, but word work is one instructional method that is used widely to help make explicit connections between phonemes (i.e., the discrete sounds of speech) and graphemes (i .e., the single letters or letter combinations that represent individual phonemes) through multisensory techniques using manipulative materials. Throughout the literature, methods to promote decoding often make use of magnetic letters (Campbell, Helf, & C ooke, 2008; OConnor & Jenkins, 1995; Tunmer & Hoover, 1993); letter tiles or cards (Blachman, Tangel, Ball, Black, & McGraw, 1999; Blachman et al., 2004; Ehri & Wilce, 1987a; McCandliss, Beck, Sendak, & Perfetti, 2003); colored chips or blocks (Linan-Thompson & Hickman-Davis, 2002; Uhry & Shepherd, 1993); and Elkonin boxes ( Joseph, 1998 2000a, 2000b, 2002a; Maslanka & Joseph, 2002; Ryder, Tunmer, & Greaney, 2008; Tunmer & Hoover, 1993; Vadasy, Sanders, & Peyton, 2005). During word work, materi als such as these

PAGE 15

15 are used to manipulate (i.e., delete, substitute, add, blend, and/or segment) phonemes within words, thereby connecting an otherwise abstract auditory task with a concrete visual-kinesthetic task for enhanced memory and learning (Campbell et al. 2008). Researchers have shown that decoding interventions that include the use of word work with manipulatives are beneficial for at -risk and struggling readers (Campbell et al., 2008; Iversen & Tunmer, 1993; Joseph, 1998; Lane et al., 2009; McCandless et al., 2003; Pullen, 2000; Pullen et al., 2005). Furthermore, word work with manipulatives has been identified as an integral component within a multi -component literacy intervention (Lane et al., 2009) as well as an effective independent interven tion ( Joseph, 2000a, 2000b ; Pullen, 2000). Although studies have identified word work as an effective instructional method for teaching decoding to beginning or struggling readers, very few studies have examined what features of word work contribute to ef fectiveness. The present study intended to extend the research on word work by investigating an aspect of words used for instruction. Specifically, this study examined the role of pseudowords in decoding instruction with beginning readers. Currently, th ere is limited research on the use of pseudowords for instructional purposes despite theoretical support for potential benefits in the development of decoding skill. Research Problem and Professional Significance Studies have shown that interventions incl uding word work are effective at improving reading and spelling outcomes in at -risk or struggling readers; however, research has provided limited insight into what makes word work more or less effective. Most published studies that include word work, do s o as part of a multi-component literacy intervention; and, of the few that have isolated word work as an independent

PAGE 16

16 variable (i.e., Lane et al., 2009; Pullen, 2000), none have sought to examine the type of words used during word work. It is common for teachers, researchers, and developers of phonics programs to place emphasis on the use of real words or even recommend against the use of pseudowords for decoding instruction. This practice is largely based upon conventional wisdom, or a reasonable belief t hat practice with real and meaningful words will promote the reading of real and meaningful text (Adams, 2011). Yet, using pseudowords during decoding instruction may prove to be beneficial for students who are just beginning to learn how to decode. In E nglish, pseudowords, sometimes called nonsense words or non -words, consist of a pronounceable string of letters that conforms to the orthographic patterns of the language but holds no meaning (Pseudowords, n. d.). To illustrate, wub, flot and straff adhere to the spelling patterns of English and are pronounceable but have no meaning in the English language. Pseudowords are often used in psycholinguistic studies of lexical processing to examine or compare the neural pathways evoked by words (Cibelli, 2012). In studies of word reading, theories of processing have been proposed in which pseudowords utilize either separate processing pathways from real words (i.e., the dual-route theories; see Coltheart et al., 1993) or shared processing pathways with real words (i.e., the connectionist theories; see Seidenberg & McClelland, 1989). These theories have provided a foundation upon which much of the current literature on reading has been built (Perfetti, Van Dyke, & Hart, 2001). Pseudoword reading has long been regarded as a reliable predictor of reading proficiency (Gough, 1983 ). As such, pseudoword reading is widely used in the measurement of decoding (e.g., Ehri et al., 2007; Pullen et al., 2005; Shankweiler et al.,

PAGE 17

17 1999; Snowling, 1981; Uhry & Shepherd, 1993). According to Good, Baker, and Peyton (2009), the rationale for using pseudowords as a measure of decoding skill is derived from the extensive research base on learning to read in an alphabetic writing system, such as English (p. 35). These researchers suggest that skills that are necessary for decoding are isolated from the ability to read words by sight during pseudoword reading because the reader cannot rely on past experience with a pseudoword. Instead, the reader relies on their understandi ng of the letter -sound relationships (i.e., the alphabetic principle) and the rules that govern those relationships and then applies this knowledge to decode. A similar rationale can be proposed for using pseudowords in decoding instruction, particularly for students who are in the beginning phases of learning to decode. For beginners, the use of pseudowords during instruction limits the possibility to rely upon memory for a word and instead forces the novice reader to apply knowledge of the alphabetic principle to decode (Cardenas, 2009). While it is possible that knowledge of real words may assist a more advanced reader to identify pseudowords through analogy, beginners will not have a large enough lexicon or the skills necessary to read pseudowords thi s way (Ehri, 2005). Therefore, pseudoword reading ensures that beginning readers are truly tapping into phonemic awareness and application of the alphabetic principle skill areas specific to decoding and, once learned, these skills can be applied to a ny word. Moreover, pseudowords offer increased opportunities to practice decoding through exposure to more letter and sound combinations than would be provided with real words alone. As such, experience with pseudoword reading may increase independent de coding success and generalization of

PAGE 18

18 decoding skill to a wider variety of unfamiliar words. Because of these potential benefits, research that examines the use of pseudowords for decoding instruction has the potential to contribute to improved practices in decoding instruction. A few researchers have mentioned including pseudowords in studies of decoding intervention, and participants who have been exposed to these interventions have improved on measures of reading (e.g., Ehri & Wilce, 1987b; Pullen et al., 2005), but only one study has specifically examined the effects of pseudoword instruction (Cardenas, 2009). In her dissertation, Cardenas investigated whether or not a phonics curriculum comprised only of pseudowords would increase kindergarteners success in phonetic decoding. This researcher found that students who were provided pseudoword phonics for 28 school days in place of real -word phonics made greater gains during and after the treatment than students who received only real -word phonics. The findings of this study offer preliminary support for the use of pseudowords during phonics instruction with beginning readers. This present study serves to extend the examination of pseudowords to te ach decoding to beginning readers by comparing decoding outcomes of three groups of students: those who receive word work with real words and pseudowords, those who receive word work with real words only, and those who receive no word work at all. Broadly, this study examined the effects of incorporating pseudowords during decoding instruction on the decoding development of kindergarteners who are at risk for reading difficulties. Specifically, the following four research questions were addressed: 1. What a re the effects of decoding instruction that includes pseudowords on real word decoding accuracy? 2. What are the effects of decoding instruction that includes pseudowords on pseudoword decoding accuracy?

PAGE 19

19 3. What are the effects of decoding instruction that includes pseudowords on real word decoding automaticity? 4. What are the effects of decoding instruction that includes pseudowords on pseudoword decoding automaticity? Theoretical Framework This study draws on three different theories: Connectionist Theory, Ehri s Phases of Word Reading, and the theory of Recombinative Generalization. The following section outlines the key tenets of each theory and how each relates to this study. Theorists of reading development largely agree on at least one thing the ultimate goal of reading is to comprehend written language. While it is true that reading comprehension involves the synergy of many complex mental processes, the complexities of reading can be simplified into two primary parts: word recognition and linguistic comprehension (Gough & Tunmer, 1986). Each part is critically important to reading ability, but linguistic comprehension will serve little value if printed words are not first recognized. In order to recognize and ultimately understand printed words, multiple sources of information must be processed. As Adams (1990) proposes in her model of parallel distributed processing of skilled reading, word recognition involves a system of associative orthographic, phonological, semantic, and co ntextual processing. This model is connectionist in nature in that these four processors work in mutual coordination to send and receive information to and from one another (Adams, 2001). Visual observation of the spelling of a word will activate the system at the level of t he orthographic processor, but recognition of the word will require associative connections with one or more of the other processors. The semantic processor will be activated with

PAGE 20

20 the readers accumulated knowledge of the words meaning; the phonological processor will be activated with the pronunciation of the word; and the context processor will be activated when the word has been primed or can be anticipated from within context. Early phonemic activation during word recognition is known to play a majo r role in the processing of letter strings (Perfetti & Bell, 1991; Perfetti, Bell, & Delaney, 1988; Ziegler & Jacobs, 1995). Thus, the formation of strong connections between the phonological and orthographic processors is foundational to word recognition The more frequently these processors exchange information, the stronger the connections become, and the stronger the connections between any two processors, the more automatic and accurate word recognition will be (Adams, 1990). The development of wor d recognition is often explained through developmental phase theories. Although many developmental theories of word recognition have been proposed in the literature, the work of Linnea Ehri (1991, 1998, 2005) generates one of the most widely recognized and accepted. Through years of research, Ehri proposes five phases through which readers pass during the development of word recognition. During the pre-alphabetic phase, beginners have no working knowledge of the alphabet ic system and are able to identify commonly encountered environmental words based upon non-alphabetic salient visual features alone. As these beginners learn a limited number of letters and corresponding sounds and use this limited knowledge to help in the identifi cation of words, they enter the partial-alphabetic phase. Once complete and functional knowledge of letters and sounds develops, readers enter the full -alphabetic phase. This marks a critical transition in the development of word recognition because readers a re finally able to fully decode words through spellings

PAGE 21

21 during the full-alphabetic phase. With practice, and as reading becomes more accurate and automatic readers will eventually pass from full-alphabetic reading into the consolidated-alphabetic phase. During this phase, they learn to recognize and use multi -letter units within words and to read words by analogy. Finally, readers enter the automatic phase where reading is highly accurate and automatic, and they have many strategies they can use to identify occasional u nknown words. The transition from partial-alphabetic to full-alphabetic word recognition has significant implications for the proposed study. This transition marks the point in development when beginning readers learn to use alphabetic information to deco de unknown printed words. In order for readers to become sufficient at word recognition, they must make this critical transition. Because proficient decoding requires strong orthographic and phonemic skills efforts to strengthen the associations between the orthographic and phonological processors are important during this transition. Including pseudowords during decoding instruction is one way to ensure that these associations are being activated. Once a word becomes familiar beginning readers can become reliant on the other sources of information (i.e., semantic and contextual sources) to identify the word. With a pseudoword, the correspondence between spelling and sound offers the only source for reading the word (Adams, 2011). Thus, reading pseudo words requires activation of the orthographic and phonological processors. An argument against using pseudowords for instruction exists among some researchers and educators based upon a conventional assumption that instructional stimuli for decoding must be real in order to be of value to students. For example,

PAGE 22

22 Adams (2011) makes the case against using only pseudowords for instruction, indicating that real words in meaningful contexts are better for beginning readers. While it is true that skill instruction should be paired with opportun ities for application in meaningful context (NRP, 2000), the potential contribution of pseudowords should not be dismissed. A goal of skill instruction is to promote generative responses from students. Generative responses demonstrate transfer of behavior or learning from trained stimuli to stimuli that have not been directly trained or practiced (Schumaker & Sherman, 1970). In other words, familiar stimuli are recombined in novel ways. The process by which generat ive responses are formed is referred to as recombinative generalization (Suchowierska, 2006). In linguistics, recombinative generalization plays a role in the development of language as previously learned language units are formed into novel arrangements (Goldstein, 1983). Instruction that promotes recombinative generalization should contribute to functional, rather than rote, knowledge that can generalize across stimuli, responses, and time (Suchowierska, 2006). For decoding instruction, this implies that, as graphophonemic units are learned, they will be useful to students when encountered in novel arrangements, or unfamiliar words. Because pseudowords conform to the orthographic and phonemic patterns of English, decoding instruction and practice with the graphophonemic units in pseudowords should result in generative responses. Conclusion This chapter provided an overview of the problem addressed by this study. Despite a wealth of research literature on the topic of beginning reading instruction, man y children continue to struggle. Effective decoding instruction is essential for the development of proficient word reading. A more complete understanding of the salient

PAGE 23

23 features of effective decoding instruction will help improve instructional practices and ultimately eliminate many of the struggles common to beginning readers. This study aims to elucidate the role of pseudowords in decoding instruction with beginning readers. Chapter 2 will present the literature on early reading development, beginnin g with a focus on Linnea Ehris phases of word recognition and the importance of decoding development. Next, the deficits leading to poor decoding will be summarized. Then, the findings of research on effective decoding instruction and intervention for a t-risk and struggling readers will be presented. Chapter 3 will describe the methods used to conduct this study. Chapter 4 will present results of the data analysis. Chapter 5 will discuss the findings and provide the limitations and implications of thi s study.

PAGE 24

24 CHAPTER 2 REVIEW OF RELATED LITERATURE To be proficient at reading, one must ultimately be able to comprehend the meaning of print, yet the ability to comprehend print is largely dependent upon the ability to first convert print into recognizable spoken words (Hoover & Gough, 1990). Accordingly, word recognition has been credited as the foundation of the reading process (Gough, 1984; Stanovich, 1991). Reading words may be approached in several ways (Ehri, 2004; Ehri & Snowling, 2004 ; NRP, 2000). Readers may decode words by converting letters into speech sounds ( i.e., phonemes ) and then blending those sounds to form words. They may also read words through analogy, which involves accessing from memory the spellings and phonemes of wo rds already learned to help read unfamiliar words that contain similar spellings. Context clues and background knowledge may also be used to anticipate or guess unfamiliar words. These three approaches to word reading are useful when reading words that h ave not been read before; however, as spellings and sounds within words that have been read before become familiar, words are stored in memory (Ehri, 2004; Ehri & Snowling, 2004). These words become part of the readers lexicon and can be read more rapidl y and efficiently by sight. The ability to read words by sight offers the greatest automaticity during reading and frees cognitive resources for text comprehension (Ehri & Snowling, 2004). Skilled readers recognize most words by sight because they have already formed complete connections between letters and sounds and have many words stored in memory. They may use decoding, analogy, and/or prediction, but they do so infrequently when encountering unfamiliar words in text. However, beginning readers mus t rely on these

PAGE 25

25 three approaches when reading because they have not yet stored the written form of many words in their lexicon. Although all three of these approaches have value, decoding provides the most accurate and reliable approach to reading unfamil iar words for the beginning reader because it provides the most reliable cues letters and their corresponding sounds (Ehri, 2004). With practice, decoding leads to greater accuracy in word reading, which contributes substantially to the storage of more words in memory, thereby promoting the development of sight word recognition. Ehris Phases of Word Recognition Development Over the years, several researchers have examined the process of developing word recognition, but the work of Linnea Ehri has made some of the most substantial contributions to our current understanding of the development of word rec ognition. Ehri (1995; 1999) has proposed a set of discernable phases of word recognition that develop over time as beginning readers experience print. Ehri and her colleagues (i.e., Ehri & Snowling, 2004; Ehri & McCormick, 1998) have described these phases with each successive phase varying by the degree to which a reader uses the alphabetic system. As developing readers pass through the phases, their command of the alphabetic system becomes stronger and word reading becomes more skilled. While the characteristics that define each phase are unique, a developing reader transitioning from one phase to another may demonstrate characteristics from both phases simultaneously. These phases of word recognition are the pre-alphabetic phase, the partial-alphabetic phase, the full-alphabetic phase, and the consolidated-alphabetic phase. On occasion, Ehri has also discussed a fifth phase, which she refers to as the automatic phase (Ehri & McCormick, 1998).

PAGE 26

26 Pre-alphabetic Phase According to Ehri (1995,1999), the pre-alphabetic phase of word recognition is characterized by a minimal working knowledge of the alphabetic system. Beginners in this phase have limited knowledge of letters and lack knowledge of the connections bet ween letters and phonemes. As such, alphabetic knowledge cannot be used to read words. Instead, these beginners attend to and memorize extraneous visual features unrelated to the printed word and associate these features with the spoken word (Ehri & Wilce, 1985; Gough, 1993). They commonly recognize visual stimuli associated with words in their environment such as familiar logos and may appear to be reading words not by recognizing or attending to any of the letters but rather by recognizing the colors, images, or patterns that remain consistent in those logos. Salient visual features within some words may also serve as cues, and some printed words may be memorized with practice but are quickly forgotten due to weak connections in memory. Consequently, connected text cannot be read during the pre -alphabetic phase (Ehri & Snowling, 2004). Because beginners in the pre-alphabetic phase rely on salient visual cues to remember words rather than their knowledge of connections between letters and sounds, it is particularly important during initial reading instruction to reduce o r eliminate extraneous cues that may detract from the alphabetic features of letters. To illustrate this, a child in this phase who learns the alphabet using a set of colorful manipulative letters may attend to the color of a given letter rather than the letter shape. This cue will not be sufficient for the child as he or she encounters the same letter in different contexts. In order to move beyond the pre-alphabetic phase of word reading, beginners

PAGE 27

27 must shift their attention from non-alphabetic visual cues to the letters in words and their corresponding sounds (Ehri & Wilce, 1985). Partial-alphabetic Phase The second phase of word recognition proposed by Ehri (1995, 1999) is characterized by the shift from reliance on visual cues to the use of phonetic c ues to recognize words (Ehri & Wilce, 1985). During the partial -alphabetic phase, developing readers begin to recognize some letters in words and associate those letters with their corresponding letter names and phonemes. This knowledge allows these beginners to use partial-alphabetic cues often initial and final letters in words in combination with other contextual clues to predict unfamiliar words (Ehri, & Snowling, 2004). In doing this, they begin to move beyond the unreliable associations they ha ve formed between non-alphabetic visual cues and meaning to more reliable graphophonemic associations. Development of graphophonemic associations is often facilitated for beginning readers when letter names contain relevant speech sounds because many chil dren learn letter names prior to learning phonemes associated with letters ( Ehri & Wilce, 1985, 1987; Templeton & Bear, 1993; Treiman, 1993). For example, the name for the letter k (i.e., kay) contains the sound /k/ in its pronunciation. Thus, recalling the correct phoneme can be facilitated when encountering the letter k in a word as opposed to encountering a letter such as w which does not contain the sound /w/ in its name (i.e., double u ). As graphophonemic associations are made, partial -alphabetic readers may begin to use the alphabetic system to support their reading efforts. Movement into the partial-alphabetic phase will begin with the recognition of a few letters and phonemes. In fact, many beginners moving into this phase learn to recognize the letters and sounds in their own names first (Ehri & Snowling, 2004).

PAGE 28

28 However, having partial knowledge of correspondence between letters and phonemes will limit accuracy when reading, particularly as the reader encounters new words, words with unfamiliar letters or sounds, or words with similar spellings. To illustrate, the word crane may easily be confused with similarly spelled words such as cane or can when the reader is only able to attend to some of the letters in the spelling. Thus, partial-alphabetic readers are not yet able to decode words. Movement from the partial alphabetic phase into the full-alphabetic phase requires extensive knowledge of how the letters are represented by phonemes in the conventional spelling system (Ehri & Snowling, 2004). Full-alphabetic Phase Readers in the full-alphabetic phase have formed complete connections between letters and sounds as well as the blending skills necessary to analyze and decode complete words (Ehri, 1999; Ehri & Snowling, 2004 ). Because full-alphabetic readers are able to use complete spellings to read, similarly spelled words may now be differentiated (Ehri & Wilce, 1987a). During this phase, readers also possess sufficient knowledge to decode spellings that they have never seen before (Ehri & Wilce, 1983, 1987b) and to read new words by analogy to words they already know (Ehri & Snowling, 2004). This promotes greater accuracy and increases reading opportunities, allowing full -alphabetic readers to be exposed to many new words. With practice the ability to decode aids the reader in the recognition of words as spellings become bonded with pronunciations in memory (Ehri & Snowling, 2004). Therefore, during the full -alphabetic phase, repeated exposure to words allows words to be committed to memory. As more words are stored in memory, word spellings, pronunciations, and mean ings may be accessed automatically upon sight (Ehri & Snowling, 2004 ) resulting in faster reading

PAGE 29

29 that no longer requires the decoding of e very word (Ehri & Wilce, 1983). For this reason, word recognition will become more accurate and automatic during the full alphabetic phase with sufficient practice. Consolidated-alphabetic Phase As a reader rapidly encounters and stores more new words in memory, they begin their transition into the consolidated -alphabetic phase. According to Ehri & Snowling (2004), working knowledge of recurring letter sequences is a hallmark characteristic of this phase. This may include knowledge of syllable pattern s, affixes, roots, and onset-rime combinations (Ehri & McCormick, 1998). As a reader is exposed to the same spelling patterns repeatedly and across many different words, the patterns become consolidated units in memory that are bonded with pronunciations. This bonding allows readers to recognize units within words that are larger than single letters, thereby speeding access of words from memory ( Ehri & Robbins, 1992; Juel, 1983). Consolidated units also assist in reading multisyllabic words. Readers in the consolidated-alphabetic phase read with greater accuracy and fluency not only because they have learned how to use the alphabetic system, but also because they now read words by analogy, they have had sufficient exposure to and practice reading many words, and they continue to rapidly develop their lexicon during this phase. As a result, basic decoding skill is no longer necessary to read most words but may be used and further developed as readers encounter new and unfamiliar words. Automatic Phase The final phase of word recognition, as described by Ehri and McCormick ( 1998), is the automatic phase. This phase is characterized by rapid and automatic identification of words that are both familiar and unfamiliar to the reader (Chall, 1983;

PAGE 30

30 Ehri & McCormick, 1998). Readers in this phase are able to read words effortlessly in or out of context, and when an unfamiliar word is encountered, they have multiple strategies involving the use of semantic and syntactic clues to help them identify and verify the word. This level of automatic and fluent reading allows the reader to direct their full attention to the meaning of the word for comprehension of the text. The Critical Transition to Skilled Decoding Ehris phases of word recognition provide an understan ding of the development from non-reader to proficient reader, but it is the point of transition between partial and full -alphabetic word recognition the point when decoding skill develops that is known to be particularly problematic for many strugglin g readers (Chall, 1983; Henry, 1993; Torgesen, 1999). To make this transition and to become a skillful decoder, readers must develop skill in phonemic awareness as well as an understanding of the alphabetic principle, both of which are known early predict ors of later reading development (Ehri et al., 2001). Phonemic awareness is the awareness of the speech sounds in spoken language. It includes the ability to identify, isolate, and otherwise manipulate individual phonemes in spoken language ( Ehri et al., 2001 ; NRP, 2000). These skills become instrumental during early reading of an alphabetic language because decoding requires the ability to identify the letters in a printed word, isolate the individual speech sounds associated with each of those letter s, and then blend them together to form the word. To do this, a reader must have an understanding of the alphabetic principle, which is knowledge of letters in the alphabet as well as the understanding that individual letters and combinations of letters, otherwise referred to as graphemes, are used in written language to represent phonemes in spoken language (Adams, 1990).

PAGE 31

31 As previously discussed, a developing reader in the partial -alphabetic phase will know some letters and the sounds commonly associated with those letters, thus, using partial-alphabetic cues along with other clues to identify words encountered in print. Because their knowledge of the alphabetic principle is not yet broad enough to decode words or to recognize words by larger units of spe lling, partial-alphabetic readers are limited to anticipating and guessing words using incomplete information from the spelling. The partial-alphabetic phase is important because it is the point in development when phonemic awareness, letter knowledge, a nd preliminary connections between letters and sounds begin to emerge. During this phase, it is important for readers to first develop the phonemic understanding that spoken words are made up of discrete sounds that can be blended and segmented. Isolating discrete sounds from within spoken language is a difficult task (Liberman, Shankweiler, Fischer, & Carter, 1974). This may be because phonemic awareness is largely abstract, and phonemes are not typically isolated for any reason during the flow of norma l continuous speech (Moats, 1998, Ehri et al., 2001). To make the abstract more concrete and to facilitate the transition from partialto fu ll-alphabetic reading, auditory and verbal information can be represented by visual and tactile supports such as manipulative sound markers or letters. Once initial letter-sound connections have been established for a few consonants and at least one vowel, letters may be combined to form simple words for decoding practice (Beck, 2006). However, as Ehri (1995, 1999, 2005) has explained, partial knowledge of the letters and their corresponding sounds will be insufficient to rely upon

PAGE 32

32 during reading for several reasons. First, knowing only some of the letters and sounds contained within a word leaves a reader with no choice but to guess the word, which decreases accuracy during reading. Second, many words share similar letters and sounds, so confusion between similarly spelled words is inevitable. If the reader has seen the word house and knows it contains the letters h and s, other words that are visually similar (e.g., horse or hose) or words that are not-so-similar but contain the letters h and s (e.g., has or his) are likely to be confused for house. Third, there is no way for readers with only partial-alphabetic knowledge to read all of the unfamiliar words they will encounter in print with accuracy and automaticity. To avoid these problems, developing readers must eventually learn how most conventio nal spelling patterns correspond to the sounds in spoken words. The transition from partialto full -alphabetic reading begins as readers start to fill in the gaps that exist in their knowledge of the alphabetic system. Full -alphabetic reading is not completely developed until all of the letters of the alphabet and the most common letter-sound relationships are learned. According to Ehri and Snowling (2004), word reading is generally slow and lacks fluency during the early part of the full alphabetic phase. This is attributed to the attention the emerging reader may be affording to every letter encountered in print as they deliberately sound out every word. However, this characteristic of full-alphabetic reading should be temporary as the alphabetic system is applied repeatedly and words are stored in memory (Ehri & Snowling, 2004). As the transition culminates, full -alphabetic readers will be able to move beyond basic decoding and begin to recognize spelling patterns across words, which may then be applied when reading new and unfamiliar words.

PAGE 33

33 Learning to decode marks a critical point in the development from partial to full alphabetic reading because it allows increased accuracy and automaticity with word reading, which in turn allows for greater opp ortunity to read more words. With sufficient practice, the sight word vocabulary of the full-alphabetic reader should grow rapidly and more attention may be shifted to comprehension of text (Ehri & Snowling, 2004). Reading connected text may continue to be slow and lack fluency at times depending upon how many unfamiliar words the reader encounters in text; thus, further development of a readers lexicon will depend upon sufficient practice with text at appropriate reading levels (Ehri & Snowling, 2004). Problems Learning to Decode Unfortunately, many children experience difficulty learning to read, and word recognition is the most common source of difficulty among struggling readers (Chall, 1983; Torgesen, 1999). Because decoding and word recognition ar e critical to reading development (Adams, 1990; Ehri, 1995, 1999) deficits that interfere with the development of decoding have been of great interest to researchers. It is a widely held view that most children with reading disabilities have a core phonological deficit that interferes with their ability to develop phonological awareness, discover the alphabetic principle, and learn to decode (McCardle, Scarborough, & Catts, 2001, p.233). According to Blachman (1994), as many as 90% of children and adults with serious reading difficulties demonstrate problems with phonological processing. Dyslexia, a well-researched specific learning disability that primarily affects the ability to read, has been most strongly associated with impairments in phonological processing, and particularly at the phoneme level (Ehri & Snowling, 2004).

PAGE 34

34 Other deficit models of reading implicate auditory, visual, or linguistic processing; processing speed; or memory as core deficits for reading disabilities; however, there is less evidence supporting causation for these models (Ehri & Snowling, 2004; McCardle et al., 2001). Moreover, while there are many models or theories that seek to explain reading difficulties, most account for a weakness in the ability to process the phonological features of language (McCardle et al., 2001). Instruction that Promotes Decoding Because decoding is the most common area of difficulty for struggling readers, instruction that facilitates the skills needed for decoding is beneficial to all beginning readers and critical for struggling readers (Adams, 2001). Although decoding may emerge for some students without much instruction, gaining a complete understanding of the alphabetic system offers the most promising results and is best achieved through explicit, systematic instruction (Adams, 1990; Ehri & Snowling, 2004; Snow, Burns, & Griffin, 1998), and this type of instruction is particularly important for strugglin g readers (Chall, 1967, 1983; Tunmer & Hoover, 1993). The National Reading Panel completed an extensive examination and meta -analyses of controlled experimental research to evaluate the effectiveness of phonemic awareness and phonics instruction. They published their findings in The Report of the National Reading Panel: Teaching Children to Read (2000). In the sections that follow, phonemic awareness and phonics instruction will each be discussed, and the findings of the NRP in each area of instruction will be summarized. Phonemic Awareness Instruction The goal of phonemic awareness instruction is to foster an awareness of and the ability to manipulate the sounds in spoken language (Ehri, 2004). Specific skills

PAGE 35

35 associated with phonemic awareness inclu de detecting, identifying, isolating, deleting, blending, and segmenting discrete phonemes in spoken words. Awareness of spoken language at the phoneme level is instrumental in preparing beginning readers for the blending and segmenting skills they will u se when decoding and encoding words in written language. However, discovery of the discrete phonemic units of language is not always obvious to developing readers and should be aided by explicit instruction that teaches how the system of speech sounds wor k (Ehri et al., 2001). The value of phonemic awareness skill as a predictor of reading development has been demonstrated repeatedly in the research ( Blachman, 1984, 1994; Hulme et al., 2002 ; Lenchner, Gerber & Routh, 1990; Mann, 1993 ; McCardle, Scarborough, & Catts, 2001; Nation & Hulme, 1997; Share, Jorm, Maclean, & Matthews, 1984). Consequently, phonemic awareness instruction has also gained appreciable attention from reading educators and researchers alike. Prior to the work of the NRP, nume rous experimental studies were conducted to evaluate the effectiveness of phonemic awareness instruction, but not all adhered to acceptable levels of methodological rigor (Ehri et al., 2001; Troia, 1999). For this reason, the NRP set out to examine the methodological design of controlled experiments that had been conducted as well as to determine the overall effectiveness of phonemic awareness instruction. The overall findings of the NRP (2000) meta -analysis support the consensus of many researchers that phonemic awareness instruction is beneficial for helping children develop as readers. The NRP reported that phonemic awareness instruction was found not only to improve phonemic awareness skills, but it was also found to facilitate transfer to various reading and spelling skills in preschool and elementary students

PAGE 36

36 including normally progressing readers as well as younger at -risk and older struggling readers. Many studies included in the analysis had already independently concluded that phonemic awarenes s instruction consistently produced reading outcome advantages over comparison approaches or no instruction at all (Ball & Blachman, 1991; Blachman et al., 1994; Bradley & Bryant 1983, 1985; Byrne & Fielding -Barnsley 1991, 1993, 1995; OConnor & Jenkins, 1 995; Urhy & Shepherd, 1993). The NRP meta-analysis found significant, moderate effect sizes for word reading (d = 0.61) and pseudoword reading (d = 0.56) (i.e., measures of word recognition and decoding skill ), strengthening the claims of these and other studies. Furthermore, even though the NRP reported larger effect sizes for studies with stronger designs, significant effects were found regardless of how rigorous studies were designed. This finding is promising because it suggests that weaknesses of some studies do not undermine claims of effectiveness of phonemic awareness instruction (Ehri et al., 2001). Although there is strong evidence that phonemic awareness instruction is beneficial during the early development of reading, the application of phon emic awareness skills must remain the ultimate goal of instruction. Ehri et al. (2001) warns that phonemic awareness should not be taught for its own sake but rather for its value in helping children understand and use the alphabetic system to read and w rite (p. 279). Larger effects of phonemic awareness instruction on reading outcomes have been found when children are taught not only phonemic awareness, but also the correspondence of speech sounds to letter s as well as how to apply phonemic awareness skills to reading and writing (Cunningham, 1990). As such, the NRP (2000) recommends that letters be incorporated into phonemic awareness instruction to

PAGE 37

37 provide visual stimuli and a concrete medium for the otherwise abstract tasks of phoneme manipulation a nd to help beginners establish initial connections between speech sounds and letters. As the connections between phonemes and letters are formed, phonemic awareness skills may be applied for decoding and the focus of instruction may shift toward phonics ( Ehri et al., 2001; NRP). Phonics Instruction Phonics refers to instruction that teaches the associations between letters and sounds as well as how to use the alphabet ic system to read and spell (Ehri, 2004). Specifically, the goal of phonics instruction is to teach beginners how to decode print. Approaches to phonics instruction are characterized throughout the literature in a variety of ways. For example, Stein, Johnson, and Gutlohn (1999) discuss two contrasting approaches to phonics instruction know n as implicit phonics and explicit phonics. In an implicit approach to phonics, students learn letter -sound connections indirectly and always through the context of whole wo rds rather than in isolation, whereas, in an explicit approach, the teacher direct s student attention to the connections between letters and sounds. Phonics instruction is generally considered explicit when individual letters and corresponding sounds are taught in isolation, usually through direct instructional approaches, and then app lied to and practiced in whole words and text (Mesmer & Griffith, 2005). Explicit phonics approaches are considered to be superior to indirect, implicit approaches, particularly for students who are at -risk or struggling (Adams, 1990; Chall, 1967, 1983). A systematic approach to phonics instruction is also described often throughout the literature (NRP, 2000). Phonics instruction is systematic when all of the major letter sound associations are presented in a clearly defined sequence (Ehri, 2004; Ehri &

PAGE 38

38 Snowling, 2004). In a well-designed systematic phonics approach, lessons are taught in a sequence that reveals the logic of the alphabetic system and allows students to apply what has already been learned to new learning (Adams, 2001, p.66). Although there has been disagreement in the past about the best method for teaching beginning reading, phonics instruction that is both explicit and systematic is known to be highly effective in promoting the development of decoding in beginning readers as well as in struggling readers (NRP). Several variations to systematic phonics instruction are found throughout the literature included in the NRP (2000) meta -analysis. The report identifies the most common variations as synthetic phonics, analytic phonics, embed ded phonics, analogy phonics, and phonics through spelling. Common to all of these approaches is the planned, sequential, and explicit presentation of phonic elements; however, these approaches also vary in one or more ways. Synthetic approaches to phonics instruction teach learners to focus on individual letters and sounds and to blend sounds to form words. In contrast, analytic phonics teaches learners to analyze letter -sound relationships, but only after words are identified. Individual letter -sound relationships are not taught in isolation with this approach. Embedded phonics teaches letter-sound correspondences in the context of reading. During this approach, learners are encouraged to use knowledge of letters and sounds as well as context clues t o identify unfamiliar words. The analogy approach to phonics emphasizes the use of familiar word parts to identify unknown words. Phonics through spelling teaches learners to segment spoken words into letters and write words. Most phonics programs inclu de one or any combination of these systematic approaches.

PAGE 39

39 The results of the NRP (2000) meta-analysis indicate that phonics instruction has a significant, moderate effect on reading and benefits decoding and word reading as well as text comprehension and s pelling in many readers including young at -risk students as well as older students with reading disabilities. In addition, synthetic and larger-unit approaches to phonics (i.e., analogy using onset -rime or phonograms) produce similar advantages, as do ind ividual, small-group, and whole-class delivery methods. More importantly, systematic phonics produces greater reading outcomes than non -systematic approaches to phonics and all other forms of instruction. Although it is clear that struggling readers need effective decoding instruction, and the NRP (2000) helps to establish that phonemic awareness and phonics instruction are both effective and important for the development of decoding skill, questions remain about what specific aspects or features of instru ction contribute to its effectiveness. In the next section, several features of instruction for improving decoding and word recognition in at -risk and struggling students will be discussed. Because the NRP has already summarized findings of a large body of experimental research conducted prior to 2000, the following review will focus primarily on research that has been conducted since that time. Finally, implications for future research and practice will be discussed. Intervention to Develop Decoding Skills in Struggling Readers The studies on phonemic awareness and phonics interventions that have been conducted in the 12 years since the report of the NRP have several things in common. Most of the studies have concentrated on instruction for at -risk or struggling firstand second-grade students. Most have implemented supplementary interventions with small-groups of students or one-on-one. Many have studied interventions that focus

PAGE 40

40 solely on phonics while others have investigated more comprehensive inte rventions that include phonics as one component. Almost all have incorporated the use of manipulatives to teach phonemic awareness and/or phonics. Although there are commonalities throughout this body of research, there is variability as well, making it difficult to draw conclusions about specific features that may or may not contribute to the effectiveness of instruction. The following sections will focus on several common features of decoding instruction found within the literature. Phonics with Phoneme Level I nstruction One study investigated the effects of phonological awareness intervention when added to an early reading program. Hatcher, Hulme, and Snowling (2004) were interested in the inclusion of supplemental phonological awareness instruction for young beginning readers participating in a structured reading program with a strong phonics component. Classes of 4to 5 -year -old children were divided into matched groups and randomly assigned to one of 4 conditions. Three conditions received 10 minutes of supplemental phonological awareness instruction varying by phonological unit (i.e., one condition received rhyme instruction, one received phoneme instruction, and one received both rhyme and phoneme instruction). Students in the control conditio n received the same amount of supplemental time participating in more of the regular reading program. Supplemental instruction was provided to groups of 10 to 15 students 3 times per week for 14.5 months over the first 2 years of school. Hatcher et al. (2004) found that phonological training had no significant impact upon the reading outcome of normally developing readers; a highly structured phonics program was sufficient for most students to master the alphabetic principle and learn to read. However, students at risk for reading problems who received phonological

PAGE 41

41 awareness instruction at the phoneme level made significant gains in word reading whereas, at-risk students who received only rhyme instruction or no phonological awareness instruction did not. Furthermore, the level of phoneme awareness skill attained by at -risk students at the end of the intervention period proved to be a powerful predictor of final word and non-word reading at follow-up. These findings indicate that young students struggling to develop reading skills may need something more than typically developing readers, and instruction that focuses on the phoneme unit may be a critical component for students who are at -risk. Mu ltisensory word work Several studies included in the NRP report provided interventions that included word work as a method of decoding instruction (e.g., Blachman, Ball, Black, & Tangel, 1994; Blachman, Tangel, Ball, Black, & McGraw, 1999; Iversen & Tunmer, 1993; OConnor & Jenkins, 1995; Tunmer & Hoover, 1993). Word work is an instructional approach commonly used to teach students how to decode by making connections between phonemes and graphemes explicit through the use of manipulative materials that provide tactile and visual phoneme representations. A mul tisensory word work approach is particularly beneficial for struggling students because it provides a concrete medium for teaching the abstract tasks of phoneme manipulation (Campbell, Helf, & Cooke, 2008). Throughout the literature, word work is conducted in a variety of ways that typically make use of magnetic letters, letter tiles, colored chips or blocks, and/or Elkonin boxes during tasks that involve the manipulation of phonemes within words. Most of the studies in this review incorporate one of two common multisensory approaches to word work during decoding instruction In the first approach, Elkonin boxes, also referred to as word boxes or graphic boxes, are used to assist in the

PAGE 42

42 development of phonemic awareness. Elkonin boxes are a series of co nnected squares drawn on paper to represent a word, with each individual square representing a single speech sound (i.e., phoneme) within the word ( Joseph, 2002b; Williams, 1984). Students begin by placing small discs or other place -markers, one at a time, into the squares as they segment the word into its phonemes. Later, discs are replaced with letters to represent each phoneme. Elkonin ( 1973) reported that students who were taught with this method of phonemic segmentation improved in other areas of li teracy beyond segmenting. One of the first researchers to use Elkonin boxes was Marie Clay (1993) in her work with Reading Recovery, a one-on-one tutoring model. In Reading Recovery, Elkonin boxes are incorporated as a method to help students spell by se gmenting words into their individual sounds. Several other one -on-one tutoring models have also included Elkonin boxes as part of a comprehensive intervention including Reading Rescue (Hoover & Lane, 1997), Sound Partners (Vadasy, Sanders, & Peyton, 2005), and University of Florida Literacy Initiative (UFLI; Hayes, Lane, & Pullen, 1999). The research on these models has demonstrated efficacy. Ehri, Dreyer, Flugman, and Gross (2007) conducted a study of Reading Rescue and found it to be effective for increasing word reading as well as comprehension skills. Sound Partners was investigated by Vadasy et al. ( 2005) and found to be effective for increasing decoding, spelling, fluency, and comprehension, especially when paired with passage reading in decodable text. While both of these tutoring models were found to be effective, neither study specifically examined the unique contributions of Elkonin boxes.

PAGE 43

43 In a study conducted by Lane, Pullen, Hudson, and Konold (2009), the effects of Elkonin boxes were spec ifically examined within the UFLI model. These researchers found word work with Elkonin boxes to be an integral component of UFLI. When all components of this model were provided to participants, UFLI tutoring was effective; yet when word work with Elkonin boxes was removed from the model, participants who received the reduced model failed to demonstrate significant improvements in decoding or word recognition. Other researchers have examined the use of Elkonin boxes as an isolated instructional approach. In an early study, Joseph ( 1998) introduced word boxes to six secondthrough fourth-grade students who were identified as having learning disabilities. Using a multiple-baseline design across participants, Joseph measured word identification and spelling for each participant across baseline, treatment, and maintenance conditions. Through visual inspection of the data, a functional relationship was observed between instruction using word boxes and student performance. For each participant, a notable increase in performance was observed for word identification and spelling. More recently, Joseph extended her research and found word boxes to be effective in combination with word sorts for elementary-aged children with mental retardation (Joseph, 2002a) and in combination with repeated readings for adolescent struggling readers (Devault & Joseph, 2004). Schmidgall and Joseph (2007) compared the use of word boxes to drill and practice and intersp ersing techniques and found word boxes to be more effective than the other methods of decoding instruction. In addition, a social validity measure indicated that students preferred word boxes over the other instructional methods.

PAGE 44

44 The second popular approach to conducting word work makes use of manipulative letters, or individual moveable letters such as magnetic letters, tiles or cards with letters printed on them, or letter blocks. In this approach, letters are used to form a word, and then one or more letters are removed and replaced to form a new word. The process is repeated to provide multiple opportunities for decoding, and e ach successive word typically shares one or more sounds in common with the previous word to demonstrate relationships between common sounds and spellings Several early studies have used word work with manipulative letters as an approach to phonics instruction, but without providing any empirical evidence to specifically support the use of this type of word work For example, Cunningham, Hall, and Defee (1991) developed Making Words, an approach to word work that employs letter cards to help students learn to decode and spell words. Students are given a limited number of letters and begin by making s hort, simple words, which they later build upon to make longer, more complex wor ds. Making Words activities have been used as part of a multi-component instructional model (Cunningham & Cunningham, 1992) that has become popular among classroom teachers; however, no empirical studies have been conducted to examine the effectiveness of this approach (Rasinski, 1999). Several studies identified by the NRP (2000) also included phonemic awareness instruction that incorporated the use of manipulative letters (e.g., Bradley & Bryant, 1985; Ehri & Wilce, 1987a; OConnor & Jenkins, 1995 ; Uhr y & Shepherd, 1993). Perhaps the earliest support for the use of manipulative letters during phonemic awareness instruction was offered by Bradley and Bryant (1985). From experience, these authors believed manipulative letters were an extremely effective approach for

PAGE 45

45 helping students struggling with reading to make the connection between sounds of speech and print during reading and spelling tasks. They rationalized that struggling readers often have difficulty with the phonological aspect of reading bec ause it is abstract; thus, by incorporating letters into sound categorization, students would be provided a concrete medium with which to work. As such, Bradley and Bryant (1985) compared the reading and spelling performance of children across four experimental conditions: (a) children trained to categorize words based upon sounds, (b) children trained to categorize words based upon sounds using plastic letters, (c) children trained to categorize words based upon concepts, and (d) children who received no training. After two years of training, these researchers found that children trained in both sound categorization conditions read and spelled better than children trained with conceptual categorization, and children trained to categorize sounds using plas tic letters surpassed all groups on reading and spelling measures. Ehri and Wilce (1987) also conducted an early experiment using letter tiles to train kindergarten students to either match letters with isolated sounds or spell nonsense words. They found that students who learned to spell words learned to read words more effectively at posttest than students who learned to match letters to sounds. Uhry and Shepherd (1993) also examined the effects of segmenting and spelling instruction that used manipulative letter blocks with first-grade students. Students in the control group received the additional instructional time without segmenting or spelling using manipulative letters. These researchers found that segmenting and spelling instruction using the manipulatives resulted in significant gains and better performance on measures of timed word reading, nonsense word reading, and timed oral passage

PAGE 46

46 word reading. Finally, OConnor and Jenkins (1995) examined the effects of providing supplemental segmenting and spelling instruction with manipulative letters to kindergarten students with developmental disabilities who were receiving Reading Mastery instruction. They found no difference between treatment and control on phonological blending and segmenting ta sks; however, they did find that treatment students made significant improvements over control students on measures of spelling and reading familiar and new words. Although most studies conducted since the report of the NRP have included some form of word work with manipulatives, three studies were identified that specifically examine word work with manipulative letters. In the first study, Pullen (2000) specifically examined the effect of instruction using letters to manipulate, blend, and segment lette rs and sounds with first-grade students struggling to learn to read. Three groups were compared: (a) an intervention including repeated reading and manipulative letter practice, (b) an intervention using only repeated reading, and (c) a no-treatment control. Results indicated that both intervention groups outperformed the control group on measures of phonological awareness and sight word knowledge, and the intervention group that used manipulative letters outperformed both the comparison and the control group on measures of decoding. No difference was found across groups on reading fluency and comprehension. In an attempt to replicate the findings of the first study, Pullen Lane, Lloyd, Nowak, and Ryals (2005) conducted a multiple-baseline single-subject experiment that examined the functional relationship between the introduction of word work instruction using manipulative letters and the decoding ability of first -grade students struggling to

PAGE 47

47 read. These researchers found that decoding skill increased as students learned to manipulate, blend, and segment the sounds of words using manipulative letters In this review, the only study to isolate the effects of word work with manipulative letters in a multi-component intervention was conducted by Lane, Pul len, Hudson, and Konold (2009). These researchers examined the components of a one -on-one tutoring model (i.e., the University of Florida Literacy Initiative), which was designed to help novice teachers to tutor struggling readers in first grade. They we re particularly interested in determining if the removal of word work, sentence writing with word work, or genre and text structure instruction from the UFLI model would affect reading outcomes. One hundred participants were randomly assigned to one of fo ur experimental conditions or a control group. Lane et al. found that students who received all five of the UFLI steps performed better on decoding and word recognition tasks than control students who did not receive UFLI tutoring. However, when either word work using manipulative letters or written word work during sentence writing was removed from the model, there were no significant differences between the treatment and control conditions. The findings provide evidence that word work is an integral component of this multi-component tutoring model when implemented in a one -on-one setting. From this research, there is an abundance of evidence to suggest that specific methods of instruction can be highly effective with students who are struggling to lea rn to read. Interventions that explicitly addressed phoneme manipulation were most beneficial for struggling students, and this was particularly true when multisensory support was incorporated into instruction. Word work using either Elkonin boxes or

PAGE 48

48 man ipulative letters two common approaches to multisensory instruction was found to be beneficial for struggling students learning to decode. Using Pseudowords A few researchers have incorporated pseudowords as part of a decoding intervention, and participants who have been exposed to these interventions have improved on measures of reading. However, research examining the role of pseudowords in decoding skill development is extremely limited. Ehri and Wilce (1987b) conducted an experiment to identify d ifferences between readers who were at two different developmental points in word reading: phonetic -cue reading and cipher reading. The participants were kindergarteners who knew most letters, some letter-sound pairs, and could read at least three primer -level words, but few of the words to be taught during training. Participants were pre -tested and then half of them were trained to mastery on decoding basic nonsense words (i.e., cipher training), while the other half were trained to mastery on certain le tter -sound relationships (i.e., cue training). On post -test, the cipher readers were stronger at printed real-word learning, decoding nonsense words, and spelling. In a few of the previously described studies, interventions included pseudowords, but specific information about how they were used was not provided. For example, in the Pullen et al. (2005) study, pseudowords were included in word work instruction using manipulative letters but there was no information provided regarding the number of pseudowords or the number of opportunities for practice The intervention used in the Lane et al. (2009) study also included pseudowords during word work with manipulative letters, but the description of the intervention does not provide details

PAGE 49

49 about how pseudoword practice is incorporated. Both studies found positive outcomes in decoding. The only study to specifically examine pseudowords as a dependent variable was conducted by Cardenas (2009). Thirty kindergartners in two intact classes participated in the study. All students were assessed at four points in time using the Phonetic Word List Assessment (PWLA), which was created by the researcher Following T est 1, both classes received an introductory phase of phonics instruction that used real words only. Following Test 2, the second phase of instruction, which lasted 28 school days, began. In this phase, one class received a pseudoword phonics intervention, while the other class continued working with real words only. Following Test 3, the third and final phase began, in which both classes received instruction with real words only. All students were assessed again (Test 4) at the conclusion of the third phase of instruction. The findings indicate that students who were provided pseudoword phonics in place of real-word phonics made greater gains from Test 2 to Test 3 and from Test 3 to Test 4 than students who received only real-word phonics. Although this study did not include random assignment to treatment, the findings indicate that p honics instruction that included pseudowords was beneficial in the development of decoding skill. Instructional Time An important instructional variable that was addressed in the report of the NRP (2000) is the amount of instructional time needed to develop phonological and alphabetic proficiency in beginning or struggling readers. The NRP draws two general conclusions concerning the amount of phonemic awareness instruction needed to be effective. First, they indicate that studies providing anywhere from 5 to 18 total hours of

PAGE 50

50 instruction yielded the largest effects on phonemic awareness acquisition, and transfer to reading was greatest when less than 20 hours of instruction were provided. Second, they point out that average training sessions lasted 25 minu tes with few lasting more than 30 minutes. From this conclusion, they make the recommendation that sessions last no more than 30 minutes. For phonics instruction, the NRP provides even less direction with no conclusions or specific recommendations made r egarding instructional time. Unfortunately, these conclusions shed very little light on the amount of time needed to sufficiently develop decoding skill. The reasons for this may be that very few researchers have specifically examined instructional time as an independent variable. It may also be that the research demonstrating effective phonemic awareness and phonics instruction varies in the amount of instruction that has been provided as well as in confounding variables (i.e., amount of time needed may vary depending upon student age or ability, teacher or program quality, etc.). Most research that has examined phonemic awareness and phonics instruction in the last 12 years has provided supplemental instruction to at -risk or struggling readers for 30 minutes or less per session. However, across studies, there is a range of instructional time that has been effective for students who are at -risk or struggling. For example, the results from the study by Hatcher et al. (2004) suggest that adding as little as 10 minutes of supplemental phonemic awareness instruction 3 times per week for 14.5 months can result in significant reading improvements for struggling beginners who are already receiving phonics instruction. Berninger et al. (2003) also found 10 min utes of supplemental instruction in the alphabetic principle with 10 additional minutes of reflective work 2 times per week to be beneficial for second-grade struggling readers,

PAGE 51

51 and this was after only 24 sessions. In contrast, McCandliss, Beck, Sendak, a nd Perfetti (2003) found 50-minute phonics sessions to be effective with second -grade struggling readers when provided 3 times per week for a total of 20 sessions. Two studies that specifically manipulated instructional time as an independent variable were identified for this review. Simmons et al. (2007) examined instructional time along with specificity of program design to determine if supplemental code -based programs that vary in amount of instructional time and specificity result in different outc omes in kindergarteners identified at risk for reading difficulties. They compared three conditions: 30 minutes of highly-specified code-based instruction, 15 minutes of highly-specified code-based instruction and 15 minutes of non -code-based instruction, and 30 minutes of moderately-specified code-based instruction. These researchers found that 15 minutes of highly-specified instruction was sufficient for increasing phonemic awareness in at -risk kindergarteners, but 30 minutes of highly-specified code instruction was significantly more effective for increasing decoding and word reading skills in the students who entered kindergarten most at -risk. Furthermore, highlyspecified instruction was significantly more effective at increasing phonemic decoding for all at-risk kindergarteners, and word attack and word reading for most -at-risk kindergartners, than 30 minutes of moderately-specified instruction. Vadasy, Sanders, and Peyton (2005) also compared three conditions that varied in instructional time, but participants were first graders scoring in the lowest quartile for reading. Treatment groups received either (a) 30 minutes of Sound Partners (i.e., a phonics-based program teaching letter -sound correspondence and providing word reading opportunities), or (b) 15 to 20 minutes of Sound Partners and 10 to 15 minutes

PAGE 52

52 of practice reading decodable text. The control group received no supplementary intervention. The instruction was provided 4 times per week for 30 -minute sessions for a total of 8 months Fol lowing the intervention, both treatment groups significantly outperformed the control group on reading accuracy, comprehension, fluenc y, and spelling measures with greater effects for word reading accuracy than for word or passage reading fluency. The treatment groups performed comparably on most other measures, but the group that received reading practice with decodable text significantly outperformed the Sound Partners -only group on passage reading fluency and accuracy. The findings here support the rec ommendations of the NRP (2000) and Ehri et al. (2001) in that less time may be required to teach specific phonics skills when students are also given time to apply those skills in real text. Although more research is needed to determine the most effect ive amount of supplemental instruction for decoding development, the growing evidence seems to suggest sessions lasting 30-minutes or less will be sufficient for most students struggling with reading. This factor will ultimately depend upon additional varia bles such as student ability or need, teacher quality, and program design. What remains unknown from the existing research is the frequency of sessions and/or duration of program that are optimal for improving decoding in struggling readers. Group Size Research has supported the value of alternatives to whole -group instruction for teaching reading to students with disabilities (Elbaum, Vaughn, Hughes, & Moody, 1999). While it is true that interventions provided to individual students can significantly improve reading outcomes, small-group instruction may be just as beneficial for students struggling with reading when instruction is well designed and provided by a

PAGE 53

53 highly-qualified instructor (Elbaum, Vaughn, Hughes, & Moody, 2000). The NRP (2000) also acknowledges student -grouping practices as a variable worth examining for phonemic awareness and phonics instruction. They included three instructional group sizes for comparison in their analyses. For phonemic awareness instruction, they found small-group instruction resulted in larger reading effect sizes than instruction with individuals or whole classes. Group size did not make a difference for phonics instruction though, with all group sizes producing significant moderate effects that did not differ statistically. Most of the studies identified for this review implemented interventions with small groups or individual students with positive results. Hatcher et al. (2004) conducted the only study in which phonemic awareness instruction was provided to la rger groups of 10 to 15 students, and found improved phonemic awareness and reading skills in young students who were at -risk for reading delay. Only one study specifically included group size as an independent variable. Vaughn et al. (2003) compared the reading outcomes of students receiving supplemental reading intervention provided one -on-one or in groups of three or ten students. The intervention was comprehensive in that it included fluency practice, phonemic awareness instruction, leveled reading, and explicit instruction in the alphabetic principle (i.e., word study). These researchers found that most students made significant gains from preto post-test on all reading outcome measures including measures of decoding and word recognition regardles s of group size, and these gains were maintained four to five weeks later at follow -up. Based on effect sizes, individual instruction and instruction with groups of three were each highly effective, and neither was superior to the other.

PAGE 54

54 From the converging evidence on instructional group size, it is clear that struggling readers can benefit from explicit instruction regardless of group size. However, recent evidence suggests struggling students may benefit most when instruction is provided to smaller groups. Furthermore, it is of practical importance that small-group instruction may be just as effective as individual instruction. Instructor Variables The NRP (2000) report does not provide conclusions about teachers that are most effective at providing decoding instruction. They do, however, provide a few recommendations for future research in this area. First, the NRP recommends that more research is needed to determine what teachers need to know to provide effective phonics instruction. In addition, they suggest that research is needed to determine what motivates teachers to provide explicit and systematic phonics instruction, which is often scripted and lacks appeal to many teachers. Since the publication of the NRP report, a few researchers have demonstrated that supplemental phonemic awareness and phonics instruction may be provided effectively by paraprofessionals. Three studies were found that examined the effects of systematic and scripted instruction provided by trained par aprofessionals. Ryder, Tunmer, and Greaney (2008) studied supplementary phon emic awareness and phonemically-based decoding instruction for 6-7 year old children with early reading difficulties in a whole-language environment. The treatment consisted of 5 6 highlysequenced and scripted small-group lessons delivered 4 times per week for 20 to 30 minute sessions by a trained and monitored teachers aide. Matched controls continued to receive whole-language instruction and supplementary activities for strug gling readers as provided by their classroom teachers. Results indicated that treatment

PAGE 55

55 students significantly outperformed controls on phonemic awareness, decoding, a nd context free word reading, though no significant differences were found for reading words in context or for comprehension. At two-year follow-up, the treatment continued to significantly outperform control on context -free word recognition and also on reading words in context. Two studies conducted by Vadasy and colleagues also explored t he effectiveness of supplementary phonics -based interventions provided by paraprofessionals. In a study described earlier, Vadasy, Sanders, and Peyton (2005) compared two phonics-based intervention groups to a control that received no supplementary intervention. First graders struggling with reading who received supplementary, one-on-one, phonics-based intervention from trained paraprofessionals significantly outperformed controls on decoding measures as well as various reading and spelling measures. In a second study, Vadasy, Sanders, and Tudor (2007) examined the effects of one-on-one supplementary phonics intervention provided by trained paraprofessionals to second and third-grade struggling readers. The treatment group received 15 minutes of phonics instruction and 15 minutes of passage reading 4 times per week and the control received no supplementary intervention. Students who received the phonics intervention outperformed control students on post -test measures of reading accuracy and fluency, and performance levels were maintained at follow -up three months later. Taken together, the results of these three studies provide evidence that paraprofessionals can successfully provide phonics instruction to at -risk and struggling students. However, the generalizability of these findings must be considered. The

PAGE 56

56 programs used in these studies were scripted, each paraprofessional was trained to use a specific program, and they were either monitored by an on -sight literacy resource teacher or provided weekly or bi-weekly coaching and modeling throughout the year. In addition these paraprofessionals were monitored for fidelity by research staff. Therefore, it is unclear whether interventions can be successfully provided by paraprofessionals under less structured conditions. Summary and Conclusions Learning to decode marks a critical milestone in the development of reading. Without decoding skill, readers are not well equipped to tackle new and unfamiliar words, and this becomes increasingly problematic as text becomes more difficult (Ehri, 1995, 1999) Moreover, when word recognition is compromised, reading comprehension is likely to suffer. Instruction that promotes phonemic awareness and an understanding of the alphabetic system is paramount for many developing readers (Adams, 1990). Decades of research provide strong support for explicit and systematic phonemic awareness and phonics instruction (NRP, 2000). However, there is still much to learn about the aspects of instruction that facilitate learning. Research conducted after the NRP report on reading instruction continues to shed light on some specific features of effective decoding instruction. For students who are at-risk for reading difficulties, instructional methods that explicitly teach phonological awareness at the phoneme level are found to be beneficial. There is also evidence to support methods that utilize multisensory word work as part of a comprehensive reading intervention or as an isolated instructional approac h with students who are either learning to read or struggling to learn.

PAGE 57

57 Recent research also suggests that the amount of supplementary phonemic awareness and phonics instruction needed to improve decoding and word reading will vary; however, there is growing support that interventions lasting 30-minutes or less per session are sufficient for most students struggling with reading. In addition, it is promising for teachers that given the numerous demands and time constraints inherent in the school day, students benefit from instruction provided in a variety of gr ouping formats and by a variety of instructional providers Contrary to common belief struggling students may not necessarily require one-on-one instruction. Growing research continues to suggest that small-group instruction can be just as effective, and paraprofessionals are able to provide effective phonics instruction to at -risk and struggling students when given adequate training and support. Finally, there is preliminary evidence for the use of pseudowords during decoding instruction with beginning readers. The Cardenas (2009) study challenges a widespread belief that phonics instruction should involve real words only. Improving our understanding of various aspects of effective decoding instruction is important for improving instructional practices Of course, these research findings should be received with caution because the effectiveness of any practice will ultimately depend upon other contextual variables such as student ability or need, teacher quality, and program design. However, these studies also provide important clues about how to provide effective decoding instruction to emerging readers, and they provide direction for future research. Implications for Research Although research continues to provide direction for educators and researchers to improve decoding inst ruction, there is still more to learn. We may have more

PAGE 58

58 information now about how long intervention sessions should last, but uncertainty remains about the frequency of sessions and the duration of programs. There is a great deal of variability in the research making it difficult to draw any specific conclusions about these variables. It is unknown if more lessons help consolidate the skills being taught or if interventions should continue beyond the early elementary school years. More research is needed to investigate session frequency and program duration. Questions also remain regarding the characteristics of teachers who provide the most effective decoding instruction. Despite the recommendations of the NRP (2000), there is still a paucity of research that empirically examines the relationships between teachers knowledge and practice and student reading growth because research in this area is complicated by many confounding variables. It would be important to learn what teachers need to know to provide effective decoding instruction to students beginning and struggling to learn to read. In addition, research is still needed to determine what will motivate teachers to provide explicit and systematic phonics instruction, which is often scripted and lacks appeal. Finally, many studies have demonstrated the importance of systematic and explicit instruction in phonemic awareness and phonics for the development of phonemic awareness, decoding, and word reading; yet, few studies have isolated specific instructional features to explore what makes this instruction most effective. There may be more to learn about effective decoding instruction by isolating discrete features of instruction that is effective, such as the number of words practiced or the ty pe of words presented during instruction.

PAGE 59

59 This study aimed to add to the current knowledge base about decoding development by examining a discrete feature of explicit decoding instruction. The purpose was to delineate the role of pseudowords in decoding instruction at a specific phase in early reading development Identifying ways to improve the quality of decoding instruction is important for beginning readers, particularly because so many continue to struggle with the skills needed to become effective at decoding.

PAGE 60

60 CHAPTER 3 METHODS The purpose of this study was to investigate the impact of including pseudowords during decoding instruction on decoding outcomes of developing readers who are transitioning from the partial-alphabetic to the full -alphabetic phase of word reading. An experimental pre-test post-test control group design with random assignment was used to compare three groups: a treatment group that received decoding instruction using a combination of real words and pseudowords during word work activities; a comparison group that received decoding instruction using only real words during word work activities; and a control group that did not receive instruction from the researcher. This design was selected to allow for the compa rison of post-test scores while controlling for pre-test group differences. In the sections that follow, the research questions and hypotheses will be presented. Then, the methods used to conduct the study will be described. Hypotheses To examine the ef fects of incorporating pseudowords during decoding instruction on the reading development of kindergarteners in the partial -alphabetic phase of word recognition, the following four research questions were addressed: 1. What are the effects of decoding instruction that includes pseudowords on real word decoding accuracy? 2. What are the effects of decoding instruction that includes pseudowords on pseudoword decoding accuracy? 3. What are the effects of decoding instruction that includes pseudowords on real word decoding automaticity? 4. What are the effects of decoding instruction that includes pseudowords on pseudoword decoding automaticity?

PAGE 61

61 The following null hypotheses were tested to address these research questions: H1: There is no statistically significant difference between groups on measures of real word decoding accuracy. H2: There is no statistically significant difference between groups on measures of pseudoword decoding accuracy. H3: There is no statistically significant difference betw een groups on measures of realword decoding automaticity. H4: There is no statistically significant difference between groups on measures of pseudoword decoding automaticity. Setting One primary school located in north central Florida was selected for this study. This school was selected in part due to a large kindergarten enrollment, which offered a promising pool from which to draw participants. The centralization of a large participant pool in one location also made the study more feasible given limited resources. In addition, the principal and teachers at this school communicated a willingness to support implementation of the study. At the initiation of the study, the participating school had a total of 513 students enrolled in kindergarten through second grade. The racial and ethnic profile for the student population was 49% Caucasian, 29% African American/Black, 13% Hispanic, 7% Multiracial, 1% Asian, and less than 1% Native American. Of the total population, 19% were receiving Title I services, 13% were receiving exceptional student education (ESE) services, and 5% were identified with English as a second language. Approximately 65% of the students were receiving either free or reduced -price lunch. There were 156 students attending kindergarten at t he initiation of the study. To ensure that a sufficient number of kindergarteners were within the desired developmental phase of word recognition for the study, results from the mid -year

PAGE 62

62 administration of the Broad Screen/Progress Monitoring Tool (BS/PMT) of the Florida Assessments for Instruction in Reading (FAIR, Florida Department of Education, 2009) were examined. In kindergarten, the BS/PMT assesses letter -sound knowledge and phonemic awareness during the mid -year administration. The measure yields a Probability of Reading Success (PRS ) score, which indicates a students chance of scoring at or above the 40th percentile on end-of-year reading assessments. Kindergarteners with a score of 85% or above on the BS/PMT a re considered to have a high PRS. Examination of the mid-year BS/PMT scores indicated that at least 58 kindergarteners were not demonstrating a high PRS. Participants A total of 72 kindergarten students participated in the study. An a priori power analysis was conducted prior to initiation of the study to determine the sample size needed to achieve desired power. Based upon the recommendations of Cohen (1988), power should be no less than .80 to reduce the probability of making a Type II Error to .20 To complete the analysis, an effect size of .60 was estimated based upon studies of similar interventions with decoding outcomes that yield effect sizes ranging from .35 to 2.06 (Scammacca, Vaughn, Roberts, Wanzek, & Torgesen, 2007). The effect size was then converted into an effect size f for the analysis. Using the computer program, G*Power 3 (Faul, Erdfelder, Lang, & Buchner, 2007), with power at .80, an effect size f of .34, and alpha level at .05 a total sample size of 87 was needed to achieve the desired power of .80. Students in the partial-alphabetic phase of word recognition were sought for participation in this study. As such, kindergarten was selected for the study because many students in kindergarten have limited decoding skill (i.e., they are not yet full

PAGE 63

63 alphabetic readers) in the spring, but have prerequisite alphabet and letter -sound knowledge to begin learning to decode. A screening process was used to identify all kindergarteners meeting these criteria. The students were identified based upon the results of an invented spelling screening, which is d escribed in full detail in the next section of this chapter. Invented spellings provide information about a students level of development in the areas of phonemic awarenes s and letter-sound knowledge; thus invented spelling scores were used to identify students performing within the partialalphabetic phase. In addition, all spellings were inspected by the researcher to ensure all potential participants were indeed demonstrating spellings that were characteristic of the partial-alphabetic phase. Therefore, if a students invented spelling score was marginal, but his or her spellings were largely characteristic of the partial -alphabetic phase, then the student was invited to participate in the study. A total of 96 kindergarteners met these criteria and were recruited to participate. As required by the University of Florida Institutional Review Board, approval from the IRB and the school district were obtained to complete the study. Documentation of approval from the IRB is provided in Appendix A. Informed consent letters were provided to parents or guardians of all recruited kindergarteners A total of 73 students returned informed consent to participate in the study. Each student who returned inf ormed consent also provided verbal assent to participate prior to the initial assessment. Of the 73 original participants, one student moved from the school prior to the completing all of the lessons for the study, and data for that student is not included in the descriptive or statistical data reported in this paper

PAGE 64

64 All 73 original participants were assigned by chance to one of three conditions using a set of random numbers (Gall, Gall, & Borg, 2007). This set of numbers was obtained using the number sequence generator located on random.org, a free Internet service. Participant names were listed in alphabetical order by last name and numbered. The set of random numbers was then used to assign a near equal number of students to each condition. Demographic data were collected for the 72 participant s who completed the study. These data, which are reported by group in Table 3-1, reveal that all three groups were similar in demographic composition, with the exception of the treatment group, which contained notably more females and more students identified as African American/Black than the other two groups. Table 3-1. Demographic information for sample by g roup Treatment Group Comparison Group Control Group Total Sample Total Participants 25 23 24 72 Mean Age (years-months) 6-2 6-2 6-4 6-3 Sex Male 11 (44%) 17 (74%) 17 (71%) 45 ( 62%) Female 14 (56%) 6 (26%) 7 ( 29%) 27 (38%) Race/Ethnicity African American/Black 10 (40%) 4 (17%) 7 (29%) 21 (29%) White 12 (48%) 14 ( 61%) 13 (54%) 39 ( 54%) Hispanic/Latino 2 (8%) 3 (13%) 4 (17%) 9 ( 13%) Mixed Race 1 (4%) 2 ( 9%) 0 (0%) 3 (4%) Lunch Status Full -price 11 (44%) 9 (39%) 9 (38%) 29 (40%) Free or reduced-price 14 ( 56%) 14 ( 61%) 15 ( 62%) 43 ( 60%) ESOL 1 (4%) 2 (9%) 1 (4%) 4 (18%) ESE (Speech-Language) 2 (8%) 1 (4%) 2 (8%) 4 ( 6%) Title I 7 (28%) 3 (13%) 2 (8%) 12 ( 17%)

PAGE 65

65 Measures Screening Measure As previously mentioned, scores from a measure of invented spelling (Lane & Pullen, 2004) were obtained for all kindergarteners to identify potential participants for this study. An invented spelling measure assesses an individuals abili ties to process phonemes in words and represent those phonemes with graphemes. T hus, invented spellings provide specific information related to a respondents developmental phase of word recognition. In addition, an i nvented spelling measure may be group administered making it the most practical means for screening a large number of students in a short period of time. The invented spelling meas ure used for this study consisted of 10-items that we re administered in the same manner as a traditional spelling test These items were scored based upon t he graphophonemic accuracy of the students written response (Lane & Pullen). In other words, if a student represented each sound in a word with a viable grapheme, then they would earn a score of 4 out of 4 for the word. Partial-alphabetic representation would earn a score less than 4. A student may earn up to 40 points on the measure, and a score of 19.5 in the spring of kindergarten places a student at the 50th percentile. Students scoring within the range of 15 to 28 on this screening generally demonstrated partial-alphabetic knowledge, and thus, most of these students were recruited for participation in this study. Students scoring above 28 generally demonstrated or were approaching full-alphabetic spellings, and were not eligible for the study. Students with scores that fell below 15 were not likely to have enough letter sound knowledge to participate in the study intervention. However, to ensure that any student meeting the criteria for inclusion in the study was not overlooked, all spellings

PAGE 66

66 were inspected by the researcher and several students with marginal scores (e.g., a score of 29 or 30) but with spellings indicating partial-alphabetic knowledge were also invited to participate. In addition, for all students with scores below 15 on the invented spelling measure, FAIR BS/PMT scores were inspected. The purpose of this inspection was to determine if any low-scoring spellers possessed sufficient phonemic awareness and knowledge of letter sounds for participation in the study. The FAIR is comprised of screening, diagnostic, and progress monitoring assessments that are administered to students in grades K through 12 in Florida (Florida Department of Education, 2009). The Broad Screen/Progress Monitoring Tool (BS/PMT) is typically administered to students three times per year. As mentioned earlier in this chapter, this measure yields a Probability of Reading Success (PRS) score, which indicates a students chance of scoring at or above the 40th percentile on end-of-year reading assessments. In kindergarten, the BS/PMT assesses letter -name knowledge and phonemic awareness in the fall, le tter -sound knowledge and phonemic awareness in the winter, and phonemic awareness and word reading in the spring. This measure is administered through standardized procedures, and reliability and validity have been well established, as detailed in the FAIR Technical Manual (Florida Department of Education, 2009). The PRS score is interpreted in the following way: kindergarteners with a score of 85% or above on the BS/PMT demonstrate high probability of success; kindergarteners with a score between 16% an d 84% demonstrate moderate probability for success; and kindergarteners scoring 15% or below demonstrate a low probability for success on end-of -year reading assessments. Students who score below 85% on

PAGE 67

67 the mid -year BS/PMT are likely to be experiencing difficulty on tasks that are highly predictive of reading success (i.e., letter-sound and phonemic awareness) and may be considered at risk for later reading difficulties. For the purpose of this screening and to include students for participation in this study, any student earning a score below 15 on the invented spelling measure but with a FAIR score of 75 or greater was also invited to participate. The rationale for this was that, regardless of their low invented spelling scores, these students likely possessed sufficient phonemic awareness and letter sound knowledge to participate in the intervention planned for this study. Pre-test/Post-test Measures The Word Attack subtest (WRMT-WA) of the Woodcock Reading Mastery Test s Revised/Normative Update (Woodcock, 1998) was administered at preand post-test. This subtest measures a students ability to apply phonic and structural analysis skills to decode unfamiliar words. Students are required to read aloud a list of pseudowords or words with very low frequency of occurrence in English. The list contains 45 pseudowords of increasing difficulty and students earn a raw score (i.e., total number of words read correctly), which may also be converted into a standard score that is centered around a mean of 100 with a standard deviation of 15 According to the publishers technical information, this subtest correlates with several other tests of reading including Woodcock-Johnson Psycho-Educational Battery (WJ), Iowa Tests of Basic Skills (ITBS), Iowa Tests of Educational Development (ITED), Peabody Individual Achievement Test (PIAT), and Wide Range Achievement Test (WRAT). In addition, internal consistency for the test median is reported to be .91, and split -half reliabilities range from .86 to .99 (Woodcock, 1998). For this study, the WA subtest served as a standardized measure of decoding accuracy. Raw scores were converted to standard

PAGE 68

68 scores for the analysis of data. Alternate forms of this assessment were used for preand post-test. The Phonetic Decoding Efficiency subtest (TOWRE-PDE) of the Test of Word Reading Efficiency, Second Edition (Torgesen, Wagner, & Rashotte, 2012) was also administered preand post-test. This subtest is an assessment of word-level decoding automaticity. Specifically, it measures the number of pseudowords that can be accurately decoded in 45 seconds. Students are provided a list of words to read aloud. After 45 seconds, the test is discontinued and the number of words read correctly is totaled. Average alternate-forms and test-retest reliabilities reportedly exceed .90 for this subtest (Torgesen et al.). For this study, the PDE subtest served as a standardized measure of decoding automaticity. However, raw scores were used in the analysis because the conversion table in the examiners manual did not provide discrete standard scores for some of the lowest raw scores obtained by participants. Alternate forms of this assessment were used at preand post-test. In addition to these standardized measures, several researcher-developed decoding measures were administered at preand post-test. These measures were created to provide more sensitive measures of decoding accuracy and automati city than would be expected when using standardized measures. In other words, the items on the standardized decoding assessments quickly become increasingly more difficulty, and kindergarteners would not be expected to read many words on these measures. Thus, decoding probes were designed to be more sensitive to the specific word structures targeted in the lessons created for this study. Probes consisted of words with

PAGE 69

69 patterns that were included in the lessons, which were carefully selected to represent a wide variation of decodable letter -sound combinations that appear in English. The probes were also designed to assess decoding skill across three dimensions: word type (i.e., real words vs. pseudowords), word complexity (i.e., linear vs. nonlinear words), and decoding task (i.e., accuracy vs. automaticity). For the purpose of this study, accuracy refers to the students ability to decode words correctly, and automaticity refers to the rate with which a student decodes words and is an indication of confidence with a skill. Word linearity is described and examined by Uhry and Shepherd (1993). Words that are linear consist of one -to-one letter -sound relationships permitting the reader to sound out each letter from left to right to blend the sounds together. Thus, linear words represent bas ic consonant-vowel -consonant (CVC) short-vowel words and their variations, including CCVC and CVCC combinations. Nonlinear words are phonetically regular, but letter sounds are dependent upon other letters within the word. For example, the long /i/ sound in the word kite is determined by the final silent e, and the sound of the /a/ in rain is determined by the adjacent i. Although the effects of word linearity were not specifically tested in the analysis of this study, it was important to include both non-linear and linear words in assess ment because both were included in the instructional intervention. Eight probes were developed: linear real-word accuracy, nonlinear real-word accuracy, linear pseudoword accuracy, nonlinear pseudoword accuracy, linear realword automaticity, nonlinear real-word automaticity, linear pseudoword automaticity, and nonlinear pseudoword automaticity. The accuracy probes consisted of four plates containing five words each for a t otal of 20 words per probe. Administration of an

PAGE 70

70 accuracy probe was discontinued if a participant did not read any of the words on a given plate correctly. This ceiling rule was established to reduce participant fatigue and frustration. Each accuracy probe yielded a total number of words read correctly with a total possible score of 20. Automaticity, or rate, probes contained 50 words each and yielded a total number of words read correctly during a one-minute timed reading. For all probes, words were randomly ordered using a set o f random numbers obtained from the number sequence generator located on random.org. The probes were administered to all participants in the same order at preand post-test. In addition, alternate forms were created for use at preand post-test by randomly altering the order of word presentation. The probes were administered to all participants in the same order at pre and post-test. For the analysis of this study, participants raw scores on each pair of linear and nonlinear probes were combined to create four researcher-developed measures: realword accuracy (R-Acc), pseudoword accuracy (P-Acc), real-word automaticity (R-Auto), and pseudoword automaticity (P-Auto). Forms A and B of these researcher-developed measures are included in the appendices. Appendix B contains the r eal-word accuracy protocols and stimulus plates; Appendix C contains the pseudoword accuracy protocols and stimulus plates; Appendix D contains the r eal-word automaticity protocols and stimulus plates; and Appendix E contains the p seudoword automaticity protocols and stimulus plates. Assessment Procedure All measures were administered to participants individually by either the researcher or another test administrator trained in proper assessment procedures. Assessment administrators were doctoral students and undergraduate students

PAGE 71

71 completing coursework to obtain their r eading endorsement. The assessment environment was one of two large, well-lit rooms with a dequate seating and table space. The standardized measures were administered according to the procedures detailed in the examiners manuals. For the researcher-developed probes, scripted oral directions were provided to assessors prior to testing. Measures were administered in the same order for each participant over approximately two testing sessions. Assessors were instructed to try to complete half of the assessments during the first session; however, if a participant began exhibiting signs of fatigue or frustration at any given point in time, testing was postponed until another day. Intervention Participants in the treatment and comparison conditions were scheduled to receive 15 instructional sessions of approximately 20 minut es in duration. All instruction was provided by the researcher in small-group format with 3 to 5 students per group. The small groups were formed based upon (a) skill level as determined by the invented spelling screening measure and (b) student location within the school. In other words, participants were grouped together with others of similar invented spelling scores within the same kindergarten quad. The sessions occurred daily as school and classroom schedules allowed. In order to fit all groups in per day, participants had to be pulled from all academic subject areas. To avoid pulling particular students from the same academic subject (i.e., reading, math, science/social studies) every day, each small group was pulled on an alternating schedule, thereby missing a different subject on each consecutive day. School administration and the classroom teachers preferred this scheduling method,

PAGE 72

72 and this also ensured that participants received a comparable amount of classroom reading instruction throughout the course of the study. Due to absences and occasional conflicts with classroom schedules, some individual participants or small groups missed planned sessions. In order to be included in the analysis for this study, participants were required to receive at least 13 out of the 15 instructional sessions with no more than one lesson missed per each of the three lesson types. Therefore, many make-up sessions were conducted over the course of the instructional term of the study. Some make-up sessions were conducted one-onone and others were conducted with small -groups of 3-5, and this was determined by lesson need. The average number of sessions completed for the treatment group was 14.36. T he average number of sessions completed for t he comparison group was 14.09 after dropping one student who moved, thereby missing the last three lessons and the post-test assessment. Treatment and comparison lessons were planned for each session and included a list of words and the order of presentation. These lessons were followed closely to ensure a systematic presentation of word patterns across time as well as a consistent presentation of words across treatment and comparison groups to eliminate the possibility of contamination in the comparison group Detai ls about instruction for each condition will be presented in the following sections. Treatment Condition Participants in the treatment condition received approximately 20 minutes of small-group word work instruction that incorporated real words and pseudowords equally. Each lesson provided participants with multiple opportunities to decode and encode words using manipulative letters and Elkonin boxes. An instructional word list

PAGE 73

73 was generated for each lesson, and the presentation of words across lessons followed an orthographic progression from simple to more complex (i.e., basic linear words preceded nonlinear words). Specifically, the first five lessons targeted basic, shortvowel CVC words; the next five lessons targeted CCVC and words; and the last five lessons targeted nonlinear words (i.e., CVVC words containing vowel teams and CVCe words). Each of the first 10 lessons in the treatment condition exposed students to a total of six real words and six pseudowords, which were derived from two or three letter patter ns. For example, the first lesson included CVC words from the letter patterns ap, at, and it. The last five lessons exposed students to a total o f eight real words and eight ps eudowords, with all derived from two letter patterns except the last lesson, which incorporated several of the previously learned letter patterns Words selected for all treatment lessons are provided in Appendix F. Word patterns were carefully selected for instruction based upon their frequency in English and their utility to students. The patterns offering the most possible real-word and pseudoword choices were selected to ensure enough words for instruction as well as for assessment purposes. To the greatest extent possible, real words were selected based upon t heir utility and likeliness that they would be known or u sed by kindergarteners in their oral language. A few words did not meet these criteria due to limited word possibilities for some word patterns. Participants in the treatment group were exposed to each word one time during a word work lesson resulting in a total of 12 different practice items for the first 10 lessons and 16 different practice items for the last

PAGE 74

74 5 lessons. Different letter patterns were introduced during each lesson, exposing students to new words during every lesson. Each lesson began with word work using manipulative letters. The instructor and participants in the group used a small magnetic board and magnetic letters to construct the words selected for the lesson. The instructor presented the first word by placing, and having participants place, the letters for the word onto their own boards in the correct order. The instructor model ed decoding the word by verbalizing each sound and blending the sounds t ogether, demonstrating to students how to use the index finger as a guide while blending. The participants were then asked to help the instructor read the word by verbalizing each sound and blending the sounds together while using their index fingers as a guide. Then, the instructor asked the participants to read the word on their own. After this initial training with the first word, the instructor asked students to change the first letter of the word to form the next word on t he list, and the process was repeated until all students demonstrate d understanding. Once the participants were given several opportunities to blend word s this way, the instructor incorporated encoding into the lesson by asking students to change individual letters to make a specific word. For example, with the word sit on the white board, the instructor would ask, What letter do we need if we want to change the word sit to lit? [Students respond with l.] Good, show me where to put l to make the word lit. All subsequent word work with manipulative letters was conducted in this way, alternating between decoding and encoding practice after every two to three words until all words for the lesson were presented. For the treatment group, half of the real words and half of the

PAGE 75

75 pseudowords selected for the lesson were presented during word work with manipulative letters. Once work with manipulat ive letters was completed, participants were provided with a dry-erase board and a marker to conduct word work with Elkonin boxes. The instructor began this portion of each lesson by verbalizing the first word from the list and demonstrating to the participants how to segment the word verbally while using his or her fingers to count the phonemes in the word. To preserve time, the dry-erase boards contained a set of three and a set of four connected boxes so the students would not need to draw their own boxes The instructor modeled selecting the correct number of boxes to represent each sound in the word by verbally segmenting each sound in the word while counting the sounds and then counting the number of boxes that were needed. Next, the instructor model ed the process of identifying the letter that corresponded with each sound, and guiding the students in writing each corresponding let ter into the corresponding box. Finally, the students were given the opportunity to do each step independently. This process was repeated for all subsequent words until students demonstrated understanding and gradually assume d responsibility for segmenting the words with limited instructor modeling. Participants segmented and wrote all words into Elkonin boxes following these procedures with guidance and prompting from the instructor as needed. For the treatment group, half of the real words and half of the pseudowords selected for the lesson were presented during word work with Elkonin boxes. To promote decoding automaticity, a word-reading task was also prepared for each lesson using the words that were taught in the lesson. The task involved the

PAGE 76

76 presentation of index cards with one taught word printed on each card. One-by-one, individual students were asked to decode the presented word. After a word was decoded, all students in the group were asked to repeat the word. When a student wa s unable to d ecode the word independently, the instructor guided the group in decoding that word sound-by-sound and then saying the word quickly as a whole. As a result of time constraints, and to ensure that all the target words for a lesson were presented during word work, this lesson component was not completed daily as planned. Instead, it was completed during the last lesson for each of the three lesson types and as time permitted. Thus, for all small groups in the treatment condition, this word-reading task occurred a total of three times. Comparison Condition Participants in the comparison group also received approximately 20 minutes of small-group word work instruction using manipu lative letters and Elkonin boxes. The procedures for the comparison condition were identical to those described for the treatment condition; however, participants in the comparison group were provided with opportunities to decode and encode real words only. An instructional word list was generated for each lesson, and the presentation of w ords across lessons followed the same orthographic progression from simple to more complex (i.e., basic linear words preceded nonlinear words) as was followed for the treatment group. Each of the first 10 lessons in the comparison condition exposed students to the same six real words presented in the corresponding treatment group lesson. Each of the last five lessons exposed participants to the same eight real words pre sented in the corresponding treatment lesson. Words for each lesson were derived from two or three different letter patterns. Words selected for all comparison lessons are provided in Appendix G.

PAGE 77

77 In this condition, all real words were practiced two times during a word work lesson once using manipulative letters and once using Elkonin boxes for a total of 12 practice opportunities during the first 10 lessons and 16 practice opportunities during the last five lessons. Thus, participants in the comparison group were provided with the same number of practice opportunities as participants in the treatment group, but they were exposed to repetition of the same real words resulting in less words overall and fewer letter-sound combinations The treatment group was not exposed to repetition of words but was exposed to the addition of pseudowords, which provided more words overall and a greater variety of letter-sound combinations Students in the comparison condition also completed the wo rd-reading task with lesson words printed on cards. However, as with the treatment condition, this activity was only conducted with comparison participants during the last lesson for each lesson type for a total of three times. Control Condition Participants in this group were assessed at preand post-test but did not receive any instruction from the researcher, thereby serving as a no-treatment control. This group of students remained in their classrooms and completed all regularly planned classroom instruction for the duration of the study. Although providing the control group with the same amount of additional time spent on a reading-related activities was initially considered to control for the effects of the additional time provided to both intervention groups as well as to control for reactivity (i.e. Hawthorne Effect), providing instruction to this group was not feasible due to limited resources. This must be recognized as a possible limitation of the study.

PAGE 78

78 Instructional Fidelity To assess fidelity, a to tal of 36 out of 180 scheduled lessons were observed by a graduate student who was trained on the lessons and procedures of this study. Half of the observations were completed during treatment lessons and half during com parison lessons, resulting in fidelity for 20% of the lessons in each condition. Two fidelity instruments were designed (i.e., one for treatment and one for comparison) to address five of elements of fidelity identified by the Institute of Educatio n Sciences (IES). These elements included adherence, quality, responsiveness, exposure, and contamination. The observer was asked to indicate whether or not (a) lessons were followed closely and completed (i.e., adherence); (b) the instructor provided adequate direction, wait time, feedback, etc. (i.e., quality); (c) the instructor made efforts to include all students in the lessons (i.e., responsiveness); (d) lessons lasted approximately 20 minutes and all target words were presented (i.e., exposure); and (e) pseudowords were taught in the treatment condition only (i.e., contamination) Each item required a yes or no answer. Lesson plans for both conditions were provided to the observer during observations to use as a reference while completing the fidelity forms. The fidelity instruments for the treatment and comparison conditions are provided in Appendix H and Appendix I, respectively. Using these methods, instructional f idelity was calculated to be 98% for the treatment group and 100% for the comparison group. Analysis Analysis of covariance (ANCOVA) was selected as an omnibus test of the hypothesized intervention effect to analyze each post -test score in this study, with corresponding pre-test score as covariate and condition as inde pendent variable. ANCOVA is used to determine if there are statistically significant differences between

PAGE 79

79 adjusted means of multiple groups. It is an extension of a one-way analysis of variance (ANOVA) in that it incorporates a covariate that is linearly related to the depe ndent variable. As such, ANCOVA is appropriate when post -test performance on a measure is believed to depend in part on pre -test performance of that same measure; t hus, the pre-test score is included as a covariate to statistically reduce error variance and remove the effect of extraneous variables. To determine if ANCOVA is suitable for analyzing data, se veral assumptions must be met. As with a one-way ANOVA, the assumptions of normality, homoscedasticity (i.e., homogeneity of variances ), and independence must be met. In addition, linearity and homogeneity of regression slopes must also be evaluated when using ANCOVA. There are standard procedures for testing these assumptions and evaluating the appropriateness of ANCOVA for the data. In general, the assumptions of linearity and homogeneity of regression slopes must be tested using the pre-test and post-test scores. If either assumption is violated, the ANCOVA model should not be used to explain the data If the relationship between preand post-test scores is linear, and if there is no interaction in the regression slopes, the n ANCOVA is a suitable method of analysis. The assumptions of normality and homoscedasticity should then be examined using residuals from the ANCOVA model. For this study, a separate ANCOVA was performed for each post-test measure using the procedures described above with each corresponding pre-test score as the covariate to determine whether or not there were statistically significant differences betwe en the group means. The scores from the four researcher-developed measures (i.e., R -Acc, P-Acc, R-Auto, and P-Auto) and from the two standardized measures (i.e.,

PAGE 80

80 WRMT-WA and TOWRE-PDE) each served as dependent variables in this series of analyses. SPSS software was used to complete the analyses. For each analysis, linearity was assessed first through the visual inspection of scatterplots of post-test scores against corresponding pre -test scores. The assumption of homogeneity of regression slopes was also tested to determine if there were any statistically significant interaction terms between pre-test scores and the independent variable (i.e., groups). In cases where the interaction term was significant the interaction model was used to explain the data W hen there was no significant interaction term, the ANCOVA procedure was performed and s tandardized residuals were plotted and then inspected visually to assess the assumptions of normality and homoscedasticity. Levenes Test of Homogeneity of Variance was also reported to verify homoscedasticity. When a statistically significant difference between group means was detected, follow-up pairwise comparisons were also completed using adjusted means for each ANCOVA. Because multiple ANCOVAs were applied to the data in this study the probability of a Type I error was inflated Type I error accumulates for each statistical test that is applied; t hus, the probability level at which a null hypothesis will be incorrectly rejected (i.e., !) should be set more stringently (Rutherford, 2011). For this analysis, the Bonferroni adjustment was applied t o control for an inflated Type I error rate. This adjustment is determined by dividing a standard alpha of .05 by the total number of tests run on a family of statistical tests. The dependent variables in this study, as measured by the researcher-created probes and the selected subtests of the WRMT and the TOWRE are all theoretically related decoding measures. However, decoding

PAGE 81

81 accuracy is distinctly different from decoding automaticity in the way it is assessed and in the type of information it provides about decoding. For these reasons, the three accuracy measures (i.e., R-Acc, P-Acc, and WRMT-WA) were analyzed as one family and the three automaticity measures (i.e., R-Auto, P-Auto, and TOWRE -PDE) were analyzed as a separate family. As such, the Bonferroni adjustment was set at .017 using !/3, or .05/3 for each analysis. This adjustment was used for the omnibus tests and also for follow-up pairwise comparisons.

PAGE 82

82 CHAPTER 4 RESULTS The purpose of this study was to examine the effects of using pseudowords during decoding instruction on the decoding outcomes of kindergarteners in the early development of word recognition Specifically, students in the partial-alphabetic phase of word recognition were selected for the study, many who demonstrated signs of risk for reading difficulties on the mid -year Florida Assessments for Instruction in Reading (FAIR) screening. Four null hypotheses were formulated and tested to examine the effects of pseudoword instruction on real-word and pseudoword decoding accuracy and automaticity. The specific research questions addressed through hypothesis testing were: 1. What are the effects of decoding instruction with pseudowords on real-word decoding accuracy? 2. What are the effects of decoding instruction with pseudowords on pseudoword decoding accuracy? 3. What are the effects of decoding instruction with pseudowords on real-word decoding automaticity? 4. What are the effects of decoding instruction with pseudowords on pseudoword decoding automaticity? To answer the research questions, decoding post-test performances were compared for 72 participants randomly assigned to one of three conditions: treatment, comparison, and control. In the treatment condition, participants received explicit instruction in segmenting and blending r eal words and pseudowords. In the comparison condition, students received the same instruction with real words only. A total of 15 instructional sessions were provided to pa rticipants assigned to the treatment and

PAGE 83

83 comparison groups. A third group was included in the study but did not receive any instruction from the researcher, thereby serving as a business-as-usual control group. Real-word and pseudoword decoding accuracy and automaticity were measured at pre-test and post-test. Decoding accuracy was assessed using the Word Attack subtest (i.e., WRMT-WA) of the Woodcock Reading Mastery Test s and four researcherdeveloped probes. The se accuracy probes were designed to assess linear real-word accuracy, non-linear real-word accuracy, linear pseudoword accuracy, and non -linear pseudoword accuracy. To address the research questions guiding this study, raw scores for linear and non-linear probes were combined prior to analysis to yield one measure of real-word accuracy (R-Acc) and one measure of pseudoword accuracy (PAcc). Decoding automaticity was assessed using the Phonemic Decoding Efficiency subtest (i.e., TOWRE-PDE) of the Test of Word Reading Efficiency and four researcherdeveloped probes designed to assess linear real-word automaticity, non -linear realword automaticity, linear pseudoword automaticity, and non -linear pseudoword automaticity. To address the research questions, raw scores for linear and non-linear probes were combined prior to analysis to yield one measure of real-word automaticity (R-Auto) and one measure of pseudoword automaticity (P -Auto). In the following sections, the results of this study are presented beginning w ith preliminary data th at serve to describe the groups. Then, the results of the statistical analyses for decoding accuracy and decoding automaticity will be presented. Pretest Findings The s cores for pretest assessments serve as a means for describing the sample in this study. Table 4-1 provides descriptive statistics for the treatment, comparison,

PAGE 84

84 and control groups on all decoding pre-test probes and measures. Means and standard deviations are reported. Table 4-1. Descriptive statistics for pre-tests by group Pre-test Measure Treatment Group Mean (SD) Comparison Group Mean (SD) Control Group Mean (SD) WRMT Word Attack 108.12 (8.03) 107.74 (8.76) 107.96 (9.12) TOWRE Phonemic Decoding Efficiency 6.2 0 (3.18) 5.83 (3.23) 6.13 (4.23) Real-word Accuracy 8.96 (5.47) 9.83 (5.19) 10.13 (7.58) Linear Real-Word Accuracy 8.88 (5.53) 9.83 (5.19) 9.38 (6.19) Nonlinear Real-Word Accuracy 0.08 (0.28) 0.00 (0.00) 0.75 (2.69) Pseudoword Accuracy 7.52 (5.36) 8.35 (5.36) 8.38 (8.07) Linear Pseudoword Accuracy 7.24 (5.07) 8.0 0 (5.29) 7.29 (6.64) Nonlinear Pseudoword Accuracy 0.28 (0.61) 0.35 (0.93) 1.08 (2.62) Real-word Automaticity 6.68 (4.20) 8.35 (8.90) 8.13 (8.12) Linear Real-Word Automaticity 6.48 (3.94) 7.52 (6.91) 7.3 8 (7.09) Nonlinear Real-Word Automaticity 0.2 0 (0.65) 0.83 (2.39) 0.75 (1.59) Pseudoword Automaticity 3.80 (2.38) 5.35 (5.02) 5.79 (5.56) Linear Pseudoword Automaticity 3.76 (2.37) 5.0 0 (4.59) 5.42 (5.26) Nonlinear Pseudoword Automaticity 0.04 (0.20) 0.35 (0.71) 0.38 (0.71)

PAGE 85

85 To examine the possibility of statistical difference between the three groups at pre-test, each pre-test score was analyzed using a separate one -way analysis of variance (ANOVA) with condition as independent variable. Table 4-2 provides the results of the ANOVA for each pretest mean. As expected with random assignment to condition, the se analyses reveal no statistically significant differences between groups on any of the pretest measures. Table 4-2. One -way ANOVA for pre-test means Pre-test Measure df F p WRMTWA 2, 69 .012 .988 TOWRE PDE 2, 69 .072 .930 R-Acc 2, 69 .235 .791 P-Acc 2, 69 .141 .868 R-Auto 2, 69 .375 .689 P-Auto 2, 69 1.32 0 .273 Analysis of Hypothesized Intervention Effect Analysis of covariance (ANCOVA) was selected to test the hypothesized intervention effects. Post-test scores were analyzed through a series of omnibus comparisons, with corresponding pre-test scores as covariates and condition as the independent variable to determine if there were statistically significant differences between adjusted post-test means of the three experimental groups. Pre-test scores were included as covariates to statistically reduce error. Following standard procedures, the assumptions of linearity and homogeneity of regression slopes were assessed first for each comparison to determine if ANCOVA was suitable for explaining the data. When these assumptions were met ANCOVA was performed, and the standardized residuals were used to assess the assumptions of

PAGE 86

86 normality and homoscedasticity. If significant main effects were found, then f ollow-up comparisons were also completed. These analyses and the results will be presented for each dependent measure in the following sections. Results for Decoding Accuracy Three measures of decoding accuracy were administered at pre -test and posttest. They included the real-word accuracy (R-Acc) and pseudoword accuracy (P -Acc) measures developed by the researcher and the Word Attack subtest of the Woodcock Reading Mastery Test s (WRMT-WA), which served as a standardized measure of pseudoword decoding accuracy. A series of ANCOVAs was conducted to d etermine if significant differences between groups existed at post -test for decoding accuracy. For the purposes of this analysis, the three measures of decoding accuracy were used to make the Bonferroni familywise adjustment for Type I error. Real -Word Ac curacy Preliminary analyses verified a linear relationship between pre and post-test RAcc for each group, as assessed by visual inspection of a scatterplot In addition, homogeneity of regression s lopes was verified as the interaction term was not statically significant, F(2,66) = .469, p = .628 Standardized residual scatterplots revealed the assumptions of normality and homoscedasticity were met However, there was one outlier observed in the data with a standardized residual greater than t hree standard deviations above zero. As a result, this outlier was removed from the data and the analysis was repeated. With the outlier removed, linearity was verified again through visual inspection of a scatterplot. Homogeneity of regression slopes wa s also verified as the interaction term was not statically significant, F(2,65) = .037, p = .963. As such, ANCOVA was

PAGE 87

87 performed with the interaction term removed to determine the main effect of decoding instruction with pseudowords on post -test R-Acc after controlling for pre-test R-Acc. The summary of the ANCOVA for R-Acc is provided in Table 4-3. The assumptions of normality and homoscedasticity were met based upon visual inspection of scatterplots of the standardized residuals Levenes Test of Homog eneity of Variance further verified the assumption of homoscedasticity, F(2,68) = .282, p = .755. Using the Bonferroni adjustment (! = .05/3 = .017) the results of the ANCOVA reveal a statistically significant difference between groups on the R -Acc measure, F(2,67) = 7.680, p = .001. The effect size is large ("p 2 = .187) based upon standards for partial eta squared established by Cohen (1969 ). Table 4 -3. Summary of analysis of covariance for R-Acc Source SS df MS F p "p 2 Pre R-Acc 2725.476 1 2725.476 241.489 .000 .783 Group 173.358 2 86.679 7.680 .001* .187 Error 756.141 67 11.286 *Significant at p < .017 Follow -up p airwise comparisons were also performed using the Bonferroni adjustment (! = .05/3 = .017). The summary of pairwise comparisons for R-Acc is provided in Table 4-4. A statistically significant difference was found between the treatment and control groups (p = .002) and between the comparison and control groups (p = .001), with adjusted post -test R-Acc means for the treatment group (M = 14.650 ) and the comparison group (M = 15.019) significantly higher than for the control group (M = 11.540). There was no statistically significant difference between treatment and comparison groups (p = .708). Adjusted group means are provided in Table 4 -5.

PAGE 88

88 Table 4-4. Pairwise comparisons for R-Acc (I) Group (J) Group Mean Difference (I-J) Std. Error p Control Treatment -3.110 .971 .002 Comparison -3.479 .980 .001 Treatment Control 3.110 .971 .002 Comparison -.369 .981 .708 Comparison Control 3.479 .980 .001 Treatment .369 .981 .708 *Significant at p < .017 Table 4 -5. Adjusted group means for R-Acc Group n Mean Std. Error Control 24 11.540 .686 Treatment 24 14.650 .686 Comparison 23 15.019 .701 Pseudoword Accuracy Preliminary analysis verified a linear relationship between pre and post-test PAcc scores for each group, as assessed by visual inspection of a scatterplot. The interaction term was not statistically significant F(2,66) = .295 p = .745 thereby verifying homogeneity of regression slopes Thus, an ANCOVA was performed with the interaction term removed to determine the effect of decoding instruction with pseudowords on post-test P-Acc after controlling for pre-test P-Acc. Standardized residuals for the interventions and for the overall model were normally distributed and there were equal variances as assessed visually through inspection of a scatterplot. Levenes Test of Homogeneity of Variance also verified homoscedasticity, F(2,69), = .212, p = .809. There were no outliers as all standardized residuals fell between plus and minus three standard deviations from zero.

PAGE 89

89 Using the Bonferroni adjustment ( = .05/3 = .017), the results reveal no statistically significant difference between groups on P -Acc, F(2,68) = 2.970, p = .058 partial "2 = .08 ; thus, post-hoc analyses were not performed. The summary of the ANCOVA for P-Acc is provided in Table 4-6. Adjusted group means are provided in Table 4 -7. Table 4-6. Summary of analysis of covariance for P-Acc Source SS df MS F p "p 2 Pre P-Acc 1814.483 1 1814.483 67.986 .000 .500 Group 158.515 2 79.258 2.970 .058 .080 Error 1814.867 68 26.689 Table 4 -7. Adjusted group means for P-Acc Group n Mean Std. Error Control 24 10.172 1.055 Treatment 25 12.681 1.035 Comparison 23 13.733 1.078 Woodcock Reading Mastery Tests Word Attack Preliminary analyses indicated a linear relationship between pre and post-test WRMT-WA for each group, as assessed by vis ual inspection of a scatterplot. However, the test for homogeneity of regression slopes revealed a statistically significant interaction term, F(2,66) = 3.963 p = .024 Upon further inspection, two outliers were observed with standardized residuals greater than three standard deviations from zero. These outliers were removed from the data and the analysis was repeated. After removing the outliers, linearity was verified again for each group through visual inspection of a scatterplot of the relationship between preand post-test WRMTWA. Inspection of a standardized residual scatterplot indicated the assumptions of normality and homoscedasticity were both met, a nd Levenes Test of Homogeneity of

PAGE 90

90 Variance also verified the assumption of homoscedasticity, F(2,67) = .582, p = .562. However, the test for homogeneity of regression slopes indicated another statistically significant interaction term, F(2,64) = 6.977 p = .002. As a result, the ANCOVA model was not suitable for analyzing the data for WRMT-WA. Instead, the interaction model was used and plotted for interpretation. Table 4 -8 provides the summary of the ANCOVA with the interaction term for WRMT-WA. The interaction between group and the pre-test covariate is plotted in Figure 4-1. Table 4 -8. Summary of analysis of covariance for WRMT -WA Source SS df MS F p Condition 217.700 2 108.850 7.803 .001 Pre WRMT 1680.530 1 1680.530 120.468 .000 Group*Pre WRMT 194.659 2 97.330 6.977 .002* Error 892.798 64 13.950 Total 895213.000 70 *Significant at p < .05 Figure 4-1. Interaction between group and pre -test covariate for WRMT-WA. The interaction depicts standard scores for the WRMT -WA, which are based on a mean score of 100 with a standard deviation of 15.

PAGE 91

91 The interaction illustrates that differences between groups on post -test WRMTWA are greater when pre-test scores are low than when pre-test scores are high. Of the p articipants with low pre-test scores, those in the comparison group scored higher at post-test than those in the control group, and those in the treatment group scored the highest of all three groups on post -test WRMT-WA. For participants with high pre-test scores, there are minimal differences between groups at post-test. Results for Decoding Automaticity Three measures of decoding automaticity were administered at pre -test and post-test. They included the real-word automaticity (R-Auto) and pseudoword automaticity (P-Auto) measures developed by the researcher and the Phonemic Decoding Efficiency subt est of the Test of Word Reading Efficiency (TOWRE-PDE), which served as a standardized measure of pseudoword decoding automaticity. A series of ANCOVAs was conducted to determine if significant differences between groups existed at post -test for decoding automaticity. Three measures of decoding automaticity were used to make the Bonferroni familywise adjustment for Type I error. Real -word automaticity Preliminary findings indicated a linear relationship between pre and post-test RAuto scores for each group, as assessed by visual inspection of a scatterplot. However, the test for homogeneity of regression slopes revealed a statistically significant interaction term, F(2,66) = 6.160 p = .004 Upon inspection of the standardized residuals from the ANCOVA model, two outliers were noted and removed from the data and the analysis was repeated. After removing the outliers, linearity was verified through visual ins pection of a scatter plot, and the test for homogeneity of regression slopes verified no significant

PAGE 92

92 interaction term, F(2,64), = 1.121, p = .332. The assumptions of normality and homoscedasticity were also met based on visual inspection of the standardized residual scatterplots. Levenes Test of Homogeneity of Variance further verified the assumption of homoscedasticity, F(2,67) = 2.666, p = .077. As a result, the ANCOVA model was performed again with the interaction term and two outliers removed to determine the main effect of decoding instruction with pseudowords on post -test R-Auto after controlling for pre-test R-Auto. The summary of ANCOVA for R-Auto is provided in Table 4-9. Using the Bonferroni adjustment ( = .05/3 = .017), the analysis reveals a statistically significant difference between groups on the R-Auto measure, F(2,66) = 5.121, p = .009. There is a large effect size ("p 2 = .134) based upon the standards for partial eta squared established by Cohen (1969 ). Table 4 -9. Summary of analysis of covariance for R-Auto Source SS df MS F p "p 2 Pre R-Auto 2778.292 1 2778.292 305.165 .000 .827 Group 93.592 2 46.796 5.121 .009* .134 Error 603.073 66 9.137 *Significant at p < .017 Follow -up pairwise comparisons were also performed using the Bonferroni adjustment (! = .05/3 = .017). The summary of pairwise comparisons for R-Auto is provided in Table 4-10. A statistically significant difference is found between the treatment and control groups (p = .009) and between the comparison and control groups (p = .006), with adjusted post -test R-Auto means for the treatment group (M = 10.413) and the comparison group ( M = 10.612) significantly higher than the control group (M = 8.067). There is no statistically significant difference between treatment and comparison groups (p = .825). Adjusted group means are provided in Table 4 -11.

PAGE 93

93 Table 4 -10. Pairwise comparisons for R-Auto (I) Group (J) Group Mean Difference (I-J) Std. Error p Control Treatment -2.346 .876 .009 Comparison -2.544 .895 .006 Treatment Control 2.346 .876 .009 Comparison -.198 .892 .825 Comparison Control 2.544 .895 .006 Treatment .198 .892 .825 *Significant at p < .017 Table 4 -11. Adjusted group means for R-Auto Group n Mean Std. Error Control 24 8.067 .619 Treatment 24 10.413 .618 Comparison 22 10.612 .645 Pseudoword Automaticity Preliminary findings verified a linear relationship between pre and post-test PAuto for each group, as assessed by visual inspection of a scatterplot. Homogeneity of regression slopes was verified by an interaction term that was not statistically significant, F(2,66) = .778 p = .464 As such, an ANCOVA was performed with the interaction term removed to determine the effect of decoding instruction with pseudowords on post -test P-Auto when controlling for pre-test P-Auto. The summary of ANCOVA for P-Auto is provided in Table 4-12. Visual inspection of scatterplots of the standardized resid uals indicated that the data were normally distributed and there was equal variance for all three groups. Levenes Test of Homogeneity of Variance further verified the assumption of homoscedasticity, F(2,69) = .859, p = .428. There were no outliers in the data as all standardized residuals fell between plus and minus three standard deviations from zero.

PAGE 94

94 Using the Bonferroni adjustment ( = .05/3 = .017), the analysis reveals a statistically significant difference on post -test P-Auto between the groups, F(2,68) = 4.970 p = .010. The effect size is large ("p 2 = .128) based upon the standards for partial eta squared established by Cohen (1969). Table 4 -12. Summary of analysis of covariance for P-Auto Source SS df MS F p "p 2 Pre P-Auto 2335.513 1 2335.513 117.515 .000 .633 Group 197.537 2 98.769 4.970 .010* .128 Error 1351.440 68 19.874 *Significant at p < .017 Follow -up p airwise comparisons were also performed with the Bonferroni adjustment (! = .05/3 = .017). The summary of pairwise comparisons for P-Auto is provided in Table 4-13. A statistically significant difference exists for P-Auto post-test means between treatment and control groups (p = .014) and between comparison and control groups (p = .005), with adjusted post -test P-Auto means for the treatment group (M = 9.735) and the comparison group ( M = 10.236) significantly higher than the control group (M = 6.466). There is no statistically significant difference between treatment and comparison groups (p = .702). Adjusted group means are reported in Table 4 -14. Table 4 -13. Pairwise comparisons for P-Auto (I) Group (J) Group Mean Difference (I-J) Std. Error p Control Treatment -3.270* 1.296 .014 Comparison -3.770* 1.302 .005 Treatment Control 3.270* 1.296 .014 Comparison -.501 1.301 .702 Comparison Control 3.770* 1.302 .005 Treatment .501 1.301 .702 *Significant at p < .017

PAGE 95

95 Table 4 -14. Adjusted group means for P-Auto Group n Mean Std. Error Control 24 6.466 .915 Treatment 25 9.735 .902 Comparison 23 10.236 .931 Test of Word Reading Efficiency Phonemic Decoding Efficiency Preliminary analysis verified a linear relationship between pre and post-test TOWRE-PDE scores for each group, as assessed by visual inspection of a scatterplot. The interaction term was not statically significant, F(2,66) = .998 p = .374 thereby verifying homogeneity of regression slopes Standardized residual scatterplots verified the assumptions of normality and homoscedasticity were met. However, t here was one outlier observed in the data with a standardized residual greater than three standard deviations below zero. As a result, this outlier was removed from the data and the analysis was repeated. With the outlier removed, visual inspection of a scatterplot verified a linear relationship between preand post-test TOWRE-PDE for each group. Homogeneity of regression slopes was also verified as the interaction term was not statically significant, F(2,65) = .299, p = .742. As such, ANCOVA was performed with the interaction term removed to test the main effect decoding instruction with pseudowords on post -test TOWRE-PDE after controlling for pre-test TOWRE-PDE. The summary of ANCOVA for TOWRE-PDE is provided in Table 4-15. Based on visual inspection of standardized residual scatterplots, the data were normally distributed and equal variance was observed, but Levenes Test of Homogeneity of Variance indicated that the assumption of homoscedasticity was violated, F(2,68) = 4.161, p = .020. Using the Bonferroni adjustment ( = .05/3 = .017),

PAGE 96

96 results indicate no statistically significant difference for the TOWRE -PDE between groups, F(2,67) = 1.756, p = .181 partial "2 = .05 ; thus, post-hoc analyses were not performed. Adjusted group means are provided in Table 4 -16. Table 4 -15. Summary of analysis of covariance for TOWRE-PDE Source SS df MS F p "p 2 Pre TOWRE 540.329 1 540.329 82.991 .000 .553 Group 22.868 2 11.434 1.756 .181 .050 Error 436.214 67 6.511 *Significant at p < .017 Table 4 -16. Adjusted group means for TOWRE_PDE Group n Mean Std. Error Control 23 5.305 .532 Treatment 25 6.152 .511 Comparison 23 6.704 .532 Summary The purpose of this study was to determine the effect of decoding instruction with pseudowords on the decoding out comes of kindergarteners in the partial-alphabetic phase of word-reading development. Two instructional groups were exposed to 15 sessions of word work instruction with manipulative letters and Elkonin boxes. Lessons for the treatment group exposed participants to pseudowords and real words, while lessons for the comparison group included only real words. A business-as-usual control group was included for comparison. Decoding accuracy and automaticity were assessed at preand post-test using researcher-developed measures of pseudoword and real-word decoding as well as standardized measures of pseudoword decoding. The results indicate the adjusted means of the treatment and comparisons groups are significantly higher than the adjusted me ans of the control group for the

PAGE 97

97 researcher-developed measures of real-word accuracy (R-Acc), real-word automaticity (R-Auto), and pseudoword automaticity (P-Auto). For the Word Attack subtest of the Woodcock Reading Mastery Test s (WRMT-WA), there is a statistically significant interaction term with treatment participants who scored low at pre-test scoring the highest of all three groups at post-test, and control participants who scored low at pretest scoring the lowest of all three groups at post test. There is minimal difference between groups for participants with high pre-test scores on the WRMT -WA. The adjusted post-test means for the pseudoword accuracy (P-Acc) measure and the Phonemic Decoding Efficiency subtest of the Test of Word Reading Efficiency (TOWRE-PDE) are not significantly different across groups though treatment and comparison group means are higher than the control group means for both measures.

PAGE 98

98 CHAPTER 5 DISCUSSION The development of early decoding skills is fundamental to reading success for beginning readers (Ehri, 1995, 1999), and particularly for those who are at risk for reading difficulties (Adams, 1990). Many students who struggle with reading have a primary deficit in the foundational skills that lead to decoding (Chall, 1983 ; Henry, 1993; Torgesen, 1999). Effective efforts to improve these skills are paramount. Numerous studies have shown that interventions that include word work are effective at improving decoding in at-risk or struggling readers (Campbell et al. 2008; Iversen & Tunmer, 1993; Joseph, 1998; Lane et al., 2009; McCandless et al., 2003; Pullen, 2000; Pullen et al., 2005) However, few of these studies have provided insight into what specific aspects of instruction make word work more or less effective. There has been only one study to investigate the value of pseudowords for decoding instruction; yet, the results suggest that pseudowords have a promising role in the early development of decoding (Cardenas, 2009). This finding is of interest, given that pseudowords are seldom us ed during instruction and have even been discouraged based upon the conventional belief that only real words should be used for their practical utility (Adams, 2011). Nevertheless, it is known that during pseudoword reading, the possibility to rely on memory is eliminated and t he skills required for decoding (i.e., phonemic awareness and applica tion of the alphabetic principle ) are isolated (Good, Baker, & Peyton, 2009). Thus decoding instruction tasks with pseudowords force the student to decode. For students who are in the early phases of learning to decode, it may be effective to include pseudowords during decoding instruction for this reason.

PAGE 99

99 The purpose of this study was to investigate the effects of including pseudowords during decoding instruction on several decoding outcomes of beginning r eaders. Participants were 72 kindergarteners in the partial-alphabetic phase of decoding development. A pre-test post-test control group design with random a ssignment was used to compare three groups: a treatment group that received decoding instruction using a combination of real words and pseudowords during word work activities; a comparison group that received decoding instruction using only real words duri ng word work activities; and a control group that did not receive instruction from the researcher. Decoding accuracy and automaticity measures were administered at preand post-test. This design was selected to allow for comparison of groups on post-tes t performance while controlling for pre-test performance. In this chapter, a discussion of the findings and implications of the e ffects of including pseudowords during small -group decoding instruction are presented. Each research question is addressed first with a summary of related findings. Then, limitations of the study are presented. Finally, implications for practice and implications for future research are discussed. Summary of Findings This study was designed and conducted specifically to address the following research questions: 1. What are the effects of decoding instruction with pseudowords on real-word decoding accuracy? 2. What are the effects of decoding instruction with pseudowords on pseudoword decoding accuracy? 3. What are the effects of decoding instruction with pseudowords on real-word decoding automaticity?

PAGE 100

100 4. What are the effects of decoding instruction with pseudowords on pseudoword decoding automaticity? Each research question was addressed through a separate researcherdeveloped de coding measure. Standardized measures of p seudoword accuracy and pseudoword automaticity were also administered to address the effects of instruction on pseudoword decoding outcomes for the second and fourth questions, respectively, though these measures were expected to be less sensitive to the early decoding abilities of the participants in this study. The following is a presentation and discussion of findings for each research question. The E ffects of Decoding Instruction with Pseudowords on Real -Word A ccuracy Participant performance on a researcher-developed measure of real-word accuracy (i.e., R-Acc) was used to address the first research question: What are the effects of decoding instruction with pseudowords on real -word decoding accuracy? An ANCOVA was used to test the main effect of instruction with pseudowords on post -test R-Acc while controlling for pre-test R-Acc. The results indicate a statistically significant difference between groups with a large effect size. Follow -up comparisons reveal statistically significant differences with both the treatment and comparison groups outperforming the control group on real-word accuracy. There was no statistically significant difference between the treatment and comparison groups This finding indicates that kindergarteners in the treatment and comparison conditions benefitted equally from small-group decoding instruction, regardless of whether they wer e exposed to real words and pseudowords or real words alone.

PAGE 101

101 The E ffects of Decoding Instruction with Pseudowords on Pseudoword A ccuracy Participant performances on a researcher -developed measure of pseudoword accuracy (i.e., P-Acc) and the Word Attack subtest of the Woodcock Reading Mastery Tests (WRMT-WA), a standardized measure of pseudoword accuracy, were used to address the second research question: What are the effects of decoding instruction with pseudowords on pseudoword decoding accuracy? An ANCOVA was used to test the main effect of instruction with pseudowords on post -test performance of each measure while controlling for pre-test on the same measure. The analyses for these two measures offer differing results. Results for the P-Acc measure reveal no statistically significant difference between groups. This indicates that neither instruction with or without pseudowords had an effect on this pseudoword decoding accuracy measure. However, the adjusted group means of both the treatment and control groups were better than the adjusted mean of the comparison group. For the WRMT-WA measure, the ANCOVA reveals a statistically signif icant group by covariate interaction with the means of all three groups differ ing more for participants with low pre-test scores and very little for participants with high-pre-test scores. Moreover, participants with low pre-test scores in the treatment group had the highest post-test scores of all three groups, whereas participants with comparably low pre-test scores in the control group had the lowest post -test scores of all three groups. This finding is unexpected bec ause standardized assessments are sometimes not sensitive enough to measure the early decoding skills of beginning readers.

PAGE 102

102 The E ffects of Decoding Instruction with Pseudowords on Real-Word A utomaticity Participant performance on a researcher -developed measure of real-word automaticity (R-Auto) was used to address the third research question: What are the effects of decoding instruction with pseudowords on real -word decoding automaticity? An ANCOVA was used to test the main effect of instruction with pseudowords on post test R-Auto while controlling for pre-test R-Auto, and results indicate a statistically significant difference between groups with a large effect size. Follow -up comparisons reveal that both the treatment and comparison groups statistically significantly outperformed the control group on R-Auto ; however, there was no statistically significant difference between the treatment and comparison groups. This finding indicates that kindergarteners in the treatment and comparison conditions benefit ted equally from small-group decoding instruction, regardless of whether they wer e exposed to real words and pseudowords or real words alone. The E ffects of Decoding Instruction with Pseudowords on Pseudo word A utomaticity Participant performances on a researcher-developed measure of pseudoword automaticity (P-Auto), and the Phonemic Decoding Efficiency subtest of the Test of Word Reading Efficiency (TOWRE-PDE), a standardized measure of pseudoword automaticity, were used to address the fourth research question: What are the effects of decoding instruction with pseudowords on pseudoword decoding automaticity? An ANCOVA was used to test the main effect of instruction with pseudowords on post -test performance of each measure while controlling for pre -test on the same measure. Similar to the results for the pseudoword accuracy measures, the results of the analyses for these two pseudoword automaticity measures vary.

PAGE 103

103 For the P-Auto measure, the results reveal a statistically significant difference between groups with a large effect size, and follow -up comparisons reveal that both the treatment and comparison groups statistically significantly outperformed the control group on real-word accuracy. There was no st atistically significant difference between the treatment and comparison groups. Similar to the findings for the R-Acc and R-Auto measures, these findings indicate that kindergarteners in the treatment and comparison conditions benefitted equally from instruction, regardless of whether they were exposed to real words and pseudowords or real words alone. Results for the TOWRE-PDE reveal no statistically significant difference between groups, though the adjusted group means show that both the treatment and co ntrol groups performed better than the c omparison group. Although these two pseudoword automaticity measures yielded conflicting results, it was suspected that the standardized measures would not be sensitive enough to detect the early decoding skills of the participants in this study. Therefore, the findings for the TOWRE-PDE are not surprising. Discussion and Interpretation The significant effects that were found for both groups receiving decoding instruction in this study align with the findings of numerous studies that have investigated similar instruction for beginning and at -risk readers (e.g., Campbell et al., 2008; Iversen & Tunmer, 1993; Joseph, 1998; Lane et al., 2009; McCandless et al., 2003; Pullen, 2000; Pullen et al., 2005) These researchers concur that explicit decoding instruction that focuses on word work activities and includes manipulatives helps students build decoding skill. Participants in both the treatment group and the comparison group of the present study significantly outperformed the control group on

PAGE 104

104 measures of real-word decoding accuracy (i.e., R-Acc), real-word decoding automaticity (i.e., R-Auto ), and pseudoword decoding automaticity (i.e., P-Auto). These results imply that the addition of pseudowords during small-group decoding instruction was just as effective as using real words alone with students in the partial-alphabetic phase of word recognition. Furthermore, the group-by-covariate interaction for the standardized measure of pseudoword decoding accuracy (i.e., WRMT -WA) indicates a greater positive effect for participants with low pre-test scores in the treatment group than in the comparison group, and almost no effect for those in the control group with low pre-test scores. Across all three groups, there was little difference in post -test scores for participants with high pre-test scores. This finding is somewhat surprising given the expectation that a standardized measure might not be sensitive enough to detect differences among beginning decoders with such a brief intervention period between pre -test and post-test. It is promising that for this measure, the inclusion of pseudowords was more beneficial for students who started out with the lowest performance. This finding provides some evidence that students with the poorest decoding skills may benefit the most from smallgroup, explicit instruction when this instruction incorporates pseudowords. While most of the findings from this study are promising, the analysis revealed no significant results for two measures. First, there was no significant difference between groups on the standardized measure of pseudoword decoding automaticity (i.e., TOWRE-PDE) in this study. Although it would seem logical that pseudoword -trained participants would do better than non -pseudoword-trained participants on pseudoword measures, the results of the TOWRE-PDE are not surprising because this standardized

PAGE 105

105 measure was not expected to be sensitive enough for these beginning readers to progress beyond the first few items of the measure. As with many standardized measures, the items on the TOWRE-PDE rapidly become more difficult. For beginning readers, this generally means that they will not be able to read many of the test items correctly before reaching items that are beyond their abilities. Second, there was no difference between groups on the measure of pseudo word decoding accuracy (i.e., P-Acc). These results are somewhat unexpected because they conflict with the results of all three of the other researcher -developed measures One possible explanation for the low P-Acc performance is that almost all participants receiving instruction in this study had difficulty learning more complex wo rd patterns, such as the consonant blend patterns in CCVC and CVCC words and the non-linear word patterns. In addition, many participants demonstrated a common confusion between similar short vowel sounds (i.e., short a and e were confused and short i and e were confused). It was clear that these participants would need more time practicing these particular patterns to become proficient at decoding them. The linear accuracy probes created for this study presented words in random order, so simple CVC words were mixed in with more complex CCVC and CVCC words. It is possible that broad difficulties with these specific patterns affected participant performance on the P-Acc measure. Although this would be interesting to investigate further, test administration for all accuracy measures included a discontinuation rule for students who missed all of the words on any given stimulus plate. Thus, some students with low scores did not have a chance read all of the words. If they had, the CVC words could have been examined further to explore this possibility. Again, if more time was

PAGE 106

106 available to develop some of the more complex word patterns, it is quite possible that participants receiving instruction would have performed better than control participants. Finally, the significant results for the measures of real-word accuracy and automaticity are of particular interest These results indicate that instruction supported the development of skills for decoding real words, regardless of whether it involved real words only or real words and pseudowords Because this study drew upon theories of reading development and the theory of recombinative generalization, these findings are of particular interest. It is clear that both instructional approaches in this study helped students to progress within the partial-alphabetic phase of word recognition described by Ehri (1991, 1998, 2005) As expected, instruction in the treatment and comparison conditions worked to strengthen the c onnections between the phonologic and orthographic processors resulting in greater accuracy and automaticity when decoding. Furthermore, it was proposed that students who were exposed to instruction and practice using pseudowords would be able to generalize their learning to new stimuli, novel arrangements, or unfamiliar words (Goldstein, 1983; Suchowierska, 2006). From the present finding s, it is clear that participants in the treatment and comparison conditions were equally capable of reading real words Therefore, learning with pseudowords would indeed seem to generalize to decoding real words, as the theory of recombinative generalization suggests. Given more time, instruction with pseudowords may have yielded different results. In this study, students received daily instruction for approximately 15 total sessions. In contrast, Cardenas (2009) implemented phonics instruction with pseudowords for 28 days and found greater results for the pseudoword -trained group.

PAGE 107

107 It is possible that the longer instructional period in the Car denas study contributed to those positive results. Also, the present study integrated pseudowords with real words for the duration of instruction with the treatment group. In the study conducted by Cardenas, treatment participants were provided instruction with only pseudowords during the treatment phase. Perhaps that period of concentrated focus on pseudowords made a difference. Finally, Cardenas provided a period of real word phonics instruction before and after the pseudowor d instruction was implemented. It is possible that sequence of instruction had an effect on the outcome of that study. The pr esent study sought to integrate Pseudowords and real words for treatment participants to determine if the addition of pseudowords would add any value to the instruction. This was thought to be important because using pseudowords only during instruction has been contraindicated in the past by reading experts (Adams, 2011). Given what has been learned from the results of this study, 15 daily lessons may not have been enough to determine specifically whether or not the addition of pseudowords is more beneficial than using real words alone. However, this study does demonstrate the value of a short, intensive intervention using manipulative s to teach early decoding skills to students in the early development of word recognition. Overall, the findings from this study add to the body of research on decoding instruction by offering evidence that instruction including pseudowords can be effective for students who are in the beginning phases of decoding development. Although the findings of the present study do not suggest that instruction with pseudowords is superior to instruction without pseudowords, the limitations of this study must be taken into consideration. Taken together this study and the study conducted by Cardenas

PAGE 108

108 (2009) provide preliminary empirical support for pseudoword use and a basis for additional research on this approach. Limitations There are limitations to this study that should be considered when interpreting the results. The first is the limited number of participants in the study, which reduces power and restricts the external validity of the study. To achieve the desired power of .80 as recommended by Cohen (1988) and to reduce the probability of making a Type II error, a total sample size of 87 was needed for this study. Given limited resources, the study was confined to one elementary school so that all instructional sessions could be provided by the researcher. Although there was a pool of 93 potential participants, only 73 kindergarteners returned consent for participation and one of these participants moved before the completion of the study. This reduction in power limits the ability to detect differences that may exist between the treatment and comparison groups and this is noteworthy given the purpose of this study, which was to determine if pseudoword instruction resulted in significantly different effects than real-word instruction. The sample also restricts the findings of this study to kindergarteners attending one elementary school. Although these kindergarteners and the school itself are not remarkably different than kindergarteners or schools in other areas of the country, generalization is limited beyond this grade level and the geographical location of this school. A larger sample from a variety of schools and geographical locations would have increased the external validity of this study. Finally, a limitation within the design of the study is that the control group did not receive any instruction from the researcher. It is possible that treatment and

PAGE 109

109 comparison participants were influenced by the fact that they were receiving daily attention from the researcher; thus, reactivity, or the Hawthorne Effect, may have influenced the performances of participants. It is also possible that the two groups of participants had overall better results because of the additional time spent receiving reading-related instruction. To address t his limitation, the control group would have to receive the same amount of time spent on a reading activity unrelated to decoding. In the present study, this was not possible due to limited resources. Implications for Practice Notwithstanding the limita tions of this study, the findings shed some light on decoding instruction for teachers. At the very least, it is plausible to conclude from the present findings that there are no negative effects on decoding skills when including both pseudowords and real words during daily small-group instruction that targets early decoding skills. In fact, participants who received instruction with pseudowords performed comparably to participants who received instruction without pseudowords, and both groups significantly outperformed the control group. This finding challenges the opinion of some educators and scholars in the field of reading instruction who caution against the use of pseudowords for instructional purposes (Adams, 2011). Based upon the present findings teachers may incorporate pseudowords in addition to real-words during decoding instruction for the explicit purpose of providing students opportunities to isolate discrete decoding skills These findings also have implications for teacher preparation and professional development. Because many teachers have already formed an opinion about pseudowords that is not based on empirical evidence, it would be useful to clarify the role that pseudowords can have in effective decoding instruction. Providing teachers

PAGE 110

110 with a deeper understanding of the rationale for using pseudowords should help them better understand the processes involved in skillful decoding. For example, teachers should understand that pseudowords offer more opportunities for students to practice blending and segmenting particular phonemic and orthographic patterns found within the English language, and what they learn with pseudowords may later be applied in novel ways, as the theory of recombinative generalization suggests (Goldstei n, 1983; Suchowierska, 2006). As such, pseudowords can be particularly helpful when teaching patterns that are less common in English by offering extension and repetition that will serve to strengthen the connections between the orthographic and phonologic al processors and consolidate the patterns in memory. Clearly, more research on pseudowords is needed to make specific recommendations pertaining to the use of pseudowords during decoding instruction. This study could not verify that pseudowords added value to decoding instruction, but the se results in combination with results of other studies that specifically examined pseudowords (Cardenas, 2009) or included pseudowords in a larger intervention (Lane, Pullen, Hudson, & Konold, 2009; Pullen, La ne, Lloyd, Nowak, & Ryals, 2005) imply that pseudowords can be beneficial to the development of decoding. If pseudowords are used along with real words for the specific purpose of developing discrete decoding skills with students who are beginning to dec ode, this practice will not interfere with decoding development and has the potential to bolster these skills during a critical period of development Moreover, the use of pseudowords during decoding instruction may provide an advantage for students with emergent or weak decoding skills

PAGE 111

111 It should be noted that decoding instruction with or without pseudowords was only provided to participants in this study for a total of only 15 days, yet statistically significant improvements in decoding skill were observ ed. While it is possible that pseudowords may have had a different impact if instruction had been provided for a longer duration, the impact of this short, intensive intervention should not be overlooked. This study demonstrates the power of daily, small-group decoding instruction with manipulatives for students who have not yet mastered decoding skill. Implications for Future Research The results and limitations of this study, in conjunction with the previous limited research on pseudoword instruction, provide several implications for future research. In the only study found to examine the use of pseudowords for instruction, Cardenas (2009) found that students who received phonics instruction with pseudowords alone for a period of 28 school days made greater gains on a phonetic word-reading task than students who received phonics instruction with real words alone. The present study shows that 15 daily, small-group decoding instructional sessions incorporating pseudowords with real words were also beneficial for young developing decoders ; however, they were no more beneficial than 15 daily, small-group decoding instructional sessions using real words alone. Taken together, these studies provide an impetus for future research to further explain the role of pseudowords during decoding instruction. Although the present study was designed to better control for systematic differences between groups, the limitations of this study should be addressed through additional research. To begin larger studies that isolate pseudoword instruction as the independent variable will be beneficial to address the limitations that impact the power and

PAGE 112

112 generalizability of this study. This study included 72 participants in kindergart en from one elementary school in nor th central Florida. Future replication of this study should seek to include more participants in order to increase power specifically for detection of differences that may exist between the treatment and comparison conditions The findings of this study indicated no statistical differences between means of the treatment and comparison groups on any measure, but any differences may have required greater power to be detected. Also, future studies that include participants from different grade levels will increase generalizability to students in those grades. For instance, many first graders are developing decoding skil ls within the partial-alphabetic phase of word recognition, so research should examine the effects of pseudoword s with this population of students The same holds true for older students struggling with decoding skill. First-graders are likely to have had more exposure to printed words than kindergarteners and, as a result, may be able to read more words by sight It is possible that pseudowords would provide greater benefit for first-graders who are under these circumstances and not yet decoding given the rationale that decoding pseudowords forces isolation and use of decoding skill rather than reliance upon memory for words. Moreover, real words will often act as pseudowords when they are not yet part of a students lexicon. For many kindergarteners, simple CVC words such as sop, vat, or hog are not likely to hold meaning. There is no way to know if any of the real words presented during the lessons in this study were pseudowords for these participants. For these reasons, future research should investigate pseudoword use with first-graders in early decoding development and also with older students who are

PAGE 113

113 struggling with decoding skills. Support for future research with lower performing students is also offered from the finding of this study that participants with low pre-test scores improved the most on one standardized decoding measure when they re ceived instruction with pseudowords included In addition to the issues of power and generalizability, the design of this study has its limitations. Participants in the control group were included for comparison to determine if pseudowords were effective at all, and particularly because treatment and comparison groups receiving explicit decoding instruction would each be expected to make progress on decoding skills. However due to limited resources, the control group did not receive any instruction from the researcher. Thus, reactivity, or the Hawthorne Effect, may have influenced the performances of participants in this study. In future studies, a stronger design will provide control participants the same amount of attention and instruction unrelated t o decoding development in order to eliminate this limitation. Beyond addressing the limitations of this study, future research may also address many new questions that arise from the findings of this study. Reading research has shown that instruction to develop discrete skills should be paired with opportunities to apply the skills during authentic reading opportunities ( Lane, 1994 ). The instruction provided in this study was highly focused on skill developm ent. Lessons did not provide opportunities to immediately apply decoding skill in connected text, and only limited opportunities to review learned skills were provided due to the time allotted for each instructional session. Therefore, additional studies should seek to include opportunities to apply learned skills to determine if the effects on decoding differ when students are given these opportunities

PAGE 114

114 Moreover, it will also be import ant to learn whether or not decoding instruction with pseudowords has an effect on later-developing reading skills Many of the pseudowords taught during this study may be syllables or affixes found in real multisyllabic words. It would be interesting to explore whether or not learning to decode pseudowords had any effec t on multisyllabic word reading. Also, g iven that the ultimate goal of all reading instruction is for the reader to comprehend text (Adams, 1990), longitudinal studies that examine the impact of early decoding instruction with pseudowords on later -developed reading skills including text level fluency and reading comprehension would be important. Although the effects of this study were promising, the design delimits the evidence that this intervention will work in classrooms when implemented by school personnel. Future studies can address this delimitation by implementing teacher -led instruction with pseudowords. Furthermore, it will be useful to find out if decoding instruction with pseudowords can be implemented successfully by paraprofessionals in the classroom. Several studies of decoding instruction without pseudowords have suggested that paraprofessionals may be instrumental in this way (Ryder, Tunmer, and Greaney, 2008; Vadasy, Sanders, & Peyton, 2005; Vadasy, Sanders, & Tudor, 2007). In the present study, the instruction was provided in small groups, which may or not be manageable for some teachers. Thus, further research should aim to elucidate the role of paraprofessionals. Decoding instruction was only provided to treatment and control partic ipants in this study for a total of only 15 days. It is possible that pseudowords may have had a different impact if instruction had been provided for a longer duration As previously

PAGE 115

115 mentioned, Cardenas (2009) found that students receiving instruction with pseudoword alone had more growth than students who received instruction with real words alone after only 28 days of instruction. Removing real words from instruction is not widely recommended because students benefit from exposure to words that they will experience in print. However, it is possible that providing only pseudowords for a limited pe riod of time, as Cardenas did, is more ideal because it offers a concentrated period of skill development. The theoretical underpinnings of the present study suggest that pseudowords have the capacity to add to instruction with real words by providing more opportunities to practice with more letter combinations. Future research should examine whether or not an experimental design similar to the present study will result in a significant outcome for participants exposed to both pseudowords and real words when more days of instruction are provided. Conclusions Most educators agree that effective reading instruction is critical for the development of early literacy skills. Decoding instruction is particularly important because it fosters the development of foundational phonological and orthographic skills that are beneficial for all readers, but critical for struggling readers. This study was conducted to examine the effects of including pseudowords during small -group decoding intervention with kindergarten students in the partial -alphabetic phase of word recognition as described by Ehri (1991, 1998, 1995). The findings suggest that instruction that incorporates pseudowords in addition to real words can have a positive effect on the decoding outcome of kindergarteners during this phase of development. Overall, post-test decoding performances of the pseudoword -trained participants were comparable to post-test decoding performances of participants trained with only real

PAGE 116

116 words, and both groups performed significantly better than a no -treatment control group on various measures of decoding accuracy and automaticity. Because prior research specifically examining pseudowords for instruction is limited to one study, the present study offers preliminary support. The findin gs from this study also reveal implications for practice and direction for future research. Clearly, more studies are needed to determine the true value of pseudowords for developing early decoding skills. Additional research should replicate the existing research and seek to address the limitations of these studies.

PAGE 117

117 APPENDIX A IRB APPROVAL DOCUMENATION

PAGE 118

118 APPENDIX B REAL-WORD ACCURACY PROTOCOLS AND STIMULUS PLATES Real-Word Accuracy Measure Form A 1 Scoring: This measure consists of two probes. Administer both probes. For a word read incorrectly, write the students response phonemically in the space provided and mark the word as incorrect. If the student self -corrects, write the letter S next to the response. If the student does not make an attempt to read a word in 5 seconds, say, Tell me what you think that word is. If they still do not respond, mark the word as incorrect and say, Try the next word. For each probe, discontinue the probe if the student reads no words correctly on a given plate. Training Item: Tell the student, I would like you to read some words. Show the training card and point to cat. You will read a list like this from top to bottom. Sweep your finger down the list of practice words. Read these words starting at the top. If the student is not able to read a word, or if they do not blend the sounds together, say Put them together to say the whole word. Model blending the word if necessary cat men trap Directions: Tell the student, Now, I am going to ask you to read some more words. Please start at the top of the list. Be sure to read every word carefully and do your best. You may begin. Brief positive feedback may be provided (i.e., Good work.) to motivate the student. Note that this is not a timed task. LRW Accuracy Probe 1 !"#! $ 2 %&' $ 3 ()*$ 4 %+),$ 5 -+#* $ 6 +#,$ 7 *&(. $ 8 "&, $ 9 *+/, $ 10 '/! $ 11 -#* $ 12 012$ 13 31+! $ 14 2/-! $ 15 3)2 $ 16 -/3 $ 17 41!$ 18 +1'2 $ 19 !)(.$ 20 2"&* $

PAGE 119

119 Real-Word Accuracy Measure Form A 2 NLRW Accuracy Probe 1 !"#$ 2 %"&$ 3 '"($ 4 ')*! 5 +&,-$ 6 .*!$ 7 /,.$ 8 %)*9 0&,!$ 10 ($*! 11 +1*2$ 12 -,&$ 13 +'$*& 14 0&)($ 15 31)+$ 16 &*4$ 17 2$$4 18 (,'$ 19 +$$20 (%*,# Assessment Administrator: ________________________ Student: _______________________________________ Date: _________________ Time: __________________

PAGE 120

120 cat me n trap LRW Accuracy Training Items Form A cat men trap LRW Accuracy Training Items Form A trot fin cap flag slop LRW Accuracy Form A: Plate 1 log pick rig plug nut LRW Accuracy Form A: Plate 2

PAGE 121

121 trot fin cap flag slop LRW Accuracy Form A: Plate 1 log pick rig plug nut LRW Accuracy Form A: Plate 2 sop wed melt dust mad LRW Accuracy Form A: Plate 3 sum jet lend tack drip LRW Accuracy Form A: Plate 4

PAGE 122

122 sop wed melt dust mad LRW Accuracy Form A: Plate 3 sum jet lend tack drip LRW Accuracy Form A: Plate 4 dune rule tube toad slime NLRW Accuracy Form A: Plate 1 fade wife roam glide bead NLRW Accuracy Form A: Plate 2

PAGE 123

123 dune rule tube toad slime NLRW Accuracy Form A: Plate 1 fade wife roam glide bead NLRW Accuracy Form A: Plate 2 shape mile steal globe chose NLRW Accuracy Form A: Plate 3 lake peek bite seem brain NLRW Accuracy Form A: Plate 4

PAGE 124

124 shape mile steal globe chose NLRW Accuracy Form A: Plate 3 lake peek bite seem brain NLRW Accuracy Form A: Plate 4

PAGE 125

125 Real-Word Accuracy Measure Form B 1 Scoring: This measure consists of two probes. Administer both probes. For a word read incorrectly, write the students response phonemically in the space provided and mark the word as incorrect. If the student self -corrects, write the letter S next to the response. If the student does not make an attempt to read a word in 5 seconds, say, Tell me what you think that word is. If they still do not respond, mark the word as incorrect and say, Try the next word. For each probe, discontinue the probe if the student reads no words correctly on a given plate. Training Item: Tell the student, I would like you to read some words. Show the training card and point to cat. You will read a list like this from top to bottom. Sweep your finger down the list of practice words. Read these words starting at the top. If the student is not able to read a word, or if they do not blend the sounds together, say Put them together to say the whole word. Model blending the word if necessary cat men trap Directions: Tell the student, Now, I am going to ask you to read some more words. Please start at the top of the list. Be sure to read every word carefully and do your best. You may begin. Brief positive feedback may be provided (i.e., Good work.) to motivate the student. Note that this is not a timed task. LRW Accuracy 1 !"#$ 2 %&'( $ 3 )*# $ 4 #+,% $ 5 -.% $ 6 %,/0 $ 7 1&*($ 8 /*%$ 9 2+.2 $ 10 2*/0$ 11 )"&2 $ 12 3"2$ 13 &.($ 14 -&.% $ 15 #'-2 $ 16 4'2 $ 17 -') $ 18 1,4 $ 19 +,( $ 20 &"4# $

PAGE 126

126 Real-Word Accuracy Measure Form B 2 NLRW Accuracy 1 !"#$ 2 %&'#$ 3 &()$ 4 *(+$ 5 !'(+ 6 +",$ 7 -'(. 8 /0*$ 9 1!$(& 10 2$$) 11 34'1$ 12 #$(+ 13 1$$. 14 #0!$ 15 %&0+$ 16 .0&$ 17 #-(0, 18 1&0.$ 19 -"&$ 20 14(2$ Assessment Administrator: ________________________ Student: _______________________________________ Date: _________________ Time: __________________

PAGE 127

127 cat me n trap LRW Accuracy Training Items Form B cat men trap LRW Accuracy Training Items Form B wed plug mad drip sop LRW Accuracy Form B: Plate 1 pick flag cap trot tack LRW Accuracy Form B: Plate 2

PAGE 128

128 wed plug mad drip sop LRW Accuracy Form B: Plate 1 pick flag cap trot tack LRW Accuracy Form B: Plate 2 melt jet log slop dust LRW Accuracy Form B: Plate 3 nut sum fin rig lend LRW Accuracy Form B: Plate 4

PAGE 129

129 melt jet log slop dust LRW Accuracy Form B: Plate 3 nut sum fin rig lend LRW Accuracy Form B: Plate 4 tube globe lake fade toad NLRW Accuracy Form B: Plate 1 dune roam wife steal peek NLRW Accuracy Form B: Plate 2

PAGE 130

130 tube globe lake fade toad NLRW Accuracy Form B: Plate 1 dune roam wife steal peek NLRW Accuracy Form B: Plate 2 chose bead seem bite glide NLRW Accuracy Form B: Plate 3 mile brain slime rule shape NLRW Accuracy Form B: Plate 4

PAGE 131

131 chose bead seem bite glide NLRW Accuracy Form B: Plate 3 mile brain slime rule shape NLRW Accuracy Form B: Plate 4

PAGE 132

132 APPENDIX C PSEUDOWORD ACCURACY PROTOCOLS AND STIMILUS PLATES Pseudoword Accuracy Measure Form A 1 Scoring: This measure consists of two probes. Administer both probes. For a word read incorrectly, write the students response phonemically in the space provided and mark the word as incorrect. If the student self -corrects, write the letter S next to the response. If the student does not make an attempt to read a word in 5 seconds, say, Tell me what you think that word is. If they still do not respond, mark the the word as incorrect and say, Try the next word. For each probe, discontinue the probe if the student reads no words correctly on a given plate. Training Item: Tell the student, I would like you to read some words that are not real. When you read these, its ok if these words dont make sense. Show the training card and point to mip You will read a list like this from top to bottom. Sweep your finger down the list of practice words. Read these words starting at the top. If the student is not able to read a word, or if they do not blend the sounds together, say Put them together to say the whole word. Model blending the word if necessary. Directions: Note that this is not a timed task. Tell the student, Now, I am going to ask you to read some more words. Please start at the top of the list. Be sure to read every word carefully and do your best. You may begin. Brief positive feedback may be provided (i.e., Good work.) to motivate the student. LPW Accuracy Probe 1 !"#$ 2 %&'#$ 3 ()*$ 4 +,"* $ 5 %-%$ 6 ."*$ 7 /)%0 $ 8 12* $ 9 !,-' $ 10 ."3*$ 11 (2+4$ 12 5&6$ 13 !7-# $ 14 8&' $ 15 ,2# $ 16 3),*$ 17 5&3*$ 18 9-6 $ 19 /)0 $ 20 672# $ mip fet glim

PAGE 133

133 Pseudoword Accuracy Measure Form A 2 NLPW Accuracy Probe 1 !"#$% 2 &$$!% 3 '("#% 4 )*$$&% 5 +",% 6 *(&$% 7 ./"#$% 8 .,-$% 9 )$"/ % 10 +"!$% 11 0$"# % 12 1("2 % 13 & 3#$ % 14 #"&$% 15 '#,.$ % 16 4$$0% 17 5,&$ % 18 6",& % 19 )(0$ % 20 '&$"! % Assessment Administrator: ________________________ Student: _______________________________________ Date: _________________ Time: __________________

PAGE 134

134 mip fet glim LPW Accuracy Training Items Form A mip fet glim LPW Accuracy Training Items Form A bap nump zet clat nin LPW Accuracy Form A: Plate 1 jat yend fot blim jast LPW Accuracy Form A: Plate 2

PAGE 135

135 bap nump zet clat nin LPW Accuracy Form A: Plate 1 jat yend fot blim jast LPW Accuracy Form A: Plate 2 zock vug brip wum lo p LPW Accuracy Form A: Plate 3 selt vust hig yed grop LPW Accuracy Form A: Plate 4

PAGE 136

136 zock vug brip wum lo p LPW Accuracy Form A: Plate 3 selt vust hig yed grop LPW Accuracy Form A: Plate 4 pate neep soat breen vaid NLPW Accuracy Form A: Plate 1 rone flate fide beal vape NLPW Accuracy Form A: Plate 2

PAGE 137

137 pate neep soat breen vaid NLPW Accuracy Form A: Plate 1 rone flate fide beal vape NLPW Accuracy Form A: Plate 2 keat goam n ute tane stife NLPW Accuracy Form A: Plate 3 yeek zine hain boke sneap NLPW Accuracy Form A: Plate 4

PAGE 138

138 keat goam n ute tane stife NLPW Accuracy Form A: Plate 3 yeek zine hain boke sneap NLPW Accuracy Form A: Plate 4

PAGE 139

139 Pseudoword Accuracy Measure Form B 1 Scoring: This measure consists of two probes. Administer both probes. For a word read incorrectly, write the students response phonemically in the space provided and mark the word as incorrect. If the student self -corrects, write the letter S next to the response. If the student does not make an attempt to read a word in 5 seconds, say, Tell me what you think that word is. If they still do not respond, mark the word as incorrect and say, Try the next word. For each probe, discontinue the probe if the student reads no words correctly on a given plate. Training Item: Tell the student, I would like you to read some words that are not real. When you read these, its ok if these words dont make sense. Show the training card and point to mip You will read a list like this from top to bottom. Sweep your finger down the list of practice words. Read these words starting at the top. If the student is not able to read a word, or if they do not blend the sounds together, say Put them together to say the whole word. Model blending the word if necessary. Directions: Note that this is not a timed task. Tell the student, Now, I am going to ask you to read some more words. Please start at the top of the list. Be sure to read every word carefully and do your best. You may begin. Brief positive feedback may be provided (i.e., Good work.) to motivate the student. LPW Accuracy 1 !"# $ 2 %"&'$ 3 ()* $ 4 *+"# $ 5 %,-$ 6 ./0 $ 7 1/0#$ 8 2"$ 9 3+)# $ 10 4,!-$ 11 5,16 $ 12 1)1$ 13 3!)0 $ 14 78-$ 15 5,6 $ 16 9/4-$ 17 &!8$ 18 38#$ 19 9/*$ 20 784-$ mip fet glim

PAGE 140

140 Pseudoword Accuracy Measure Form B 2 NLPW Accuracy 1 !"#$ % 2 &#'(% % 3 )"*(% 4 +,*( % 5 -(#. % 6 /#*(% 7 0/,1( % 8 0"#/% 9 &#,2 % 10 *(('% 11 1.#/(% 12 34/( % 13 -"5( % 14 0*(#' % 15 5(#/ % 16 1,2(% 17 '#/(% 18 6((5% 19 7#,* % 20 -)((*% Assessment Administrator: ________________________ Student: _______________________________________ Date: _________________ Time: __________________

PAGE 141

141 mip fet glim LPW Accuracy Training Items Form B mip fet glim LPW Accuracy Training Items Form B lop zock hig grop zet LPW Accuracy Form B: Plate1 wum nump fot brip selt LPW Accuracy Form B: Plate 2

PAGE 142

142 lop zock hig grop zet LPW Accuracy Form B: Plate1 wum nump fot brip selt LPW Accuracy Form B: Plate 2 yend nin blim jat yed LPW Accuracy Form B: Plate 3 vust clat bap vug jast LPW Accuracy Form B: Plate 4

PAGE 143

143 yend nin blim jat yed LPW Accuracy Form B: Plate 3 vust clat bap vug jast LPW Accuracy Form B: Plate 4 goam vape rone zine beal NLPW Accuracy Form B: Plate 1 tane stife soat vaid neep NLPW Accuracy Form B: Plate 2

PAGE 144

144 goam vape rone zine beal NLPW Accuracy Form B: Plate 1 tane stife soat vaid neep NLPW Accuracy Form B: Plate 2 flate jute boke sneap keat NLPW Accuracy Form B: Plate 3 fide pate yeek hain breen NLPW Accuracy Form B: Plate 4

PAGE 145

145 flate jute boke sneap keat NLPW Accuracy Form B: Plate 3 fide pate yeek hain breen NLPW Accuracy Form B: Plate 4

PAGE 146

146 APPENDIX D REAL-WORD AUTOMATICITY PROTOCOLS AND STIMULUS PLATES Real-Word Automaticity Measure Form A Scoring: This is a timed word reading task The student will have 1 minute to read as many words as possible for each of two automaticity word plates. Mark any word read incorrectly with a line through it. If the student self -corrects an error, write an S next to the word. Give the student 3-5 seconds to say a word, and if they do not read the word, say, Try the next word. Circle the last word read before the timer sounds. Training Item: I would like you to read some more words for me. Read from left to across the row in this direction. [point to cat and sweep your finger from left to right under the words]. Lets practice that. [Point to cat] Start here and read across the row. [If the student does not blend the sounds together to form a whole word, ask the student to say the whole word. Model blending the word together if necessary. Repeat for each example word.] cat men trap Directions: Now, you will read some more words. Start at the top of the page and read across each row from left to right. Try to read every word and do your best. If you get stuck on a word, you can skip it and try the next word. You will only read for one minute. [Set timer to 1 minute. Place probe in front of student] You may begin now. Brief positive feedback (i.e., Good work.) may be provided if needed for motivation. LRW Automaticity rat hog jump tick got kit yum dot must bat win gust frog lab fast bug let drug swim fig flip prop bend pet hit dump plot but pack grim kick tin jug hump map dad fed lack slug rap dig tend sad skip drag drop tab grip cop last NLRW Automaticity beak hope nail cape seat feel flute goat bean shine bike pride fine plate tune dime leak prize stale slide five trail smoke home dude wipe feed rude wait robe main paid blaze keep cone goal wave choke June speed nose bake hole clean heat shake greet drive sweep float Administrator: _______________ Student: __________________________ Date: ___________

PAGE 147

147 cat men trap LRW Automaticity Training Items Form A cat men trap LRW Automaticity Training Items Form A

PAGE 148

148 rat hog jump tick got kit yum dot must bat win gust frog lab fast bug let drug swim fig flip prop bend pet hit dump plot but pack grim kick tin jug hump map dad fed lack slug rap dig tend sad skip drag drop tab grip cop last LRW Automaticity Form A

PAGE 149

149 beak hope nail cape seat feel flute goat bean shine bike pride fine plate tune dime leak prize stale slide five trail smoke home dude wipe feed rude wait robe main paid blaze keep cone goal wave choke June speed nose bake hole clean heat shake greet drive sweep float NLRW Automaticity Form A

PAGE 150

150 Real-Word Automaticity Measure Form B Scoring: This is a timed word reading task The student will have 1 minute to read as many words as possible for each of two automaticity word plates. Mark any word read incorrectly with a line through it. If the student self -corrects an error, write an S next to the word. Give the student 3-5 seconds to say a word, and if they do not read the word, say, Try the next word. Circle the last word read before the timer sounds. Training Item: I would like you to read some more words for me. Read from left to across the row in this direction. [point to cat and sweep your finger from left to right under the words]. Lets practice that. [Point to cat] Start here and read across the row. [If the student does not blend the sounds together to form a whole word, ask the student to say the whole word. Model blending the word together if necessary. Repeat for each example word.] cat men trap Directions: Now, you will read some more words. Start at the top of the page and read across each row from left to right. Try to read every word and do your best. If you get stuck on a word, you can skip it and try the next word. You will only read for one minute. [Set timer to 1 minute. Place probe in front of student] You may begin now. Brief positive feedback (i.e., Good work.) may be provided if needed for motivation. LRW Automaticity hit got bend bat cop last flip drop map let frog skip dot lab must tick yum bug swim lack tin jug plot dump fed kit kick gust fig rap fast tend hog grip pack rat dig but jump prop sad drag hump dad win pet slug grim tab drug NLRW Automaticity hole dime goat seat prize tune cape float bean leak greet cone nose speed bike keep plate feel stale rude wait robe shake sweep bake paid feed trail smoke five home heat shine blaze wipe fine nail goal flute choke wave dude main pride slide hope beak june drive clean Administrator: _______________ Student: __________________________ Date: ___________

PAGE 151

151 cat men trap LRW Automaticity Training Items Form B cat men trap LRW Automaticity Training Items Form B

PAGE 152

152 hit got bend bat cop last flip drop map let frog skip dot lab must tick yum bug swim lack tin jug plot dump fed kit kick gust fig rap fast tend hog grip pack rat dig but jump prop sad drag hump dad win pet slug grim tab drug LRW Automaticity Form B

PAGE 153

153 hole dime goat seat prize tune cape float bean leak greet cone nose speed bike keep plate feel stale rude wait robe shake sweep bake paid feed trail smoke five home heat shine blaze wipe fine nail goal flute choke wave dude main pride slide hope beak june drive clean NLRW Automaticity Form B

PAGE 154

154 APPENDIX E PSEUDOWORD AUTOMATICITY PROTOCOLS AND STIMULUS PLATES Pseudoword Automaticity Measure Form A Scoring: This is a timed word reading task The student will have 1 minute to read as many words as possible for each of two automaticity word plates. Mark any word read incorrectly with a line through it. If the student self -corrects an error, write an S next to the word. Give the student 3-5 seconds to say a word, and if they do not read the word, say, Try the next word. Circle the last word read before the timer sounds. Training Item: I would like you to read some more words for me. Read from left to across the row in this direction. [point to mip and sweep your finger from left to right under the words]. Lets practice that. [Point to mip ] Start here and read across the row. [If the student does not blend the sounds together to form a whole word, ask the student to say the whole word. Model blending the word together if necessary. Repeat for each example word.] mip fet glim Directions: Now, you will read some more words. Start at the top of the page and read across each row from left to right. Try to read every word and do your best. If you get stuck on a word, you can skip it and try the next word. You will only read for one minute. [Set timer to 1 minute. Place probe in front of student] You may begin now. Brief positive feedback (i.e., Good work.) may be provided if needed for motivation. LPW Automaticity Probe spag yop plim lat mig bot clug mab blop ned crat glot nop pend nust frop grug smip tum blag flot lin zig jelt dap jit zick rog lum rast plog jad yelt shug vap hast zump vad zot hab lut sug crip zin ped nend vack zug ret glim NLPW Automaticity Probe fane meap gleat nide boak heek cobe pake rean fute drate lote pime bape greap leam fime jeed spail veek vope baid doan rile sneep chake fune glope leat wike mobe shime tain hoat breet blait rive tude fain zoat noke flabe zail yate zake stike toam blime keet vate Administrator: _______________ Student: __________________________ Date: ___________

PAGE 155

155 mip fet glim LPW Automaticity Training Items Form A mip fet glim LPW Automaticity Training Items Form A

PAGE 156

156 spag yop plim lat mig bot clug mab blop ned crat glot nop pend nust frop grug smip tum blag flot lin zig jelt dap jit zick rog lum rast plog jad yelt shug vap hast zump vad zot hab lut sug crip zin ped nend vack zug ret glim LPW Automaticity Form A

PAGE 157

157 fane meap gleat nide boak heek cobe pake rean fute drate lote pime bape greap leam fime jeed spail veek vope baid doan rile sneep chake fune glope leat wike mobe shime tain hoat breet blait rive tude fain zoat noke flabe zail yate zake stike toam blime keet vate NLPW Automaticity Form A

PAGE 158

158 Pseudoword Automaticity Measure Form B Scoring: This is a timed word reading task The student will have 1 minute to read as many words as possible for each of two automaticity word plates. Mark any word read incorrectly with a line through it. If the student self -corrects an error, write an S next to the word. Give the student 3-5 seconds to say a word, and if they do not read the word, say, Try the next word. Circle the last word read before the timer sounds. Training Item: I would like you to read some more words for me. Read from left to across the row in this direction. [point to mip and sweep your finger from left to right under the words]. Lets practice that. [Point to mip ] Start here and read across the row. [If the student does not blend the sounds together to form a whole word, ask the student to say the whole word. Model blending the word together if necessary. Repeat for each example word.] mip fet glim Directions: Now, you will read some more words. Start at the top of the page and read across each row from left to right. Try to read every word and do your best. If you get stuck on a word, you can skip it and try the next word. You will only read for one minute. [Set timer to 1 minute. Place probe in front of student] You may begin now. Brief positive feedback (i.e., Good work.) may be provided if needed for motivation LPW Automaticity lat yop jelt mab sug zig rast dap zug ped smip nop jad nend zin mig crip shug frop vap clug hast blop lum nust zick crat zot glim jit ned zump yelt bot rog glot lut vack ret blag hab flot plim pend grug lin plog vad spag tum NLPW Automaticity bape stike jeed vate toam flabe pake vope blime meap heek rile breet leat rive chake fime cobe fute glope shime tude greap wike baid rean boak pime yate mobe spail hoat keet drate tain sneep fane veek leam lote fune noke zail gleat blait zake zoat nide fain doan Administrator: _______________ Student: __________________________ Date: ___________

PAGE 159

159 mip fet glim LPW Automaticity Training Items Form B mip fet glim LPW Automaticity Training Items Form B

PAGE 160

160 lat yop jelt mab sug zig rast dap zug ped smip nop jad nend zin mig crip shug frop vap clug hast blop lum nust zick crat zot glim jit ned zump yelt bot rog glot lut vack ret blag hab flot plim pend grug lin plog vad spag tum LPW Automaticity Form B

PAGE 161

161 bape stike jeed vate toam flabe pake vope blime meap heek rile breet leat rive chake fime cobe fute glope shime tude greap wike baid rean boak pime yate mobe spail hoat keet drate tain sneep fane veek leam lote fune noke zail gleat blait zake zoat nide fain doan NLPW Automaticity Form B

PAGE 162

162 APPENDIX F WORDS FOR TREATMENT LESSONS LESSON 1: CVC LESSON 2: CVC -at, -ap, -ip -in, -ig, -ug Word Work Elkonin Boxes Word Work Elkonin Boxes vap map mat zat sit dit dap sap gat sat vit bit fin min mig dig dug sug zin bin lig big jug nug LESSON 3: CVC LESSON 4: CVC -ad, -ed, -et -um, -ut, -ot Word Work Elkonin Boxes Word Work Elkonin Boxes mad zad zed bed bet ret cad bad led ped jet fet lum gum gut sut sot dot yum tum nut dut not mot LESSON 5: CVC LESSON 6: CCVC & CVCC -og, -op, -ab -ot, -at, -ast Word Work Elkonin Boxes Word Work Elkonin Boxes dog not nop cop cab sab pog jog dop mop zab tab spot flot flat smat past rast plot prot blat scat fast dast LESSON 7: CCVC & CVCC LESSON 8: CCVC & CVCC -ag, -og, -elt -ip, -ug, -op Word Work Elkonin Boxes Word Work Elkonin Boxes drag smag smog snog melt zelt blag flag frog spog velt felt flip frip slug flug stop glop slip glip clug plug chop frop

PAGE 163

163 LESSON 9: C CVC & CVCC LESSON 10: CCVC & CVCC -end, -ump, -ust -im, -ick, -ack Word Work Elkonin Boxes Word Work Elkonin Boxes rend send fump jump must zust zend bend bump tump nust just slim stim yick pick pack gack swim glim kick bick dack rack LESSON ll: Non -Linear LESSON 12: Non-Linear i_e, a_e o_e, u_e Word Work Elkonin Boxes Word Work Elkonin Boxes fine five rive rike rake bake bape fape bite bike fime pime fane cake cape bame hope hote note noke tube fune June lune hole cobe dobe nose dude dune gute fute LESSON 13: Non-Linear LESSON 14: Non -Linear ee, ai ea, oa Word Work Elkonin Boxes Word Work Elkonin Boxes peek feek feed zeed rain gain taid baid feel peel jeek jeed bait wait zain fain heat leat bean fean goat goal doal doak feap beat leam heap boak boam load toad LESSON 15: Non -Linear i_e, o_e, a_e, ee Word Work Elkonin Boxes slide blide choke chome flake flape greet smeet glide blime chose glope shake flape green sneep

PAGE 164

164 APPENDIX G WORDS FOR COMPARISON LESSONS LESSON 1: CVC LESSON 2: CVC -at, -ap, -ip -in, -ig, -ug Word Work Elkonin Boxes Word Work Elkonin Boxes sap map mat sat sit fit sap map mat sat sit fit fin bin big dig dug jug fin bin big dig dug jug LESSON 3: CVC LESSON 4: CVC -ad, -ed, -et -um, -ut, -ot Word Work Elkonin Boxes Word Work Elkonin Boxes mad bad bed led let jet mad bad bed led let jet yum gum gut nut not dot yum gum gut nut not dot LESSON 5: CVC LESSON 6: CCVC & CVCC -og, -op, -ab -ot, -at, -ast Word Work Elkonin Boxes Word Work Elkonin Boxes dog jog mop cop cab tab dog jog mop cop cab tab spot plot flat scat past fast spot plot flat scat past fast LESSON 7: CCVC & CVCC LESSON 8: CCVC & CVCC -ag, -og, -elt -ip, -ug, -op Word Work Elkonin Boxes Word Work Elkonin Boxes drag flag frog smog melt felt drag flag frog smog melt felt flip slip slug plug stop chop flip slip slug plug stop chop

PAGE 165

165 LESSON 9: C CVC & CVCC LESSON 10: CCVC & CVCC -end, -ump, -ust -im, -ick, -ack Word Work Elkonin Boxes Word Work Elkonin Boxes send bend bump jump just must send bend bump jump just must slim swim kick pick pack rack slim swim kick pick pack rack LESSON ll: Non -Linear LESSON 12: Non-Linear i_e, a_e o_e, u_e Word Work Elkonin Boxes Word Work Elkonin Boxes fine five bite bike bake rake cake cape fine five bite bike bake rake cake cape hope hole note nose tube dude dune June hope hole note nose tube dude dune June LESSON 13: Non-Linear LESSON 14: Non-Linear ee, ai ea, oa Word Work Elkonin Boxes Word Work Elkonin Boxes feed feel peel peek rain gain bait wait feed feel peel peek rain gain bait wait bean beat heat heap load toad goat goal bean beat heat heap load toad goat goal LESSON 15: Non -Linear i_e, o_e, a_e, ee Word Work Elkonin Boxes slide glide choke chose flake shake green greet slide glide choke chose flake shake green greet

PAGE 166

166 APPENDIX H FIDELITY INSTRUMENT FOR TREATMENT CONDITION 1 Fidelity Checklist: Treatment Condition Instructor: School: Observer: Group: Date: Lesson start time: End time: Session Length: Adherence Step 1: Word Work with Manipulative Letters Start: _________ End: _________ The instructor guided students in reading and spelling real and pseudowords using manipulative letters. Yes No Notes Instructor guided students in encoding words. Instructor guided students in decoding words. Real words and psuedowords were taught during Step 1. Instructor presented only the words selected for Step 1. Step 2: Word Work with Elkonin Boxes Start: _________ End: _________ The instructor guided students in segmenting and spelling real and pseudowords using Elkonin boxes. Yes No Notes Instructor guided students in segmenting and counting sounds in words. Instructor guided students in spelling/writing words. Real words and psuedowords were taught during Step 2. Instructor presented only the words selected for Step 2. Step 3: Word Review with Index Cards (when appropriate) The instructor provided students with opportunities to read word cards containing words selected for instruction. Yes No N/A Notes Instructor presented words selected for instruction on word cards. Instructor allowed students to attempt to read words first. When necessary, instructor assisted students to blend sounds and then read whole words. Focus of this step was on decoding.

PAGE 167

167 2 Quality Instructor demonstrated elements of high quality instruction. Yes No Notes Instructor was prepared and organized with lesson plan and materials readily available before and during lesson. Instructional pace allowed for complete presentation of selected words in approximately 15-20 minutes. Instructor modeled sounds accurately and demonstrated blendable sounds when possible. Elements of direct instruction (gradual release of responsibilitiy) were observed during lesson. Insturctor provided positive feedback. Instructor provided specific corrective feedback as needed (i.e., modeled blending and/or segmenting sounds and guided students to correct reading and/or spelling of word). Responsiveness Instructor provided opportunites for all students to participate and respond during lesson. Yes No Notes Instructor took measures to include all students during lesson. Instructor encouraged all students to respond and become engaged in lesson. Instructor made clear the expectation that all students should participate in lesson. All students were engaged in lesson. Exposure Instructor exposed students to all real and pseudoword stimuli selected for Steps 1 and 2 of lesson. Yes No Notes Instructor presented all words selected for Step 1 of lesson. Instructor presented all words selected for Step 2 of lesson. Additional Observation Notes:

PAGE 168

168 APPENDIX I FIDELITY INSTRUMENT FOR COMPARISON CONDITION 1 Fidelity Checklist: Comparison Condition Instructor: School: Observer: Group: Date: Lesson start time: End time: Session Length: Adherence Step 1: Word Work with Manipulative Letters Start: _________ End: _________ The instructor guided students in reading and spelling real words using manipulative letters. Yes No Notes Instructor guided students in encoding words. Instructor guided students in decoding words. Instructor presented only the words selected for Step 1. Step 2: Word Work with Elkonin Boxes Start: _________ End: _________ The instructor guided students in segmenting and spelling real using Elkonin boxes. Yes No Notes Instructor guided students in segmenting and counting sounds in words. Instructor guided students in spelling/writing words. Instructor presented only the words selected for Step 2. Step 3: Word Review with Index Cards (when appropriate) The instructor provided students with opportunities to read word cards containing words selected for instruction. Yes No N/A Notes Instructor presented words selected for instruction on word cards. Instructor allowed students to attempt to read words first. When necessary, instructor assisted students to blend sounds and then read whole words. Focus of this step was on decoding. Contamination Students were exposed to real words only during lesson. Yes No Notes Instructor presented real words only for instruction during this lesson.

PAGE 169

169 2 Quality Instructor demonstrated elements of high quality instruction. Yes No Notes Instructor was prepared and organized with lesson plan and materials readily available before and during lesson. Instructional pace allowed for complete presentation of sel ected words in approximately 15-20 minutes. Instructor modeled sounds accurately and demonstrated blendable sounds when possible. Elements of direct instruction (gradual release of responsibilitiy) were observed during lesson. Insturctor provided positive feedback. Instructor provided specific corrective feedback as needed (i.e., modeled blending and/or segmenting sounds and guided students to correct reading and/or spelling of word). Responsiveness Instructor provided opportunites for all students to participate and respond during lesson. Yes No Notes Instructor took measures to include all students during lesson. Instructor encouraged all students to respond and become engaged in lesson. Instructor made clear the expectation that all students should participate in lesson. All students were engaged in lesson. Exposure Instructor exposed students to all real word stimuli selected for Steps 1 and 2 of lesson. Yes No Notes Instructor presented all words sel ected for Step 1 of lesson. Instructor presented all words selected for Step 2 of lesson. Additional Observation Notes:

PAGE 170

170 LIST OF REFERENCES Adams, M. J. (1990). Beginning to read: Thinking and learning about print. Cambridge: MIT Press. Adams, M. J. (2001). Alphabetic anxiety and explicit systematic phonics instruction: A cognitive science perspective. In S. B. Neuman & D. K. Dickinson (Eds.), Handbook of early literacy research (Vol. I, pp. 66 -80). New York, NY: The Guilford Press. Adams, M. J. (2011). The relation between alphabetic basics, word recognition, and reading. In S. J. Samuels and Alan E. Farstrup (Eds.), What research has to say about reading instruction: 4th Edition (pp. 4-24). Newark, DE: International Reading Association Ball, E. W. & Blachman, B. A. (1991). Does phoneme awareness training in kindergarten make a difference in early word recognition and developmental spelling? Reading Research Quarterly 26(1), 49-66. Beck, I. L. (2006). Making sense of phonics: The hows and whys. New York, NY: Guilford Press. Berninger, V. W., Vermeulen, K., Abbott, R. D., McCutchen, D. Cotton, S., Cude, J., !Sharon, T. (2003). Comparison of three approaches to supplementary reading instruction for low-achieving second-grade readers. Language, Speech, and Hearing Services in Schools, 34, 101 -116. Biancarosa, G., & Snow, C. E. (2004). Reading next: A vision for action and research in middle and high school literacy: A report to Carneg ie Corporation of New York. Washington, DC: Alliance for Excellent Education. Blachman, B. A. (1984). Relationship of rapid naming ability and language analysis skills to kindergarten and first -grade reading achievement. Journal of Educational Psychology, 76(4), 610-622. Blachman, B. A. (1994). What we have learned from longitudinal studies of phonological processing and reading, and some unanswered questions: A response to Torgesen, Wagner, and Rashotte. Journal of Learning Disabilities, 27 287 -291. Blachman, B., Ball, E., Black, R., & Tangel, D. (1994). Kindergarten teachers develop phoneme awareness in low-income, inner-city classrooms: Does it make a difference? Reading and Writing: An Interdisciplinary Journal, 6 1-18. Blachman, B. A., Tangel, D. M., Ball, E. W., Black, R., & McGraw, C. K. (1999). Developing phonological awareness and word recognition skills: A two -year intervention with low-income, inner-city children. Reading and Writing: An Interdisciplinary Journal, 11, 239 -273.

PAGE 171

171 Blachman, B. A., Schatschneider, C., Fletcher, J. M., Francis, D. J., Clonan, S. M., Shaywitz, B. A., & Shaywitz, S. E. (2004). Effects of intensive reading remediation for second and third graders and a 1-year follow-up. Journal of Educational Psychology, 96(3), 444-461. Bradley, L., & Bryant, P.E. (1983). Categorizing sounds and learning to read: A causal connection. Nature, 30, 419 -421. Bradley, L., & Bryant, P. (1985). Rhyme and reason in reading and spelling Ann Arbor: University of Michigan Press. Byrne, B., & Fielding -Barnsley, R. (1991). Evaluation of a program to teach phonemic awareness to young children. Journal of Educational Psychology, 83 (4), 451-455. Byrne, B., & Fielding-Barnsley, R. (1993). Evaluation of a program to teach phonemic awareness to young children: A 1-year follow-up. Journal of Educational Psychology, 85(1), 104-111. Byrne, B., & Fielding-Barnsley, R. (1995). Evaluation of a program to teach phonemic awareness to young children: A 2and 3 -year follow-up and a new preschool trial. Journal of Educational Psychology, 87(3), 488-503. Campbell, M. L., Helf, S., & Cooke, N. L. (2008). Effects of adding multisensory components to a supplemental reading program on the decoding skills of treatment resisters. Education and Treatment of Children, 3 3(3), 267-295. Cardenas, J. M. (2009). Phonics instruction using pseudowords for success in phonetic decoding. (Doctoral Dissertation). FIU Electronic Theses and Dissertations (Paper 139) Chall, J. S. (1967). Learning to read: The great debate New York: McGraw-Hill. Chall, J. S. (1983). Stages of reading development. New York: McGraw-Hill. Cibelli, E. (2012). Shared early pathways of word and pseudoword processing: Evidence from high-density electrocorticography. UC Berkley Phonology Lab Annual Report Clay, M. M. (1993). Reading Recovery: A guidebook for teachers in training Portsmouth, NH: Heinemann. Cohen, J. (1969). Statistical power analysis for the behavior sciences New York, NY: Academic Press. Cohen, J. (1988). Statistical power analysis for the behavior sciences. (2nd ed.). Hillsdale, NJ: Erlbaum.

PAGE 172

172 Coltheart, M., Curtis, B., Atkins, P., & Haller, M. (1993). Models of reading aloud: Dualroute and parallel-distributed-processing approaches. Psychological Review, 100(4), 589-608. Cunningham, A. E. (1990). Explicit versus implicit instruction in phonemic awareness. Journal of Experimental Child Ps ychology, 50(3), 429-444. Cunningham, P. M., & Cunningham, J. W. (1992). Making words: Enhancing the invented spelling-decoding connection. The Reading Teacher, 46(2), 106-115. Cunningham, P. M., Hall, D. P., & Defee, M. (1991). Non-ability-grouped, multilevel instruction: A year in a first-grade classroom. The Reading Teacher, 44(8), 566571. Devault, R., & Joseph, L. M. (2004). Repeated readings combined with word boxes phonics technique increases fluency levels of high school students with severe reading delays. Preventing School Failure, 49(1), 22-27. Ehri, L. C. (1995). Phases of development in learning to read words by sight. Journal of Research in Reading, 18(2), 116-125. Ehri, L. C. (1998). Grapheme-phoneme knowledge is essential for lear ning to read words in English. In J. Metsala & L. Ehri (Eds.), Word recognition in beginning literacy (pp. 3-40). Mahwah, NJ: Lawrence Erlbaum Associates. Ehri, L. C. (1999). Phases of development in learning to read words. In J. Oakhill & R. Beard (Eds.). Reading development and the teaching of reading: A psychological perspective (pp. 79-108). Oxford, UK: Blackwell. Ehri, L. C. (2004). Teaching phonemic awareness and phonics: An explanation of the National Reading Panel Meta -Analyses. In P. McCardle & V. Chhabra (Eds.), The voice of evidence in reading (pp. 53-187). Baltimore, MD: Paul H. Brookes Publishing Co. Ehri, L. C. (2005). Learning to read words: Theories, findings, and issues. Scientific Studies of Reading, 9(2), 167-188. Ehri, L. C., Dreyer, L.G., Flugman, B., & Gross, A. (2007). Reading Rescue: An effective tutoring intervention model for language -minority students who are struggling readers in first grade. American Educational Research Journal, 44 (2), 418-448. Ehri, L. C., & McCormick, S. (1998). Phases of word learning: Implications for instruction with delayed and disabled readers. Reading and Writing Quarterly: Overcoming Learning Difficulties, 14 (2), 135-163. Ehri, L. C., Nunes, S. R., Willows, D. M., Schuster, B. V., Yaghoub-Zadeh, Z., & Shanahan, T. (2001). Phonenic awareness instruction helps children leanr to read:

PAGE 173

173 Evidence from the National Reading Panel's meta-analysis. Reading Research Quarterly, 36(3), 250-287. Ehri, L. C., & Robbins, C. (1992). Beginners need some decoding skill to read words by analogy. Reading Research Quarterly, 27(1), 13-26. Ehri, L. C., & Snowling, (2004). Developmental variation in word recognition. In C. Stone, E. Silliman, B. Ehren, & K. Apel (Eds.), Handbook of language and literacy: Development and dis orders (pp. 433-460 ). New York: Guilford Press. Ehri, L. C., & Wilce, L. S. (1983). Development of word identificat ion speed in skilled and less skilled beginning readers. Journal of Educational Psychology, 75 3-18. Ehri, L. C., & Wilce, L. S. (1985). Movement into reading: Is the first stage of printed word learning visual or phonetic? Reading Research Quarterly, 20(2), 163-179. Ehri, L., & Wilce, L. (1987a). Does learning to spell help beginners learn to read words? Reading Research Quarterly, 12, 47 -65. Ehri, L. C., & Wilce, L. S. (1987b). Cipher versus cue reading: An exper iment in decoding acquisition. Journal of Educational Psychology, 79(1), 3-13. Elbaum, B., Vaughn, S., Hughes, M., & Moody, S. W. (1999). Grouping practices and reading outcomes for students with disabilities. Exceptional Children, 65(3), 399415. Elbaum, B., Vaughn, S., Hughes, M., & Moody, S. W. (2000). How effective are one-toone tutoring programs in reading for elementary students at risk for reading failure? A meta-analysis of the intervention research. Journal of Educational Psychology, 92(4), 605-619. Elkonin, D. B. (1973). U.S.S.R. In J. Downing (Ed.), Comparative reading (pp. 551-579). New York: Macmillan Faul, F., Erdfelder, E., Lang, A., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39(2), 175-191. Florida Department of Education. (2009). Florida Assessments for Instruction in Reading (FAIR). Tallahassee, FL. Francis, D. J., Shaywitz, S. E., Stuebing, K. K., Shaywitz, B. A., & Fletcher, J. M. (1996). Developmental lag versus deficit models of reading disability: A longitudinal, individual growth curves study. Journal of Educational Psychology, 88, (1) 3-17. Gall, M. D., Gall, J. P., & Borg, W. R. (2007). Educational research: An introduction (8th ed.). New York: Pearson.

PAGE 174

174 Goldstein, H. (1983). Recombinative generalization: Relationships between environmental conditions and the linguistic repertoires of language learners. Analysis and Intervention in Developmental Disabilities, 3 279-293. Good, R. H., Baker, S. K., & Peyton, J. A. (2009). Making sense of nonsense word fluency: Determining adequate progres s in early first-grade reading. Reading & Writing Quarterly, 25(1), 33-56. Gough, P. (1983). Context, form, and interaction. In K. Rayner (Ed.), Eye movements in reading: Perceptual and language processes (pp. 203-211). New York: Academic. Gough, P. (1984). Word recognition. In R. Barr, M.L. Kamil, & P. Mosenthal (Eds.), Handbook of reading research, (Vol. I, pp. 225-253). New York, NY: Longman. Gough, P. B. (1993). The beginning of decoding. Reading and Writing: An Interdisciplinary Journal, 5, 181-192. Gough, P., & Tunmer, W. (1986). Decoding, reading, and reading disability. Remedial and Special Education, 7, 6-10. Hatcher, P. J., Hulme, C., & Snowling, M. J. (2004). Explicit phoneme training combined with phonic reading instruction helps young children at risk of reading failure. Journal of Child Psychology and Psychiatry, 45 (2), 338-358. Hayes, L. F., Lane, H. B., & Pullen, P. C. (1999). University of Florida literacy initiative: Tutoring for beginning readers. Henry, M. K. (1993) The role of decoding in reading research and instruction. Reading and Writing, 5(2), 105-112. Hoover, N. L., & Lane, H. B. (1997). Reading rescue for elementary students: A literacy intervention for struggling readers Gainesville, FL: Author. Hoover, W. A., & Gough, P. B. (1990). The simple view of reading. Reading and Writing: An Interdisciplinary Journal, 2, 127 -160. Hulme, C., Hatcher, P. J., Nation, K., Brown, A., Adams, J., & Stuart, G. (2002). Phoneme awareness is a better predictor of early reading skill than onset -rime awareness. Journal of Experimental Child Psychology, 82 (1), 2-28. Iversen, S. & Tunmer, W. (1993). Phonological processing skills and the Reading Recovery Program. Journal of Educational Psychology, 85 112 -126 Joseph, L. M. (1998). Word boxes help children with learning disabilities identify and spell words. The Reading Teacher, 52 (4), 348-356.

PAGE 175

175 Joseph, L. M. (2000a). Developing first graders phonemic awareness, word identification, and spelling: A comparison of two contemporary phonic instructional approaches. Reading Research and Instruction, 39 (2), 160-169. Joseph, L. M. (2000b). Using word boxes as a large group phonics approach in a first grade classroom. Reading Horizons, 39(2), 117-126. Joseph, L. M. (2002a). Facilitating word recognition and spelling using word boxes and word sort phonic procedures. School Psychology Review, 31 (1), 122-129. Joseph, L. M. (2002b). Helping children link sound to print: Phonics procedures for small-group or whole-class settings. Intervention in School and Clinic, 37 (4), 217221. Juel, C. (1983). The development and use of mediated word identification. Reading Research Quarterly, 18, 306 -327. Juel, C. (1988). Learning to read and write: A longitudinal study of 54 children from first through fourth grades. Journal of Educational Psychology,80, (4) 437-447. Lane, H. B. (1994). The effects of explicit instruction in contextual application of phonological awareness on the reading skills of first -grade students. Unpublished doctoral dissertation, University of Florida. Lane, H. B., & Pullen, P. C. (2004). Phonological awareness assessment and instruction: A sound beginning Boston, MA: Allyn & Bacon. Lane, H. B., Pullen, P. C., Hudson, R. F., & Konold, T. R. (2009). Identifying essential instructional components of literacy tutoring for struggling beginn ing readers. Literacy Research and Instruction, 48, 277 -297. Lenchner, O., Gerber, M. M., & Routh, D. K. (1990). Phonological awareness tasks as predictors of decoding ability: Beyond segmentation. Journal of Learning Disabilities, 23(4), 240-247. Liberman, I. Y., Shankweiler, D., Fischer, F. W., & Carter, B. (1974). Explicit syllable and phoneme segmentation in the young child. Journal of Experimental Child Psychology, 18, 201-212. Linan-Thompson, S., & Hickman-Davis, P. (2002). Supplemental readin g instruction for students at risk for reading disabilities: Improve reading 30 minutes at a time. Learning Disabilities Research & Practice, 17 (4), 242-251. Lyon, R. (2001). G. R. (2001, March 8). Measuring success: Using assessments and accountability to raise student achievement Statement of Dr. G. Reid Lyon, Chief, Child Development and Behavior Branch, National Institute of Child Health and Human Development, National Institutes of Health. Subcommittee on Education Reform, Committee on Education and t he Workforce, U.S. House of

PAGE 176

176 Representatives. Washington, DC. Retrieved from http://www.nrrf.org/lyon_statement3-01.htm Mann, V. A. (1993). Phoneme awareness and future reading ability. Journal of Learning Disabilities, 26(4), 259-269. Maslanka, P., & Joseph, L. M. (2002). A comparison of two phonological awareness techniques between samples of preschool children. Reading Psychology, 23, 271 288. McCandliss, B., Beck, I. L., Sandak, R., & Perfetti, C. (2003). Focusing attention on decoding for children with poor reading skills: Design and preliminary tests of word building intervention. Scientific Studies of Reading, 7 (1), 75-104. McCardle, P., Scarborough, H. S., & Catts, H. W. (2001). Predicting, explaining, and preventing childrens reading difficulties. Learning Disabilities Research & Practice, 16(4), 230-239. Mesmer, H. A. E., & Griffith, P. L. (2005). Everybodys selling itbut just what is explicit, systematic phonics instruction? The Reading Teacher, 59(4), 366-376. Moats, L. C. (1998). Teaching Decoding. American Educator, 22(1-2), 1-9. Nation, K. & Hulme, C. (1997). Phonemic segmentation, not onset -rime segmentation, predicts early reading and spelling skills. Reading Research Quarterly 32(2), 154167. National Reading Panel. (2000). Teaching children to read: An evidence based assessment of the scientific research literature on reading and its implications for reading instruction. National Institute of Child Health and Human Development. Washington, DC. OConnor, R. E., & Jenkins, J. R. (1995). Improving the generalization of sound symbol knowledge: Teaching spelling to kindergarten children with disabilities. The Journal of Special Education, 29(3), 255-275. Perfetti, C. A., & Bell, L. (1991). Phonemic activation during the first 40 ms of word identification: Evidence from backward masking and priming. Journal of Memory and Language, 34 (4), 473-485. Perfetti, C. A., Bell, L. C., & Delaney, S. M. (1988). Automatic (prelexical) phonetic activation in silent word reading: Evidence from back ward masking. Journal of Memory and Language, 27(1), 59-70. Perfetti, C. A., Van Dyke, J., & Hart, L. (2001). The psycholinguistics of basic literacy. Annual Review of Applied Linguistics, 21 (1), 127-149.

PAGE 177

177 Pullen, P. C. (2000). The effects of alphabetic word work with manipulative letters on the reading acquisition of struggling first -grade students (Unpublished doctoral dissertation). University of Florida, Gainesville. Pullen, P. C., Lane, H. B., Lloyd, J. W., Nowak. R., & Ryals, J. (2005). Effects of explicit instruction on decoding of struggling first grade students: A data -based case study. Education & Treatment of Children, 28(2), 63-76. Pseudowords. (n. d.). Trustees of the University of Pennsylvania. Retrieved from http://www.med.upenn.edu/bbl/downloa ds_pseudowords.shtml Rasinski, T. (1999). Making and writing words. Reading Online, September, 1999. Retrieved from http://www.readingonline.org/articles/art_index.asp?HREF=/articles /words. Rutherford, A. (2011). ANOVA and ANCOVA: A GLM approach. Hoboken, NJ: John Wiley & Sons, Inc. Ryder, J. F., Tunmer, W. E., & Greaney K. T. (2008). Explicit instruction in phonemic awareness and phonemically based decoding skills as an intervention strategy for struggling readers in whole language classrooms. Reading and Writing, 21(4), 349369. Scammacca, N., Vaughn, S., Roberts, G., Wanzek, J., & Torgesen, J. K. (2007). Extensive reading interventions in grades k -3: From research to practice. Portsmouth, NH: RMC Research Corporation, Center on Instruction. Schmidgall, M., & Joseph, L. M. (2007). Comparison of phonic analysis and whole word-reading on first graders cumulative words read and cumulative reading rate: An extension in examining instructional effectiveness and efficiency. Psychology in the Schools, 44(4), 319-332. Seidenberg, M. S., & McClelland, J. L. (1989). A distributed developmental model of word recognition and naming. Psychological Review, 96, 523-568. Share, D. L., Jorm, A. F., Maclean, R., & Matthews, R. (1984). Sources of individual differences in reading acquisition. Journal of Educational Psychology, 86 (6), 13091324. Schumaker, J., & Sherman, J. A. (1970). Training generative verb usage by imitation and reinforcement procedures. Journal of Applied Behavior Analysis, 3, 273 -287. Shankweiler, D., Lunquist, E., Katz, L., Stuebing, K. K., Fletcher, J. M., Brady, S., Shaywitz, B. A. (1999). Comprehension and decoding: Patterns of association in children with reading difficulties. Scientific Studies of Reading, 3(1), 69-94. Simmons, D. C., Kameenui, E. J., Harn, B., Coyne, M. D., Stoolmiller, M., Santoro, L. E., Kaufman, N. K. (2007). Attributes of effective and efficient kindergarten

PAGE 178

178 reading intervention: An examination of instructional time and design specif icity. Journal of Learning Disabilities, 40 (4), 331-347. Snow, C. E., Burns, S. M., & Griffin, P. (1998). Preventing reading difficulties in young Children. Washington, DC: National Academy Press. Snowling, M. J. (1981). Phonemic deficits in developmental dyslexia. Psychological Research, 43(2), 219-234. Stanovich, K. E., (1986) Matthew effects in reading: Some consequences of the individual differences in the acquisition of reading. Reading Research Quarterly, 21, 360 -407 Stanovich, K. (1991). Word recognition: Changing perspectives. In R. Barr, M. Kamil, P. Mosenthal, & P. Pearson (Eds.), Handbook of reading research (Vol. II, pp. 383417). White Plains, NY: Longman Stein, M., Johnson, B., & Gutlohn, L. (1999). Analyzing beginning reading programs: The relationship between decoding instruction and text. Remedial and Special Education, 20(5), 275-287. Suchowierska, M. (2006). Recombinative generalization: Some theoretical and practical remarks. International Journal of Psychology, 41 (6), 514-522. Templeton, S., & Bear, D. (Eds.). (1993). Development of orthographic knowledge: The foundation of literacy. Hillsdale, NJ: Erlbaum. Torgesen, J. K. (1999). Phonologically based reading disabilities: T oward a coherent theory of one kind of learning disability. In R. J. Sternberg & L. Spear -Swerling (Eds.), Perspectives on learning disabilities (pp. 231-262). New Haven: Westview Press. Torgesen, J., Wagner, R., & Rashotte, C. (2012). Test of word reading efficiency, second edition (TOWRE-2). Austin, TX: Pro-Ed. Treiman, R. (1993). Beginning to spell. New York, NY: Oxford Press. Troia, G. A. (1999). Phonological awareness intervention research: A critical review of the experimental methodology. Reading R esearch Quarterly, 34(1), 28-52. Tunmer, W.E., & Hoover, W.A. (1993). Phonological recoding skill and beginning reading. Reading and Writing: An Interdisciplinary Journal, 5, 161-179. Uhry, J. K., & Shepherd, M. J. (1993). Segmentation/spelling instruction as part of a first-grade reading program: Effects on several measures of reading. Reading Research Quarterly, 28(3), 219-233.

PAGE 179

179 Vadasy, P. F., Sanders, E. A., & Peyton, J. A. (2005). Relative effectiveness of reading practice or word-level instruction in supplemental tutoring: How text matters. Journal of Learning Disabilities, 38 (4), 364-380. Vadasy, P. F., Sanders, E. A., & Tudor, S. (2007). Effectiveness of paraeducatorsupplemented individual in struction: Beyond basic decoding skills. Journal of Learning Disabilities, 40(6), 508-525. Vaughn, S., Linan-Thompson, S., Kouzekanani, K., Bryant, D. P., Dickson, S., & Blozis, S. A. (2003). Reading instruction grouping for students with reading difficult ies. Remedial and Special Education, 24 (5), 301-315. Williams, J. P. (1984). Phonemic analysis and how it relates to reading. Journal of Learning Disabilities, 17(4), 240-245. Woodcock, R. W. (1998). Woodcock reading mastery t ests Revised/Normative Update (WRMT-R/NU) Circle Pines, MN: American Guidance Service. Ziegler, J. C., & Jacobs, A. M. (1995). Phonological information provides early sources of constraint in the processing of letter strings. Journal of Memory and Language, 34, 567 -593.

PAGE 180

180 BIOGRAPHICAL SKETCH Keri Marie Madsen received a Bachelor of Science in Communication Sciences and Disorders in 1997 from East Carolina University and a Master of Science in Speech-Language Pathology in 2002 from the University of South Florida. While att ending graduate school, Keri worked for one year as a speech -language pathologist in an elementary school in Hillsborough County, Florida. After graduating with her masters degree, Keri resumed her work as a school -based speech-language pathologist in Alachua County, Florida, for six years. During her time in Alachua County, Keri served many students with reading difficulties. As a result, she became interested in literacy development and began pursuing an Education Specialist degree in Curriculum and Instruction at the University of Florida. Her studies focused on differentiated instruction, inclusion, and language and literacy development, assessment, and instruction. In 2004, Keri began teaching reading to elementary students struggling with litera cy skills, and she later assumed the role of reading coach for elementary school teachers In 2007, Keri received her specialist degree. While completing her doctoral studies in Special Education at the University of Florida, Keri served as a graduate teaching assistant and lead instructor for several face-to-face and web-based language and literacy courses in the College of Education and supervised student teachers during their practicum and internship placements. Keri also assumed various roles on grant -funded research projects including a leadership role as Professional Development Coordinator on a large project funded by the Insti tute of Education Sciences. She has presented her work in the areas of literacy instruction,

PAGE 181

181 professional development, and teacher knowledge about reading at local, state, and national conferences. In the future, Keri plans to pursue teaching and research in the areas of language and literacy development as well as in the professional preparation and development of teachers and/or speech-language pathologists She hopes to pursue these goals during a career in higher education.