A longitudinal test of the lag theory of developmental dyslexia

MISSING IMAGE

Material Information

Title:
A longitudinal test of the lag theory of developmental dyslexia
Physical Description:
x, 81 leaves : ill. ; 28 cm.
Language:
English
Creator:
Fennell, Eileen Brennan, 1942-
Publication Date:

Subjects

Subjects / Keywords:
Reading disability   ( lcsh )
Dyslexia   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Includes bibliographical references (leaves 72-79).
Statement of Responsibility:
by Eileen Brennan Fennell.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000086057
notis - AAK1412
oclc - 05356233
System ID:
AA00003905:00001

Full Text













A LONGITUDINAL TEST OF THE LAG THEORY
OF DEVELOPMENTAL DYSLEXIA











By

EILEEN BRENNAN FENNELL


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






UNIVERSITY OF FLORIDA


1978













ACKNOWLEDGEMENTS

The opportunity to think, to learn and to grow has

been immeasureably enhanced by the members of my committee:

Dr. Hugh Davis, Dr. Kenneth Heilman, Dr. W. Keith Berg

and most recently, Dr. Randy Carter. There is also a

special sense of gratitude to Dr. Jacquelin Goldman, Co-

Chair, whose support and encouragement were always there

whether or not I could ask. And for Dr. Paul Satz, Chair-

man, there is a complex mixture of feelings which grow

out of our long relationship. He hired me, fresh out of

undergraduate school, and taught me "to do" neuropsychology.

Many years later when I was finally ready to pursue a

graduate degree, he offered his support again. Now he is

there to complete the cycle .how appropriate. In

between, I grew and he grew, and growing up was sometimes

painful. Paul's decency and caring eased the pain and

fostered the growth for this I am grateful.

Others helped, too. The critical thinking and moral

support of Dawn Bowers and Tom Van den Abell was tremen-

dously helpful. Margi Tintner did so much more than type

the manuscript and help navigate the mysteries of NERDC.

Molly Harrower's optimism was still another resource.

Audrey Schumacher's wisdom let me see the choices I had,

and own them.







And finally, to Bob, Shannon and Chris, who didn't

complain too loudly about dinners at Burger Doodle, this

work is dedicated.


iii
















TABLE OF CONTENTS


Page


ACKNOWLEDGEMENTS . ....

LIST OF TABLES .. . ...

LIST OF FIGURES .

ABSTRACT . . .

INTRODUCTION . .

METHOD . . .

RESULTS . .

DISCUSSION . .

APPENDIX A: DESCRIPTION OF TEST BATTERY .

REFERENCES . .

BIOGRAPHICAL SKETCH .


S. v



. viii




. 19
. 19

. 23

. 55

. 69

. 72

. 80

















LIST OF TABLES

Table

1 Mean scores on Finger Localization Test
for the dyslexics and normal controls
over time . . .

2 Pair-wise post hoc comparisons for each of
the reading groups on Factor I measures .

3 Mean raw scores and per cent correct on
Recognition Discrimination Test for the
dyslexics and normal controls over time .

4 Mean raw scores and per cent correct on
Embedded Figures Test for the dyslexics
and normal controls over time .

5 Mean age equivalent scores on the Beery Test
of Visual Motor Integration for the dyslexics
and normal controls over time .

6 Mean raw scores on Similarities Test for the
dyslexics and normal controls over time .

7 Pair-wise post hoc comparisons for each of
the reading groups on Factor II measures .

8 Mean mental age equivalent scores on Peabody
Picture Vocabulary Test for the dyslexics
and normal controls over time .


9 Mean number of words recalled on Verbal
Fluency Test for the dyslexics and
normal controls over time .

10 Mean number of letters recited on Alphabet
Recitation for the dyslexics and normal
controls over time . .


. 46


. 48


11 F ratio and Significance levels of tasks which
fitted a linear or quadratic model .


Page


. 33


. 37


. 39


. 43


. 51









Table


Page


12 Mean slopes by groups of dependent measures
and Duncan's post hoc pair-wise
comparisons . . .... 53














LIST OF FIGURES

Figure Page

1. Mean raw scores on Finger Localization
Test by groups over time . 26

2. Mean raw scores on Recognition Discrimination
Test by groups over time . ... 30

3. Mean percentage correct on Recognition
Discrimination Test by groups over time 31

4. Mean raw scores on Embedded Figures Test
by groups over time . 34

5. Mean percentage correct on Embedded Figures
Test by groups over time . 35

6. Mean Beery VMI age equivalent scores by
groups over time . ... 38

7. Mean raw scores on Similarities Test by
groups over time . . 40

8. Mean mental age equivalent scores on the
Peabody Picture Vocabulary Test by groups
over time . . 44

9. Mean number of words on Verbal Fluency Test
by groups over time . 47

10. Mean number of letters recalled on Alphabet
Recitation Test by groups over time .. 49


vii














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



A LONGITUDINAL TEST OF THE LAG THEORY
OF DEVELOPMENTAL DYSLEXIA

By

Eileen Brennan Fennell

August 1978

Chairman: Paul Satz
Major Department: Psychology

A current theory hypothesizes that the disorder of

specific developmental dyslexia is a consequence of imma-

turity of the central nervous system among these children.

As a result of this immaturity, development is postulated

to yield different age-related effects. Younger dyslexic

children would show greater deficits in earlier develop-

ing cognitive-perceptual skills. Over time, as their

neural system presumably matures, they would begin to

"catch up" to normal readers on these tasks. In contrast,

later developing verbal cognitive skills would produce

only minimal performance deficits in younger dyslexics.

Over time as these skills are delayed in their acquisition,

older dyslexics would fall further behind normal readers,

presumably because the foundations for later developing

language skills are delayed.


viii








The present study compared the performance of a select

group of severe dyslexics (N=35) and three normal control

groups (N=103) on two sets of neuropsychological measures.

Factor I tasks involved four tests taken from the Satz

Predictive Battery and generally reflected sensorimotor

and cognitive perceptual skills. Factor II tasks, also

from the Satz Predictive Battery, measured performance on

four interrelated language conceptual functions. Measure-

ments were taken in Kindergarten (Time 1), Grade 2 (Time 2)

and Grade 5 (Time 3).

Significant main effects for Reading Group and for

Time were obtained for both the Factor I and Factor II

tasks. Significant Group x Time interactions were also

observed. With the exception of two tasks which had ceil-

ing effects by Time 2 (Finger Localization and Alphabet

Recitation), the performance of the dyslexics did not

approach that of the normal readers for either Factor I

or Factor II tests.

Failure to find that dyslexics eventually caught up

with the normal readers on the sensory-perceptual tasks

was interpreted to cast doubt on the lag theory of develop-

mental dyslexia. Similar findings for the conceptual-

language tasks were interpreted as consistent with either a

deficit or a lag model of dyslexia. Finally, no support

was found for a unitary deficit explanation of develop-

mental dyslexia.








Future research involving other approaches to model

testing and examination of the cognitive strategies em-

ployed by dyslexics and normal readers was discussed.













INTRODUCTION

Since the mid 1960's, dyslexia or disordered reading

has been the focus of a burgeoning body of research litera-

ture in neuropsychology, developmental psychology, pediatric

neurology and education. In 1968, the World Federation of

Neurology proposed two general definitions of this disorder:

(1) specific developmental dyslexia a disorder manifested

by difficulty in learning to read despite conventional in-

struction, adequate intelligence and sociocultural opportu-

nity; and (2) dyslexia a disorder in children who, despite

conventional classroom experience, fail to attain the lan-

guage skills of reading, writing and spelling commensurate

with their intellectual abilities (Critchley, 1970).

The first definition encompasses those instances of

reading disability thought to be constitutional in origin

(Critchly, 1970; Benton, 1975; Witelson, 1977), while the

second subsumes a broader category of reading backwardness

regardless of intelligence (Rutter & Yule, 1975; Gibson &

Levine, 1975).

Research in reading disorders has a long history of

conflicting opinions as to etiology and multiple conceptual

frameworks (Park & Linden, 1968; Klasen, 1972; Critchley,

1975). Historically, reading retardation has been seen as

a psychosocial deviation (Natchez, 1968), an educational







problem (Bannatyne, 1971), a neurological syndrome (Money,

1966) or as an interaction of all three factors (Anderson,

1970). Methodological weaknesses in empirical research

(Applebee, 1971; Benton, 1975), the absence of a theoretical

orientation (Satz, 1977) and problems even in defining the

subject of study (Ross, 1976) have still further obscured

the presumptive goals of this research. The goals of this

research have been defined as: (1) early detection and

prediction of children destined to show difficulty in learn-

ing to read; (2) determination of etiology in specific cases;

and, (3) development of effective individualized intervention

and remediation techniques (Applebee, 1971; Thompson, 1973;

Torgeson, 1975; Schain, 1977).

The present study is concerned with the second of these

goals. Specifically, it focuses on one neurogenic model of

specific dyslexia the developmental lag theory proposed by

Satz and Sparrow (1970).

Psychological Correlates of Reading Disabilities

Several recent reviews have described a number of per-

ceptual, motor, cognitive and neurological correlates of

developmental disabilities in reading (Benton, 1975; Halla-

han, 1975; Torgeson, 1975; Evans, 1977; Rosenthal, 1977;

Satz, 1977; Schain, 1977; Witelson, 1977). The most fre-

quently reported problems include: (1) delayed early lan-

guage development; (2) family history of reading problems;

(3) visuoperceptual disorders such as poor form discrimina-

tion; (4) impaired directional sense, especially right-left







orientation; (5) deficient intersensorv integration, espe-

cially auditory-visual integration; (6) impaired oral lan-

guage skills such as articulation problems, deficient vo-

cabulary and poor verbal memory; (7) impaired speech-sound

discrimination; (8) deficiencies in arithmetic skills and

in other cognitive-symbolic functions; (9) mixed or deviant

hand-eye preference; (10) deficiencies in short-term memory;

(11) deficiencies in information processing; and, (12)

disturbances in CNS function as reflected in EEG anomalies,

soft neurological signs, motor incoordination and abnor-

malities in auditory and visual evoked potentials.

As Ross (1976) notes, though correlational studies in

no way imply causality, the persistent finding of a number

of deficiencies in performance on tasks other than reading

itself has lent credence to the notion of a neurogenic origin

for at least some cases of developmental dyslexia. Unfortun-

ately, the specific etiology or etiologies remain unclear

(Benton, 1975; Rosenthal, 1977).

Neurological Theories of Dyslexia

Two broad classes of theories of the neurological ab-

normality underlying developmental dyslexia have been pro-

posed. The first suggests that developmental dyslexia is

the result of either focal maldevelonment of the brain or

of defective organization of cerebral function (Benton, 1975;

Kinsbourne, 1975; Rourke, 1976). The second asserts that

developmental dyslexia is the result of immaturity of the

cerebral cortex (Bender, 1958; Money, 1966; Satz & Sparrow,







1970). The first model can be viewed as a deficit model

since primary support is sought in demonstrating perform-

ance deficits among dyslexics. The second model is gener-

ally called a developmental lag model since primary support

is found in demonstrating age-related differences between

dyslexics and normals (Satz, Rardin & Ross, 1971; Satz &

Van Nostrand, 1973; Darby, 1974; Rourke, 1976; Usprich, 1976).

Deficit Model

The earliest explanations for childhood dyslexia were

founded on adult models of acquired word blindness (Critch-

ley, 1970; Benton, 1975; Satz, 1976; Satz, 1977; Witelson,

1977). Benton (1975) describes these models as "parietal"

models because of the assumption of some focal abnormality

in the posterior regions of the cortex (left angular gyrus,

left inferior parietal lobule and later, biparietal lobe

disease). The parietal association areas are felt to be

implicated in cases of developmental dyslexia because of

the critical role these regions play in intersensory asso-

ciations such as required in reading (Benson & Geshwind,

1969; Gibson & Levine, 1975).

Orton (1928) rejected the idea that developmental

reading disorders in children were the product of a specific

lesion analogous to adult acquired dyslexia. Instead, he

proposed that the obstacle to reading acquisition in chil-

dren was a ". failure to establish the physiological

habit of working exclusively from the engrams of one hemi-

sphere"; that is, the failure to establish unilateral cerebral







dominance for the symbolic functions of language. He chose

the label "Strephosymbolia" (twisted symbols) to delineate

this disorder from acquired reading difficulties. Thus,

he shifted emphasis from a focal or regional disorder to a

disorder at the level of cortical organization and function-

al integration. His proposal, however, found little empiri-

cal support in the subsequent literature (Ross, 1976;

Springer & Eisenson, 1977).

Zangwill (1960, 1962) proposed yet another model of

cerebral disorganization for specific dyslexics. Noting

that there was a subgroup of dyslexics who showed mixed or

left handedness and a history of impaired motor skills,

speech development and perceptual deficiencies, he suggested

that these children were more vulnerable to learning dis-

ability. Poor lateral dominance in association with other

factors such as environmental or emotional stress led to

a learning disorder.

Both Orton and Zangwill's positions purport to explain

the etiology of reading disability yet neither provide for a

specific mechanism or provide predictions as to when var-

ious deficiencies should be manifest. Further, like all

neurogenic theories, they rely upon unobservable events

(e.g., cerebral dominance) inferred from behavioral signs

with few age-specific norms and in the absence of demon-

strable CNS pathology. Recent experimental work has pro-

duced a number of "deficit" explanations for dyslexia which

focus on performance rather than upon cortical mechanisms.







Proposals include an information-processing deficit (Saba-

tino, 1968; Sabatino & Ysseldyke, 1972), a verbal-linguistic

deficit (Vellutino, Steger & Kandel, 1972; Vellutino, Smith,

Steger & Kaman, 1975; Vellutino, 1977; Vellutino, Bentley

& Phillips, 1978), perceptual-processing deficiencies (Bryan

& Bryan, 1974; Doehring, 1976; Witelson, 1977) and atten-

tional-arousal defects (Douglas, 1976; Sheer, 1976; Ross,

1976; Milberg & Whitman, 1978).

Developmental Lag Model

Critchley (1970) reports that as early as the mid-1920's

developmental dyslexia began to be conceptualized as a

functional rather than anatomical developmental delay. By

the 1960's, the notion of dyslexia as a maturational lag

or developmental delay had a number of proponents (Bender,

1958; Money, 1962; Rabinovitch & Ingram, 1968; de Hirsh,

1968). As Bender described it: maturationall lag signi-

fies a slow differentiation" in the maturation of functional

brain areas and "did not indicate a structural defect,

deficiency or loss" (1958, p. 160). By 1969, Money asserted

that "the basic common denominator in all cases of

developmental dyslexia is a maturational lag with respect

to one or more of the component variables integral to the

functional process of reading successfully" (p. 377). How-

ever, as Usprich (1976) notes, most of the support for the

delay model was found in discussion sections of papers as

a posteriori explanations for age-related differences in

performance by dyslexics.







In 1970, Satz and Sparrow presented a more formal model

of the etiology of specific developmental dyslexia from

which testable hypotheses could be derived. In brief, the

theory states that: "reading disabilities reflect a lag

in the maturation of the brain which delays differentially

those skills which are in primary ascendancy at different

chronological ages. Consequently, those skills which dur-

ing childhood develop ontogenetically earlier (e.g., visuo-

perceptual and cross-modal sensory integration) are more

likely to be delayed in younger children who are matura-

tionally immature. Conversely, those skills which during

childhood have a later or slower rate of development (e.g.,

language and formal operations) are more likely to be

delayed in older children who are maturationally immature"

(Satz, Taylor, Friel & Fletcher, 1977, p. 10).

Differences between the Deficit and Delay Models

Kinsbourne (1975) suggests that the deficit model is a

traditional medical model of organic etiology and empha-

sizes a limited malfunction of the brain. In contrast, the

delay model, although also an organic model, places more em-

phasis on the rate of acquisition of abilities relative to

chronological age. Thus, the child is seen as selectively

immature or unready in certain functional areas. A child

with developmental delay differs not in the qualitative man-

ner but in the age at which certain skills are acquired and

therefore looking like a normal younger child.







Rourke (1976), Ross (1976), Usprich (1976) and Witel-

son (1977) have criticized the delay concept primarily be-

cause of the unsolved question of at what point does a

persistent delay become a functional deficit. Only a lim-

ited number of longitudinal studies have followed children

early identified as at risk for reading failure (de Hirsh,

Jansky & Langford, 1966; Satz, Friel & Rudegaier, 1976).

To date, primary emphasis has been on the predictive effi-

ciency of precursors to later developing reading problems

(Satz, 1977). Further, both the delay model and the

deficit models assume implicitly relatively static models

of the brain (Kinsbourne, 1975; Usprich, 1976). As Usprich

(1976) states, "what is happening in the brain while this

selective and ontogenetically varying retarding is going

on?" (p. 42-43). Recent evidence (Epstein, 1974a, 1974b)

suggests that between the ages of three to 14, there are

spurts in brain growth which are correlated with spurts in

intellectual function. Thus, the course of brain matura-

tion may not be that of an age-related monotonic increase

in size and functional complexity. Differences in rates

of physical maturation have also been related to perfor-

mance on verbal and perceptual tasks (Weber, 1978). Finally

Gottlieb (1978) suggests that a variable developmental

course may increase or decrease the facilitative potency of

experience upon later-life behavioral competency.







Research on the Developmental Lag Theory

To date, those studies which have addressed the devel-

opmental lag hypothesis have been indirect tests of the

theory. The series of Florida Longitudinal Studies (Satz

et al., 1977) have primarily focused on the predictive

validity of a group of language and perceptual-motor tasks.

In general, Satz et al. have shown that visuo-perceptual

tasks given in the early school years (Kindergarten and

Grade 1) are powerful predictors of later reading achieve-

ment while language tasks predict reading performance only

for older children. These findings are in agreement with

earlier factor analytic studies of reading retardates

(Lyle, 1969; Chissom, 1971) and provide indirect support

for the Satz-Sparrow model.

Dykman, Ackerman, Clements and Peters (1971) compared

the performance of carefully selected reading disabled

children and matched controls in two age groups (below age

ten and above ten) on measures of conditionability, im-

pulsivity and attention. The authors interpreted the over-

all poorer performance, slower motor reaction times, mini-

mal heart rate changes and lower amplitude contingent

negative variation (correlates of attention) to reflect

neurologically based attentional deficiencies in the learn-

ing disabled children. They hypothesized that these re-

sults supported the viewpoint that the "main cause of

learning disabilities is a maturational lag" with the best

evidence of this lag being the number of "soft signs"







shown by these children on a special neurological exam.

This study was one of several reviewed by Ross (1976) in

developing his proposition that reading disabled children

suffered from a specific developmental lag in selective

attention. This view is also shared by Douglas (1976) and

Sheer (1976).

Symmes and Rapoport (1972), dealing with the issue of

sample source (clinic vs. school) raised by studies like

Dykman et al. (1971), selected children from urban public

and private schools whose problems in reading had "no ob-

vious reason." Children were studied who had at least one

standard deviation between WISC Full Scale IQ and the Wide

Range Achievement Test Reading Score (WRAT). They found no

evidence consistent with immature or delayed neurological

development for the total group (ages 7-13) and for three

age subgroups (7-9, 9-11, 11-13). However, this finding

is not striking since children were eliminated from the

study if there were abnormalities found on EEG and neuro-

logical exams and if there was evidence of high-risk for

neurological disorders. Thus, by selection criteria, a

group of dyslexic children of above average intelligence

was created (FSIQ = 113) whose failure to learn to read

was both unexpected and unexplainable. Yet the authors pro-

posed that deviations noted on the WISC subscales reflected

an "information processing deficiency" rather than develop-

mental disorder. Further, their analysis of WISC subtest

scatter ignored the often contradictory evidence gleaned








from numerous WISC studies of learning disables children

(Huelsman, 1970; Torgeson, 1975; Hallahan, 1975; Benton, 1975).

In a preliminary test of the Satz-Sparrow theory,

Satz, Rardin and Ross (1971) compared the performance of

matched dyslexic and normal reader controls at ages 8 and

11. Non-language dependent measures were the Bender-Gestalt

Test, Recognition-Discrimination, and an auditory-visual

integration task. Verbal measures were the WISC Verbal IQ,

a verbal fluency test and a dichotic listening task. Age-

related differences between dyslexic and control children

were found on the Bender-Gestalt test and on the auditory-

visual task, with significantly lower scores only for the

younger-aged dyslexics. Poorer verbal performance was

significant only for the older dyslexics. Thus, the pre-

diction of age-related differences in performance on lan-

guage and perceptual measures was supported. Weaknesses

of the study were the small sample size at each age (N=10)

and the reading criteria for poor readers (six months

behind grade level).

Satz and Van Nostrand (1973) undertook a larger and

more direct test of the Satz-Sparrow theory and employed a

larger sample size, controlling for the effects of IQ and

SES. A total of 40 dyslexic and 40 matched normal readers

were studied at each of two age levels (7-8 and 11-12).

Two major composite analyses of covariance were run. The

first, based on "earlier" developing skills consisted of

eight different perceptual and perceptual-motor tasks with







VIQ and SES as the covariates. Significant main effects

for age and group were found but there was no significant

age x group interaction. The second analysis of "later

developing" skills included three different measures of

language performance with WISC PIQ and SES as the covari-

ates. Significant main effects for age and group as well

as a significant age x group interaction were found. Thus

minimal support for the "earlier skills" hypothesis was

found while substantial support occurred for the "later

skills" hypothesis. Again, however, this study employed

different subjects at the two age levels.

Mattis, French and Rapin (1975) described three dif-

ferent subtypes of dyslexia observed in a large number of

reading disabled subjects: language disordered, articula-

tory-graphomotor dyscoordination and visuo-spatial percep-

tual disorder. They speculated that those children who

demonstrated the visuo-perceptual disorder syndrome may

manifest "uneven but continuing maturation of the percep-

tual processes rather than a fixed deficit" (p. 158) which

would be consistent with a maturational lag hypothesis.

More recently, Denckla and Rudel (1976) and Denckla

(1977) report the persistence of below-age signs in dyslexic

as opposed to other learning disabled children. Denckla

(1977) specifically dismisses Critchley's (1970) proposal

that soft neurological signs among dyslexic children are

merely epiphenomena. However, Erickson (1977) compared sus-

pected neurologically impaired and known brain damaged







dyslexic children and found that none of the neurological

signs (hard or soft) was significantly correlated with

chronological age. She concluded that the typical tests

for MBD could not discriminate between reading disabled

children and normal controls, arguing against the view that

dyslexic children are neurologically impaired. The neuro-

logical tests were more closely related to IQ than to

reading level. A major weakness of this study was the

selection of two criterion measures of questionable relia-

bility (Slosson IQ and Slosson Oral Reading Test). In

addition, since the sample consisted of only second grade

children (age range 82-105 months), it did not allow for

analysis of age-related effects in the neurological exam

(Dykman et al., 1971).

Muehl and DiNello (1976) administered the Harrison

Stroud Reading Readiness Profile and WISC to a random sample

of normal first grade males enrolled in a mid-western school

system. Reading achievement scores were obtained for the

six subsequent years for 56 of the subjects. Multiple re-

gression analyses were utilized to determine the relation-

ship between various combinations of the 19 predictor vari-

ables (Harrison Stroud Readiness Subtests and WISC subtests)

to reading achievement in Grades 1 through 7. Performance

in Grades 1 through 3 were found to be related to tasks

involving visual discrimination and the ability to make

ideo-graphic symbolic associations. Later reading perfor-

mance was more related to reasoning and problem solving







ability. Interestingly, letter naming was found to be a

significant predictor to reading achievement at all grade

levels. They concluded that reading deficiencies reflected

not a unitary deficit but difficulty in a number of intel-

lectual processes which mature at different rates.

Four very recent studies provide at least partial sup-

port for the developmental lag theory of reading disabili-

ties.

Sobotka, Black, Hill and Porter (1977) compared the

test performance of 24 dyslexic boys and 24 normal readers

at four different age levels (7, 9, 11, 13). All children

were prescreened to rule out neurological or sensory defects.

Subjects were matched on their Peabody Picture Vocabulary

Test (PPVT) scores with minimal criteria being a PPVT IQ of

90 or above. Nonverbal dependent measures were the WISC

Performance IQ, the Bender-Gestalt test (Koppitz scoring)

and an audio-visual integration task. Verbal dependent

measures were the WISC IQ, a verbal fluency test and two

dichotic listening tests. Normal readers obtained signifi-

cantly higher scores than the dyslexics on all nonverbal

tests combined. Significant age x group interactions were

found for the WISC PIQ and the Bender-Gestalt test in the

seven year old group and in the WISC PIQ only for the nine

year old group but by age 11 there were no significant group

differences. Auditory-visual integration was significantly

impaired among dyslexics at all age levels. There were no

significant age x group interactions on the verbal measures,







although normal readers scored significantly better on all

verbal tasks. This latter finding to some extent may have

been a function of selecting and matching the two groups

on a verbal measure (PPVT).

Tarver, Hallahan, Cohen and Kauffman (1977) tested

Ross's (1976) hypothesis of a developmental lag in selective

attention utilizing Hagen's Central-Incidental Task. Cen-

tral and incidental recall scores were compared between

disabled and normal readers at ages 8, 10, 13 and 15. They

found that constant age-related increases occurred in cen-

tral recall for both the normal and disabled readers. Inci-

dental recall, however, did not decline among the disabled

readers until age 15. This finding was felt to support the

proposal that learning disabled boys demonstrated develop-

mental lags in selective attention and verbal rehearsal

strategies. This lag could produce delayed development of

higher cognitive processes such as reading which depend

upon efficient operation of these more basic strategies.

Fletcher and Satz (1978a) examined developmental

changes in the longitudinal performance patterns of reading

disabled and normal children followed for six years. The

focus of this study was on the predictive and concurrent

validity of a series of visuo-perceptual and language re-

lated measures (e.g., Beery Test of Visuo-Motor Integration,

Embedded Figures, WISC similarities, verbal fluency). Meas-

ures of sensori-motor perceptual skills were found to be

significantly correlated with reading achievement for reading







disabled subjects and normals at grades K and 2. On the

other hand, measures of verbal-conceptual functions were

more highly correlated with reading achievement at the end

of Grade 5. Factor analyses yielded two independent age-

related factors: Factor I labelled a sensorimotor con-

ceptual factor consisted of four tasks: the Beery VMI,

Recognition-Discrimination, Embedded Figures and Finger Lo-

calization; Factor II labelled a verbal-conceptual factor -

consisted of the similarities subtest of the WISC, Peabody

Picture Vocabulary IQ, and verbal fluency. Results were

interpreted as providing indirect support for the Satz-Spar-

row (1970) developmental lag hypothesis.

Finally, Strete, Stamm, Kreder and Lovich (1978) em-

ployed a simultaneous matching-to-sample (SMS) and a delayed

matching-to-sample (DMS) task with learning disabled readers

(X age = 8.7 years) and a control group of normal readers

(age range 5 to 8 years). Normal readers demonstrated sig-

nificant age-related reductions in errors on both tasks.

Responses of the disabled learners, however, were signifi-

cantly poorer than the 8 year old controls. The authors

interpreted their poor performance to reflect to develop-

mental lags of 24 months on the SMS task and 31 months on

the DMS task. Scores on these tasks were not correlated

with intelligence level. Although this study provides some

support for a developmental lag hypothesis, its effectiveness

is weakened by the inclusion of 10 children on psychotropic

medication among the learning disabled group (six on Ritalin,







three on Mellaril and one on Dilantin and Mysoline). Thus,

almost one-third of the learning disabled group fit equally

well into a minimal brain dysfunction group. A further

difficulty involves the use of only younger-aged disabled

readers which did not allow for examination of the per-

sistence of these lags in older ages (i.e., beyond age 8).

Conclusions and Hypotheses

As the preceding review suggests, there have been a

number of theoretical explanations for developmental dyslexia.

There have been an even larger number of empirical studies

which demonstrate differences between dyslexics and normal

readers in various cognitive and perceptual functions. Few

studies, however, have attempted an empirical test within

the framework of a specific theoretical model. Where there

is empirical support for the developmental lag hypothesis

of dyslexia, methodological weaknesses lessen the impact.

Specifically, those studies addressed to the question have

employed either small sample sizes, narrow or restricted age

levels, a cross-sectional rather than longitudinal approach

or loosely defined dyslexic or normal control groups. All

too frequently, lags have been discovered through post hoc

interpretations of differences between dyslexics and nor-

mals. Those longitudinal studies which have been published

have been primarily concerned with predicting reading fail-

ure or have inferred support for lag mechanisms out of

changing correlations of the predictive measures.







The present study, in contrast, utilizes the same

groups of dyslexic and normal readers who have been fol-

lowed since early kindergarten. Repeated measurements are

utilized on eight different psychological measures which

have been found to be highly related to reading performance

across the age-span examined (ages 5 to 11). With the

passage of six years since the inception of the Florida

Longitudinal Project, the specific predictions of the Satz-

Sparrow Model of Specific Developmental Dyslexia can now

be tested.

In addition to expected main effects for group (dys-

lexics vs. normal readers) and for time (grades kindergar-

ten, two and five), two specific hypotheses will be tested:

(1) Hypothesis I At the younger ages, dyslexic children

will be more impaired than normals on sensorimotor percep-

tual skills (Factor I measures). Over time, performance

differences between dyslexics and normal readers will de-

crease, producing a significant group x time interaction;

and (2) Hypothesis II Dyslexic children will be more im-

paired than normals on language conceptual tasks (Factor II

measures) at the older ages. Performance differences be-

tween dyslexics and normal readers will increase over time,

yielding a significant group x time interaction in the

opposite direction to the Factor I measures. Hypothesis II

is consistent with either a lag or a deficit position. A

critical test of a lag model, however, rests in Hypothesis I,

that is, in finding that over time dyslexics "catch up" to

the normals in their performance.













METHOD


Subjects

Subjects for the present investigation were selected

from the entire population of white males who entered

Kindergarten in Alachua County, Florida, in 1970 (Sample I)

and the first cross-validation sample of white males

entering Kindergarten in the same county in 1971 (Sample

II). Characteristics of this population have been fully

described elsewhere (Satz et al., 1977).

At the end of the sixth year of follow up, a total of

238 boys from the original samples of 678 who were still

in residence in Alachua County were given the Wide Range

Achievement Test (WRAT) (Jastak & Jastak, 1965). On the

basis of a computed discrepancy score (chronological age

in months WRAT Reading Level age in months), 138 subjects

were selected who met the criteria for the four following

reading groups: (1) Dyslexics (N=35) Those subjects

whose WRAT discrepancy score was -25 months or less (two

years or more behind in reading). All dyslexics met at

least one or more of the World Federation Criteria for

specific developmental dyslexia (Taylor, Satz & Friel, 1978).

The mean WRAT discrepancy score was -33.68 (s.d.=6.19);

(2) Normal Control One Norm 1 (N=42) Those subjects

whose WRAT discrepancy score was between 0 months and 12







months (up to one year ahead in reading). The mean WRAT

discrepancy score was 6.67 (s.d.=3.95); (3) Normal Control

Two Norm 2 (N=18) Those subjects whose WRAT discrepancy

score was between 13 months and 24 months (up to two years

ahead in reading). The mean WRAT discrepancy score was

17.71 (s.d.=3.75); and (4) Normal Control Three Norm 3

(N=43) Those subjects whose WRAT discrepancy score was

25 months or more (more than two years ahead in reading).

The mean WRAT discrepancy score was 41.26 (s.d.=13.68).

Thus, the subjects for this study consisted of a group of

severely retarded readers dyslexicss) and three normal con-

trol groups.

Neuropsychological Measures

Performance on eight neuropsychological tasks which

were found to be highly predictive of reading achievement

was examined (Satz et al., 1977). Each subject was ad-

ministered these tasks at Time 1 (early Kindergarten; mean

age 66.2 months), Time 2 (third follow up year; mean age

92.5 months) and Time 3 (sixth follow up year; mean age

130.34 months).

Factor I (sensorimotor-perceptual) tasks included: (1)

the Finger Localization Test (FL) total correct score;

(2) the Recognition Discrimination Test (RD) total cor-

rect at time limit; (3) the Embedded Figures Test (EF) -

total correct at time limit; and (4) the Beery Test of

Visual Motor Integration (VMI) age equivalent in months

(Beery & Buktenica, 1967). To avoid a ceiling effect at







Time 3, norms for the Recognition Discrimination and the

Embedded Figures tests were extended through Grade 5 during

the sixth year of follow up.

Factor II (conceptual-language) tasks included: (1)

the WPPSI or WISC Similarities subtest (SIM) raw score

(Wechsler, 1949; Wechsler, 1967); (2) a Verbal Fluency Test

(VF) total score; (3) the Peabody Picture Vocabulary Test

(PPVT) mental age in months (Dunn, 1959); and, (4) Alpha-

bet Recitation (ALPH) total letters recalled. Each of

these measures has been described in greater detail in an

earlier publication (Satz & Friel, 1973). No subjects were

included who had missing data on any of the measures at any

point of time.

Statistical Analyses

The independent variables were reading group (4 levels)

and time of testing (3 levels). The dependent variables

were the eight neuropsychological tests listed above.

Analyses of Variance. Separate multivariate analyses

of variance (MANOVA) were utilized to compare the dyslexics

with each of the normal control groups on the Factor I and

Factor II dependent measures. The General Linear Models

program of the Statistical Analysis System (Barr, Goodnight,

Sall & Helwig, 1976) was used for these analyses. Univari-

ate Analyses of Variance (ANOVA) were then run comparing

the four reading groups on each of the dependent measures,

utilizing the Biomedical Computer Package P series, BMD

P2V, (Brown, 1977).







Trend Analyses. Separate regression analyses were also

performed on each of the eight dependent measures for each

of the four reading groups to test for the presence of a

common linear or quadratic trend over time (Hale, 1977).

The General Linear Models regression program of the Sta-

tistical Analysis System (Barr et al., 1976) was employed

in these analyses.

All procedures were run at the Northeast Regional Data

Center, University of Florida, Gainesville.












RESULTS

Multivariate ANOVAS Factor I and Factor II Tasks

Dyslexics vs. Norm 1

Multivariate analyses of variance on the Factor I

tasks (Finger Localization, Recognition Discrimination, Em-

bedded Figures and Beery Visual Motor Integration) yielded

significant main effects for reading group (Hotelling trace

=.70, F approximation ,75=13.14, p<.001), time (Hotelling

trace=10.33, F approximation ,300=193.65, p .001) and a

significant group x time interaction (Hotelling trace=.65,

F approximation8,300=12.10, p-.001).

Similarly, multivariate analyses of variance on the

Factor II tasks (Similarities, Peabody Picture Vocabulary

Test, Verbal Fluency and Alphabet Recitation) yielded sig-

nificant main effects for group (Hotelling trace=.54, F

approximation4,75=10.07, p <.001) and time (Hotelling trace=

9.44, F approximation8,300=176.91, p <.001) and a signifi-

cant group x time interaction (Hotelling trace=.29, F

approximation8,300 =5.50, p<.001).

Dyslexics vs. Norm 2

Multivariate analyses of variance on the Factor I tasks

yielded significant main effects for reading group (Hotel-

ling trace=.84, F approximation4,51=10.76, p <.001), time

(Hotelling trace=10.32, F approximation8,204=131.59, p<.001)







and a significant group x time interaction (Hotelling

trace=.82, F approximation8,204=10.42, p<.001).

Likewise, multivariate analyses of variance of the

Factor II tasks also yielded significant main effects for

group (Hotelling trace=.57, F approximation4,51=7.21,

p<.001), time (Hotelling trace=9.15, F approximation8,202

=115.47, p<.001) and a significant group x time interaction

(Hotelling trace=.68, F approximation8,202=8.60, p<.001).

Dyslexics vs. Norm 3

Multivariate analyses of variance on the Factor I tasks

yielded significant main effects for group (Hotelling trace=

1.33, F approximation4,83=27.57, p<.001), time (Hotelling

trace=ll.ll, F approximation8,332=230.54, p< .001) and a

significant group x time interaction (Hotelling trace=l.01,

F approximation8,332=20.86, p<.001).

And again, multivariate analyses of variance of the

Factor II tasks also yielded significant main effects for

group (Hotelling trace=1.42, F approximation4,83=29.48,

p<.001), time (Hotelling trace=11.65, F approximation8,318

=231.58, p<.001) as well as a significant group x time

interaction (Hotelling trace=.94, F approximation8,318=

18.75, p<.001).

Univariate ANOVAS Factor I Tasks

Finger Localization

The mean scores on the Finger Localization Test for

each of the four groups at each time are presented in Table 1.

With the maximum possible score of 46, examination of these









Table 1

Mean scores on Finger Localization Test for the
dyslexics and normal controls over time





READING GROUPS


DYSLEXICS
(N=35)

25.31

41.11


NORM 1
(N=42)

35.76

45.95


NORM 2
(N=18)

37.24

43.29


3 44.83


NORM 3
(N=43)

37.19

44.14

45.72


TIME


45.09 45.12







Norm 3
45 Norm 2
Norm 1
Dyslexics





4o


iI



35 /
0



S30









25 -
3 I -








1 2 3
Time

Figure 1. Mean raw scores on Finger Localization Test
by groups over time.



















































































o
Z


4-1
0



OU
ct







Oci

0d



0


0 0
OO







.r-I
IC
eo

















P-l


M -


0
z


o
z


o
z







means demonstrates a ceiling effect for the dyslexics

and the three control groups by Time 3 (Figure 1). A re-

peated measures analysis of variance (Brown, 1977, p. 550)

yielded significant main effects for group (F3,133=20.45,

p .001), time (F2,266=239.07, p <.001) and a significant

group x time interaction (F6,266=15.76, p .001). For the

dyslexics, Duncan's post hoc comparisons (Winer, 1971) found

that the scores at Times 1, 2 and 3 were all significantly

different from each other (p <.05). Among the normal con-

trol groups, only Time 1 scores were significantly different

(p(.05) from Time 2 ahd Time 3 (ceiling effect). Because

of the significant group x time interaction, a Z approxima-

tion of the t-test (Searle, 1971) was utilized for post hoc

pair-wise comparisons between the groups at each time of

measurement (Table 2). At Time 1, dyslexics had signifi-

cantly lower scores than each of the normal control groups.

By Time 3, however, a ceiling effect on the task obviated

any group differences.

Recognition Discrimination

The mean raw scores and mean percentage correct for

the dyslexics and three normal control groups are presented

in Table 3. Because performance neared the maximum score

of 15 by Time 3, nine items were added to the test and norms

were extended through Grade 5 during the sixth follow up

year. Comparison of the mean raw scores (Figure 2) with

the percentage correct (Figure 3) suggests that the task

requirements were more difficult at Time 3. A repeated


















4-1




O
0


co
*r4

*v-4 0)


4-1






0 0
re






0) 4--
ou











0)





U )
P 0







4-1



0 4












U)
OU







u)
Su
cu




o N


Co


o a

->.
ur

O u



kO
0
$u
S)
T3
eX
( 0
(fl4=
4-)


-4
II
oz












cO
oz

II


C,,


oz
is-'


HuL

SII
CY)



r-


.-4 qC








Norm
Norm


20





18


0 Dyslexics
!


10 [-


8 1-


1 2 3
Time

Figure 2. Mean raw scores on Recognition Discrimination
Test by groups over time.


Norm 1


16


o 14
12




12






















/ \


Norm 3

* Norm 2


b Norm 1





b Dyslexics


60 H


50 F


1 2 3
Time

Figure 3. Mean percentage correct on Recognition Discrim-
ination Test by groups over time.


100


90 -


70







measures analysis of variance (Brown, 1977) disclosed a

significant main effect for group (F3,133=23.45, p<.001)

and for time (F2,266=539.07, p <.001); however, the group x

time interaction was not significant. For the dyslexics

as well as each of the normal control groups, Duncan's

Multiple Range Tests yielded significant differences in the

scores between each of the times of measurement. At Time 1,

Time 2 and Time 3, the dyslexics had significantly poorer

performance than each of the normal control groups. For

the normal controls, however, a variable pattern of differ-

ences in the mean scores emerged between Time 1 and Time 3.

Embedded Figures

Table 4 presents the mean raw scores and mean percen-

tage correct for the dyslexics and the three normal control

groups. Overall total score increased with the extension

of the task through Grade 5 (Figure 4), although there was

a decrease in the relative percentage correct for the con-

trol groups (Figure 5). Only the dyslexics showed a modest

gain in performance between Time 2 and Time 3. A repeated

measures analysis of variance (Brown, 1977) produced sig-

nificant main effects for group (F3,133=14.57, p<.001) and

for time (F2,266=805.58, p 4.001) as well as a significant

group x time interaction (F6,266=3.10, p <.01). For each

of the four reading groups, mean scores were significnatly

different at each of the three times of measurement (Dun-

can's Multiple Range, p<.05). Post hoc pair-wise compari-

sons suggested that a floor effect may have occurred at
























j--4
u)







-4 4J
a>








0









0*0
o e













0 0
bD-rl

CC













0) 0





w


U)
0J4











a)
0










U)
C
4W












u
0 0

u








CO


m >,


OZ














Z-
oz
Z -


r--4^
c'l

SII
OZ
2 V


U)
rH)


on]v


r--4 CN










Norm 3


Norm 1
Norm 2






0 Dyslexics


/
/


/
/
/


1 2 3

Time

Figure 4. Mean raw scores on Embedded Figures Test by
groups over time.











-A Norm 3

Norm 1

Norm 2





-,-0 Dyslexics


60 h


50 -


1 2 3
Time

Figure 5. Mean percentage correct on Embedded Figures Test
by groups over time.


90 r


80 h


70 h







Time 1 in that with the exception of the most superior

reading group (Norm 3), all other groups did not differ

from each other. By Time 2 and Time 3, dyslexics had sig-

nificantly lower scores than the normal controls. By Time

3, the same pattern of very superior performance by the

Norm 3 group again emerged (Figure 4).

Beery Test of Visual Motor Integration

Age equivalent scores in months are presented in Table

5 for the four reading groups. As can be seen in Figure 5,

there was a significant increase in scores for all groups

across time (Duncan's procedure, p <.05), with the dyslexic

group maintaining its low performance relative to the nor-

mals at each point of measurement. Significant main effects

for group (F3,133=19.36, p<.001) and for time (F2,266=313.51,

p <.001) and a significant group x time interaction (F6,266

5.90, p< .001) were observed by a repeated measures analysis

of variance (Brown, 1977). At Time 1, the dyslexic scores

were not significantly different from the first two control

groups (Table 2). By Time 2 however, the dyslexics were

significantly different from the normal controls, a posi-

tion which maintained through Time 3.

Univariate ANOVAS Factor II Tests

WPPSI/WISC Similarities

The mean total raw scores on the Similarities subtests

of the WPPSI in Time 1 and the WISC at Times 2 and 3 are

presented in Table 6. Over time, there was a significant

increase in scores (Duncan's procedure, p< .05) between








Table 5

Mean age equivalent scores on the Beery Test of
Visual Motor Integration for the dyslexics
and normal controls over time






READING GROUPS


DYSLEXICS
(N=35)

55.74

71.49


NORM 1
(N=42)

62.67

85.21


NORM 2
(N=18)

62.65

86.59


3 88.11


NORM 3
(N=43)

68.28

90.95

127.35


TIME


112.74 111.71







130





120





110


Figure 6.


Mean Beery VMI
over time.


age equivalent scores by groups


Norm 3







Norm 1
Norm 2







0 Dyslexics
/


Time









Table 6


Mean raw scores on Similarities Test for the
dyslexics and normal controls over time


READING GROUPS


DYSLEXICS
(N=35)


10.37

15.20


3 11.89


TIME


NORM 1
(N=42)

13.57


17.14


NORM 2
(N=18)

11.59

17.59


NORM 3
(N=43)

14.28

18.28

19.30


16.78 16.53










Norm 3







Norm 1
Norm 2


Q
/ \
/


b Dyslexics


10 -


5


1 2 3
Time

Figure 7. Mean raw scores on Similarities Test by groups
over time.


15








the dyslexic group (Figure 7), although the scores between

Time 2 and Time 3 were not significantly different for each

of the normal control groups. A repeated measures analysis

of variance (Brown, 1977) of the raw scores produced signif-

icant main effects for group (F3,133=26.94, p <.001) and

for time (F2,266=88.31, p <.001), with a significant group

x time interaction (F6,266=5.12, p< .001). As can be seen

in Table 7, dyslexics had significantly lower scores com-

pared to all by the second normal control group (Norm 2)

at Time 1. By Time 3, dyslexics were significantly below

all of the controls. It was not until Time 3, however,

that no significant differences were found between each of

the normal groups. Interestingly, between Time 2 and Time 3

all but the most superior reading group (Norm 3) showed a

decrease in total raw scores which could not be attributed

to changes in task requirements, with the greatest drop

occurring in the dyslexic group (Figure 7).

Peabody Picture Vocabulary Test

The mean scores in months on the PPVT for each of the

groups are presented in Table 8. As expected, each of the

groups showed a significant increase in scores over time

(Duncan's procedure, p .05), although the dyslexics main-

tained mean scores roughly half those of the most superior

reading group at each measurement (Figure 8). A repeated

measures anlaysis of variance (Brown, 1977) yielded signif-

icant main effects for group (F3,13324.05, p <.001) and

time (F2,266=881.52, p<.001), with a significant group x












































O4-4
0 U

oH







o
00




0
ti











.,- -
cl)












I
* 4
0 M


cli
o-4
0~r
cj c
CLr
U 3)
0
43>-
b0
4-1
M b


'-4 r


C14

mE
O
Z


CI


rn
X

CO1


C14

o
O
z


0
z


o

0
Z









Table 8

Mean mental age equivalent scores on Peabody Picture
Vocabulary Test for the dyslexics
and normal controls over time





READING GROUPS


TIME


DYSLEXICS
(N=35)


60.23

88.91


3 119.86


NORM 1
(N=42)

72.95

98.86

141.31


NORM 2
(N=18)

73.65

101.88

145.24


NORM 3
(N=43)

80.53

110.74

166.14








170


160


150


140


130


120


110


100


90


80


70


1 2 3
Time

Figure 8. Mean mental age equivalent scores on the Peabody
Picture Vocabulary Test by groups over time.


SNorm 3




Norm 2

SNorm 1






p0 Dyslexics







time interaction (F6,266=8.26, p<.001). Examination of

Table 7 reveals that, in general, dyslexics performed sig-

nificantly worse than the normal control groups. Differ-

ences between the normal control groups varied at each

time comparison with no clear pattern emerging.

Verbal Fluency

Mean verbal fluency scores for each of the reading

groups at each time of testing are presented in Table 9.

For each of the groups, total score increased significantly

(Duncan's procedure, p<.05) between each time of measure-

ment (Figure 9). A repeated measures analysis of variance

(Brown, 1977) produced significant main effects for group

(F3,133=8.25, p<.001) and for time (F2,266=321.10, p<.001).
However, the group x time interaction did not approach sig-

nificance (F6,266=1.22, p >.30). For each of the groups,

the scores increased significantly over time (Duncan's

procedure, p <.05), with scores roughly doubling between

Time 1 and Time 3. As Table 7 reveals, the groups did not

generally differ from each other at Time 1 which may have

reflected a floor effect on the task. By Times 2 and 3,

the dyslexics had significantly lower scores than each of

the normal control groups, while the normals did not differ

from each other.

Alphabet Recitation

Mean number of letters recited by each of the groups

at time point are presented in Table 10. Clearly the

dyslexics were far below the maximum score at Time 1 and




46



Table 9

Mean number of words recalled on Verbal Fluency Test
for the dyslexics and normal controls over time






READING GROUPS


TIME


DYSLEXICS
(N=35)


18.63

27.91


NORM 1
(N=42)

21.50

32.86


NORM 2
(N=18)

20.65

33.24


3 37.00


NORM 3
(N=43)

23.44

34.39

45.23


44.98 41.82








Norm 3
Norm 1


Norm 2


Q Dyslexics


I ------4--


Time


Figure 9.


Mean number of words on Verbal Fluency Test by
groups over time.


45 k


40 h


35


U)

0
T 30
0






25





20






15









Table 10

Mean number of letters recited on Alphabet Recitation
for the dyslexics and normal controls over time





READING GROUPS


TIME


DYSLEXICS
(N=35)


13.37

23.14


3 25.46


NORM 1
(N=42)

20.21

26.00


NORM 2
(N=18)

23.47

25.59


26.00 25.88


NORM 3
(N=43)

23.35

26.00

26.00









Norm 3
Norm 1
Norm 2
Dyslexics
Dyslexics


S--I -


Time

Figure 10. Mean number of letters recalled on Alphabet
Recitation Test by groups over time.








showed significant gains between Time 1 and Time 3 (Dun-

can's procedure, p <.05). For the normal controls, how-

ever, in general by Time 2 performance was almost flawless.

As Figure 10 depicts, a ceiling effect operated for the

normal groups by Time 2, but the dyslexics did not approxi-

mate the maximum score until Time 3. A repeated measures

analysis of variance (Brown, 1977) yielded significant main

effects for group (F3,133=17.60, p <.001) and for time

(F2,266=61.08, p<.001) as well as a significant group x

time interaction (F6,266=9.60, p <.001). Post hoc compari-

sons between each of the groups (Table 7) found that the

dyslexics were significantly below the normals at the first

two times of measurement but that by Time 3 there were no

differences among any of the groups.

Trend Analyses

Hale (1977) suggests that a trend analysis may be the

most meaningful test for significant developmental increases

in performance when intermediate level age groups are in-

volved. For this reason, regression analyses were per-

formed on each of the dependent measures for the dyslexics

and each of the normal control groups. Because there were

three data points, tests for a common linear or quadratic

trend were run utilizing the SAS (Barr et al., 1976) PROC

REGR program. Results of these analyses have been sum-

marized in Table 11.

The trend analyses were run to determine whether the

main effects due to time found on the individual ANOVAS




















C')



Z II


4M
U



0 r-4



Cd0



o Q)







0
u t






44 C)r




-U t






r -4

.,0 4-
"-4





b4
T)--
T)M






4-1l
ri



[ne


oo

oz


































U)

C)


-CA -4 -a 4


C") 0 cn
m C-1 < l








cq








-o mo C1



-4 C- 0% r--



H4 HA c')
o o ^ CD




















^ rl C r-


co 1 4
0









coq















cCl


C-) C 4 C


















CO C --C













,-cO-
o^ r' o^






o a C













C0 Co (-n









p O 04








44
'a'-
n KS a
HA e a
eC r
Q'a (-C



N / HA a-


C") -- CC)
.2 .2' *
a' a'n ~
Co i-





a-'n


C')



U-)
In


0 'CN 0-

co -





0

f>


HA
I Lr ) 4 (
0000

AVVV




. .- '








could best be described as a linear or nonlinear (i.e.,

quadratic) model. As can be seen in Table 11, the effect

of time on Finger Localization, Alphabet Recitation and

Similarities scores can generally be described by a non-

linear effect for all groups except Norm 2. In contrast,

the other five tasks (Recognition Discrimination, Embedded

Figures, Visual Motor Integration, Peabody Picture Vocabu-

lary Test and Verbal Fluency) fit linear models for the

dyslexics. For the Normal readers, results were more

variable particularly among the Factor II subtests.

For this reason, a one-way analysis of variance was

run on the slopes of the average regression lines of each

group. Because of the ceiling effect on Finger Localization

and Alphabet Recitation, these tests were excluded from

the analysis. For the Factor I measures, there was no

significant main effect due to group on Recognition Dis-

crimination (F3,133=0.55, p >.05). Significant main effects

for group were found on Embedded Figures (F3,133=3.77,

p < .05) and on Visual Motor Integration (F3,133=7.67, p<

.001). For the Factor II measures, significant main effects

for group were found on Similarities (F3,133=4.53, p <.01)

and Peabody Picture Vocabulary Test (F3,133=10.17, p<.001)

but not on Verbal Fluency (F3,133=1.62, p>.05). Table 12

presents the mean slopes and the post hoc Duncan's Multiple

Range contrasts for each measure by groups. As can be seen,

dyslexics were significantly different from the normal

readers on EF and VMI. Normal readers were generally not








Table 12


Mean slopes by groups of dependent measures*
and Duncan's post hoc pair-wise comparisons






READING GROUPS


FACTOR I


DYSLEXICS


43.29

53.29
16.19
16.19


VMI


NORM 1

41.73


66.31


NORM 2

47.06

66.32


25.04 24.53 29.53


FACTOR II


SIM

PPVT


0.76 1.61


29.84


9.19


34.18


11.74


2.47


35.79


10.59


2.51


42.80

10.90


* FL and ALPH excluded due to ceiling effect. Any two
means not joined by a continuous line are significantly
different at p (.05.


NORM 3

44.48

66.22




54


different from each other except on the PPVT where Norm 3

was significantly different from each of the other groups.














DISCUSSION

As expected, the normal control groups generally

demonstrated superior performance compared to the dyslexics

on both Factor I sensorimotorr, perceptual) and Factor II

(conceptual-language) measures at each time of measurement.

The multivariate analysis of variance on the Factor I and

Factor II tasks yielded significant main effects for read-

ing group and for time as well as significant group x time

interactions. However, examination of the univariate

analysis and trend analysis of the component tasks of each

of the Factors did not support Hypothesis I (test of the

lag theory).

For example, among the Factor I measures, a ceiling

effect was found on the Finger Localization Test by Time 3;

while a significant group x time interaction was only found

on three of the four measures (Finger Localization, Beery

VMI and Embedded Figures). The patterns of performance

shown by the dyslexics and each of the normal control groups

varied somewhat from test to test but generally were not

consistent with a unitary developmental lag phenomenon.

Instead, dyslexics did not catch up to the normal readers

as predicted in Hypothesis I. For the Factor II measures,

a ceiling effect was found on Alphabet Recitation. Signif-

icant group x time interactions were found on Similarities,








the Peabody Picture Vocabulary Test scores and on Alphabet

Recitation but not for the Verbal Fluency measure. In

general, performance on the Factor II conceptual-language

measures by the dyslexics was similar to the Factor I find-

ings, that is, they did not improve over time and, in some

instances, their performance became more impaired.

In the face of these results, reexamination of the

implications of such terms as developmental difference,

deficit and lag seems called for. As the earlier review

suggests, there are an increasing number of studies which

find that dyslexics attain lower scores than normal reading

groups. Depending upon the orientation of the investigator

and the age at which there scores are obtained, such quan-

titative differences have been interpreted to reflect either

a deficit or a lag.

What are the assumptions which underlie each position?

A deficit model implies that defective or deficient cortical

functioning is the mechanism by which the quantitative dif-

ference between dyslexics and normals is observed (Kinsbourne

1975; Usprich, 1976; Rourke, 1976; Bishop, 1977; Fletcher

& Satz, 1978b). Further, Zigler (1969) states that a defect

model rests upon demonstration of discontinuities in per-

formance (different scores) rather than upon observation of

greater or lesser amounts of the same continuous variable.

However, compensatory mechanisms which are known to operate

in the presence of brain disease (Benson 6 Geschwind, 1969;

Luria, 1973) may allow for the development of other less








adequate or more inefficient strategies to compensate for

the presumed structural lesion. Further, recent work on

the longterm consequences of early injury (Goldman, 1974)

as well as differential rates of cortical development

(Epstein, 1974a; 1974b), the effects of experience on

neural development (Gottlieb, 1978) and the characteristics

of the language orthography itself (Makita, 1968) suggests

the interactive effects of other variables which can affect

performance at different times. While evidence of a clear

neurological defect in these children is, at best, contra-

dictory, ultimately this position rests upon a hypothetical

construct (brain defect) to explain the observed behavioral

events (Satz, 1977).

In some ways, the delay position also ultimately rests

upon a hypothetical construct (brain immaturity) to explain

the observed behaviors. Unlike the deficit model, lag

theorists suggest that the quantitave differences observed

in age-matched dyslexics and normal readers reflects a

qualitatively immature performance by the dyslexics

(Kinsbourne, 1975; Satz et al., 1977). Thus, dyslexics

are more like younger normal readers and with sufficient

time should approximate normal performance, at least for

earlier developing perceptual skills (Satz & Sparrow, 1970).

This position, however, also relies ultimately upon the

quantitative level of performance achieved by dyslexics

which is interpreted as "lagging" behind normals. Much

less attention has actually been paid to qualitative








differences in strategies employed by dyslexics, in types

of errors committed or in analysis of rates of acquisition.

When cross sectional age differences are found as, for

example, in the study by Strete et al. (1978) they are

interpreted post hoc to reflect a developmental lag whether

or not both qualitative and quantitative differences are

found. Yet it was precisely on the basis of the rate

phenomenon and the quality of performance the Bender's (1958)

original developmental lag hypothesis was formulated.

A particular weakness of the Satz-Sparrow model lies

in Hypothesis II which leads to predictions which could be

just as consistent with a deficit model. Thus, according

to Hypothesis I, if dyslexics improve, they had delayed

development and, according to Hypothesis II, if dyslexics

don't improve, they also had delayed development. Yet it

would seem equally as plausible to hypothesize that failure

to improve is contradictory to the notion of delayed normal

development.

Results of the present investigation suggest that,

depending upon which task is examined, one could invoke

either a lag, a difference or a deficit model of develop-

mental dyslexia. The overwhelming evidence, however, is

consistent with a deficit model.

On initial reading, the lag model would seem to fit

the data obtained on the Finger Localization and Alphabet

Recitation tests, in the by Time 3 no differences were

found between the scores of the dyslexics and normal controls.








However, the ceiling effect found on these measures pre-

vents one from being able to determine if the dyslexics did

indeed "catch up" to the normals or if measurement con-

straints obscured the demonstration of the presence or

absence of these differences between the groups. These two

measures tap earlier developing sensory-perceptual and rote

naming skills which are highly predictive of later reading

achievement for severe dyslexics and very superior readers

(Muehl & DiNello, 1977; Satz et al., 1977). The lower

concurrent validity of these same two measures with read-

ing performance in older readers (Fletcher & Satz, 1977a)

is also consistent with the ceiling effects found in the

present study. Therefore, rather than interpret ceiling

effects as support for a lag model, a more cautious inter-

pretation is that limitations of the measures employed (i.e.,

ceiling effect) do not allow for an adequate demonstration

of a "catching up" phenomenon.

The absence of a group x time interaction on the

Recognition Discrimination and Verbal Fluency tests sug-

gests that dyslexics begin with and maintain a consistently

different level of performance than normal readers on these

perceptual-language tasks (Models 4 and 6, Rourke, 1976).

On the remaining four tasks (Embedded Figures, Beery

VMI, Similarities and Peabody Picture Vocabulary Test),

dyslexics persist in demonstrating poorer performance

relative to normals. One interpretation of these latter

results is that they are consistent with a deficit in








cortical functioning among dyslexics. Over time dyslexics

show greater and greater impairment on a variety of cogni-

tive tasks.

A second interpretation, following Lenneberg (1967,

p. 170), is that as the dyslexics grow older, the initially

slight lag in performance increases as they fall further

and further behind normals. While early milestones are

only slightly behind schedule, as the dyslexics grow older

the spacing between milestones (here defined as normal

performance on these tasks) become more prolonged without

a change in order or sequence. The present data, however,

do not allow for a clear choice between these two interpre-

tations.

The overall results are generally consistent with the

age-related predictive power of these measures observed in

earlier studies of the same population of children (Satz

et al., 1977; Fletcher & Satz, 1977; Fletcher, 1978; Taylor

et al., 1978). Performance on these tasks is related to

reading achievement, with the greatest separation between

very superior readers and severely disabled readers.

Clearly, however, neither a unitary deficit model nor a

maturational lag adequately explains the results observed.

Whether the same results would be found in the milder

cases of reading disturbance is a research question worth

pursuing.

One might argue that the present results have been

obscured by weaknesses in the selection procedures employed.








Although the WRAT is a measure of sight reading rather than

of reading comprehension, previous work has shown it to be

related to teacher judgment of overall reading skills

(r=.80, Taylor et al., 1978) and to the level of linguistic

competence among dyslexics studied (Fletcher, 1978). Fur-

ther, since only children who were always behind in reading

achievement were selected for the dyslexic group, it may

have weighted the dyslexic sample toward the most severe

end of the continuum (Rutter & Yule, 1975). Since the

sample was selected in the sixth follow up year, efforts

were made to avoid including dyslexic subjects whose dis-

ability was of recent origin. Thus, only a highly selec-

tive and narrow sample of dyslexics was employed with the

expectation that the sample was representative of children

with chronic histories of reading failure.

Because of the absence of differences in neuropsycho-

logical performance between a larger sample of disabled

readers and severe dyslexics at Time 1 and Time 3 (Taylor

et al., 1978), no distinction was made between children who

met all or most of the World Federation guidelines for

dyslexia. However, post hoc Chi-square analyses (Winer,

1971) of the four reading groups revealed that the dys-

lexics had significantly more "soft signs" on the neurolog-

ical exam administered during the third follow up year

(X2=58.82, df=6, p<.001). The dyslexics also were found

in the lower of two broad SES categories (I=low; II=middle

or above) significantly more often than the control groups








(X2=27.43, df=3, p<.001). Thus, one cannot fully rule

out the impact of neurological factors or socio-cultural

factors on the results of the present investigation. It

is worth noting, however, that earlier studies suggested

that the effects of SES on performance were not powerful

(Satz et al., 1977).

Further, the absence of an appropriate intelligence

quotient covariate is a weakness of the original design

of the longitudinal natural history study which has carried

over to the present study. Rutter and Yule's (1975) work

on specific reading retardation suggests that controlling

for the effects of general intelligence would have enabled

one to distinguish between reading backwardness and spe-

cific reading retardation among the dyslexics in the pre-

sent study. An earlier study (Satz et al., 1976) demon-

strated that identically matching dyslexics and normal

readers on PPVT IQ did not affect the power of these tests

to predict reading achievement among severe dyslexics and

superior readers. In the present study, post hoc examina-

tion of the PPVT scores revealed that by Time 3 all but

six of the dyslexics had achieved scores in the Low Average

Range. Some dyslexic subjects showed dramatic gains of as

much as 30 points between Time 1 and Time 3 reflecting the

liability of performance in younger aged disabled readers.

Nevertheless, the effect of this verbal intelligence esti-

mate on both verbal and perceptual performance is difficult

to assess in the present study.








Another sampling bias may have been introduced in the

present study since only about one-sixth of the original

population of 678 subjects was selected for study. While

the total sample size was certainly adequate (N=138), one

must always be cautious in assuming that sample results

unerringly reflect true population characteristics. Differ-

ences in performance between subjects in the present study

and the total longitudinal population may reflect the

narrower sample employed.

One additional point worth noting is that comparisons

between the dyslexics and normal readers varied as a

function of the normal control's level of reading profi-

ciency. Future research should pay greater attention to

the criteria employed in the selection of normal readers

as well as of dyslexics.

Finally, it could be argued that the selection pro-

cedure guaranteed that those dyslexics who improved over

time were not included. This argument seems specious at

best since the intent of this investigation was to examine

the neuropsychological performance of children who, despite

educational experience and remediation efforts, failed to

make adequate progress in reading achievement. Whether

"recovered" dyslexics perform differently on these measures

is a question which has yet to be addressed in the Florida

Longitudinal Study.

In conclusion, the present results are not consistent

with the Satz-Sparrow Developmental Lag Theory of dyslexia.








Dyslexics did not "catch up" with the normal readers on

higher order cognitive-perceptual tasks (Factor I) which

was necessary to support the lag theory. Over time their

performance on Factor II tasks, as predicted, fell further

behind the normal readers on complex conceptual-language

measures. It was only on earlier developing sensory-

perceptual (Finger Localization) and naming skills (Alpha-

bet Recitation) that the dyslexics eventually looked like

the normal readers. Unfortunately, the presence of a

ceiling effect on these tests does not permit one to know

whether the dyslexics actually "caught up" or whether the

tests were not sensitive measures of differences in the

older ages.

As Cruickshank (1977) and Fletcher and Satz (1978b)

note, dyslexia is a complex perceptual disorder in which

both visual-spatial and verbal mediational processes may

interact with developmental factors to affect ultimate

reading proficiency. The present data suggests that read-

ing disability involves more than just difficulty in read-

ing. Kass (1977) remarks, "It has been said that the

handicap is subtle. A better adjective would be insidious.

All that is subtle about learning disability is that we do

not understand the characteristics or sets of characteris-

tics (of the disorder)" (p. 425). To this might be added

that no single unitary model is adequate to explain the

mechanism underlying the performance of the severe dyslexics

examined in the present study.








Perhaps, the notion of developmental lag is only

applicable to those functions which develop early and which

have an automatic ceiling effect (Finger Localization and

Alphabet Recitation). For example, basic motor and language

skills such as walking and talking are acquired by most

children. Yet between children one observes variations

in both the age at which these skills are accomplished

and in the rate at which proficiency is attained. Differ-

ences in acquisition and level of proficiency are most

marked among children who are developmentally at risk (e.g.

retardation, sensory handicaps, socioculturally deprived)

but eventually most will attain these basic developmental

milestones. It may be that the concept of maturational

lag has most relevance only to the most basic, early develop-

ing milestones or skills. The concept becomes too loose

and has little heuristic value when applied to the many

complex functions characteristic of the older child. Per-

sistent difficulties on these tasks among dyslexics at

Time 3 seems to be analogous to the clinical observation of

"soft signs" among disabled readers (Peters, Dykman &

Romaine, 1975; Denckla, 1977) and may, indeed, reflect more

than a maturational delay. In the Florida Longitudinal

Study, efforts were made to avoid similar ceiling effects

on the Recognition Discrimination and Embedded Figures

tests. However, by increasing the complexity of these

tasks at Time 3, Satz et al. (1977) may have introduced

other cognitive demands upon the children which precluded








any spurious support for a lag mechanism inherent in tests

which ceiling-out at early ages.

The verbal and perceptual difficulties demonstrated by

the dyslexics is also a powerful argument against the kind

of unitary deficit hypothesis proposed by Vellutino and

co-workers (1977). Dyslexics were impaired across a num-

ber of tasks. Further, while dyslexics do improve their

performance on some perceptual and language tasks it is at

a rate which is usually different from that of the normals

(trend analyses), again except for tasks which had an

early-occurring ceiling effect. Thus, it would seem that

improvement, but at different rates, may play at least a

small part in understanding how the severe dyslexic per-

formed over time. The fact that the same dyslexic child-

ren, followed over a six year period, continue to show

problems on a variety of cognitive tasks which tap both

language and perceptual dimensions clearly suggests that a

single underlying cognitive deficit (e.g. a verbal pro-

cessing defect) cannot be invoked to explain these results.

As Weiner and Cromer (1968) note, reading is the final

outcome of a number of interrelated processes. A delay

or alteration in any of the fundamental operations can

affect the ultimate proficiency and competency of the

reader. More recently, Doehring (1976), Rourke (1976)

and Fletcher and Satz (1978b) caution against seeking

overly simplistic explanations for reading disorders in

order to obtain premature closure on a preferred explana-

tory model.








The power of the longitudinal approach of Satz et al.

(1977) is that by following an entire population of child-

ren from entry into Kindergarten through their primary

school experiences and by obtaining repeated measurements

on the same functions, one could examine the course of

their development on a variety of psychological functions.

Out of that population, the present sample was selected

to test the specific predictions formulated by Satz and

Sparrow (1970) at the outset of the longitudinal study.

The opportunity to follow these same children for six

years yielded results which suggest that a developmental

lag theory is not adequate to explain the developmental

course of the severely dyslexic children studied. Lag

mechanisms may be operative in the case of early developing

and early maturing sensorimotor and mnemonic skills.

Clearly, this is not the case in higher-order, more complex

and integrative perceptual and language skills. The severe

dyslexics in the present study are impaired across more

than one performance dimension.

It may be that dyslexics utilize different cognitive

strategies than do normal readers. Whether there strate-

gies evolved to compensate, however inefficiently, for

delayed or deficient development cannot be determined from

the present results.

Two future research directions have been suggested

by the present investigation. Applying Zigler's (1969)

approach to familial retardation, an even more direct test








of a difference versus defect model may lie in a comparison

between dyslexics and a reading achievement-matched group

of younger normal readers. This study is now in the plan-

ning stages.

The recent emphasis on understanding the cognitive

strategies employed by normal and disabled readers (Kershner,

1977; Fisher & Frankfurter, 1977; Witelson, 1977) is one

other appropriate research direction. If we refocus our

attention not on the stimulus properties of the particular

tasks employed but upon how the reader "operates" (in

the Piagetian sense) upon these properties, we may begin

to understand how the developing brain functions in the

reading process and whether adult models of brain function

are applicable to children.














APPENDIX A
DESCRIPTION OF TEST BATTERY

1. Finger Localization (FL) Somatosensory test consisting

of five levels of performance. Level 1: Shielded uni-

lateral stimulations made to the fingertips; shield re-

moved between stimulations and subject is required to

point to the finger stimulated with the index finger of

his/her free hand. Five trials per hand, starting with

the preferred hand. Level 2: Shielded unilateral

stimulations made to the fingertips; subject identifies

each stimulated finger on a corresponding diagram of an

opened hand. Level 3: Shielded, randomized series of

three bilateral and ten unilateral stimulations made

to the backs of the subjects' hands; subject waved

hands) stimulated. Only bilateral trials scored.

Level 4: Shielded unilateral stimulations made to

the fingertips; subject recalled the number of finger-

tips touched. Instructions in the numbering of each

hand given immediately before the stimulations to that

hand. Level 5: Shielded simultaneous bilateral stimu-

lations made to pairs of fingertips; subject recalled

the number of fingertips stimulated on each hand. Five

pairs of stimulations, starting with the preferred

hand. Maximum score is 46.








2. Recognition Discrimination (RD) Visuoperceptual

matching-to-sample task requiring subject to identify

a geometric stimulus design among a group of four

figures, three of which are rotated and/or similar

in shape to the stimulus figure. Maximum score is 24.

3. Embedded Figures (EF) Visuoperceptual task requiring

subject to identify in which of three choice designs

a stimulus design is embedded. Maximum score is 24.

4. Beery Developmental Test of Visual-Motor Integration

(VMI) Visuomotor task requiring subject to reproduce

(copy) 26 stimulus line drawing designs. Discontinue

testing after three consecutive failures (scoring norms

in manual). Score reported in age equivalent in months.

5. Similarities (SIM) Subtest of the Wechsler Pre-School

and Primary Scale of Intelligence (Time 1) or the

Wechsler Intelligence Scale for Children-Revised (Times

2 and 3). Scored as in manual; raw scores rather than

scale score conversions utilized.

6. Peabody Picture Vocabulary Test (PPVT) Form A Picture

recognition vocabulary test requiring subject to select

which one of four stimulus line drawings goes with a

given vocabulary word. Maximum number of items is 150.

Mental age equivalent in months utilized.

7. Verbal Fluency Test (VF) Subject is required to name

in one minute as many objects as possible in each of

three rooms in his home (kitchen, bedroom, living room)

at Time 1. At Times 2 and 3, subject is required to




71



name as many words as possible in each of three cate-

gories (food, transportation, animals). Score is the

total number of items named.

8. Alphabet Recitation (ALPH) Subject is required to

recite the ABC's. Maximum correct is 26 letters

named, regardless of order given.













REFERENCES


Anderson, R.P. A perspective on learning disabilities.
Journal of Learning Disabilities, 1970, 3(3), 24-28.

Applebee, A.N. Research in reading retardation: Two
critical problems. Journal of Child Psychology and
Psychiatry, 1971 12, 91-113.

Bannatyne, A.D. Language, reading and learning difficul-
ties. Springfield, Illinois: Charles C. Thomas,
97T1.

Barr, A., Goodnight, J., Sall, J. and Helwig, J. A user's
guide to Statistical Analysis System 76. Raleigh,
North Carolina: SAS Institute, Incorporated, 1976.

Beery, K. andr Bktenica, N. Developmental test of visual-
motor integration. Chicago: Follett Educational
Corporation, 1967.

Bender, L. Problems in conceptualization and communication
in children with developmental alexia. In P.H. Hoch
and J. Zubin (Eds.), Psychopathology of communication.
New York: Grune and Stratton, 1958, 155-175.

Benson, F. and Geschwind, N. The alexias. In P.J. Vinken
and G.W. Bruyn (Eds.), Handbook of clinical neurology.
New York: Wiley Interscience, 1969, 112-140.

Benton, A. Developmental dyslexia: Neurological aspects.
In W.J. Friedlander (Ed.), Advances in neurology,
Volume 7. New York: Raven Press, 1975, 1-45.

Bishop, D.V. Language comprehension in children with develop-
mental language disorders. Unpublished manuscript,
1977.

Brown, M. (Ed.) Biomedical computer programs: P series.
Berkeley, California: University of California Press,
1977.

Bryan, T. and Bryan, J. Understanding learning disabili-
ties. Port Washington, New York: Alfred Publishing
Company, 1974.

Chissom, B.S. A factor analytic study of the relationship
of motor factors to academic criteria for first and
third grade boys. Child Development, 1971, 42, 1133-1143.








Critchley, M. The Dyslexic Child. Springfield, Illinois:
Charles C. Thomas, 1970.

Critchley, M. Developmental dyslexia: Its history, nature
and prospects. In D. Duane and M. Rawson (Eds.), Read-
ing, perception and language. Baltimore: York Press,
1975, 9-15.

Cruickshank, W. Myths and realities in learning disabili-
ties. Journal of Learning Disabilities, 1977, 10, 57-64.

Darby, R. Ear asymmetry phenomenon in dyslexic and normal
children. Unpublished Master's Thesis, University of
Florida, 1974.

de Hirsh, K. Specific dyslexia or strephosymbolia. In
G. Natchez (Ed.), Children with reading problems.
New York: Basic Books, 1968, 97-113.

de Hirsh, K., Jansky, J. and Langford, W. Predicting
reading failure. New York: Harper and Row, 1966.

Denckla, M.B. Minimal brain dysfunction and dyslexia:
Beyond diagnosis by exclusion. In M.E. Blau, I. Rapin
and M. Kinsbourne (Eds.), Topics in Child Neurology.
New York: Spectrum Publications, 1977, 243-261.

Denckla, M.B. and Rudel, R.G. Rapid automatized naming
(RAN): Dyslexia differentiated from other learning
disabilities. Neuropsychologia, 1976, 14, 471-479.

Doehring, D.G. Evaluation of two models of learning dis-
ability. In R. Knights and D.J. Bakker (Eds.), The
neuropsychology of learning disorders. Baltimore:
University Park Press, 1976, 405-412.

Douglas, V.I. Perceptual and cognitive factors as deter-
minants of learning disabilities: A review chapter
with special emphasis on attentional factors. In R.
Knights and D.J. Bakker (Eds.), The neuropsychology
of learning disorders. Baltimore: University Park
Press, 1976, 413-422.

Dunn, L. Peabody Picture Vocabulary Test. Circle Pines,
Minnesota: American Guidance Service, 1959.

Dykman, R.A., Ackerman, P.T., Clements, S.D. and Peters,
J.E. Specific learning disabilities: An attentional
deficit syndrome. In H.R. Myklebust (Ed.), Progress in
learning disabilities, Volume II. New York: Grune
and Stratton, 1971, 56-93.







Epstein, H.T. Phrenoblysis: Special brain and mind growth
periods. I. Human brain and skull development.
Developmental Psychobiology, 1974a, 7, 207-216.

Epstein, H.T. Phrenoblysis: Special brain and mind growth
periods. II. Human mental development. Developmental
Psychobiology, 1974b, 7, 217-224.

Erickson, M.T. Reading disability in relation to perform-
ance on neurological tests for minimal brain dysfunc-
tion. Developmental Medicine and Child Neurology,
1977, .1T,-775.

Evans, J.R. Evoked potentials and learning disabilities.
In L. Tarnopol and M. Tarnopol (Eds.), Brain function
and reading disabilities. Baltimore: University
Park Press, 1977, 77-109.

Fisher, D.F. and Frankfurter, A. Normal and disabled
readers can locate and identify letters: Where's
the perceptual deficit? Journal of Reading Behavior,
1977, IX, 31-43.

Fletcher, J.M. Developmental changes in linguistic corre-
lates of reading disability: An evaluation of a
theory. Unpublished Doctoral Dissertation, University
of Florida, 1978.

Fletcher, J.M. and Satz, P. Developmental changes asso-
ciated with reading disability: A multivariate test
of a theory. Submitted for publication, 1978a.

Fletcher, J.M. and Satz, P. Some problems with the unitary
deficit hypothesis. Submitted for publication, 1978b.

Gibson, E. and Levine, H. The Psychology of reading.
Cambridge, Massachusetts: MIT Press, 1975.

Goldman, P. An alternative to developmental plasticity:
Heterology of CNS structures in infants and adults.
In D.G. Stein, J.J. Rosen and N. Butters (Eds.),
Plasticity and recovery of function in the central
nervous system. New York: Academic Press, 1974,
T49-1747

Gottlieb, G. The roles of experience in the development
of behavior and the nervous system. In G. Gottlieb
(Ed.), Neural and behavioral specificity. New York:
Academic Press, 1978, 25-54.

Hale, G. On the use of ANOVA in developmental research.
Child Development, 1977, 48, 1101-1106.







Hallahan, D.P. Comparative research studies on the psycho-
logical characteristics of learning disabled children.
In W.M. Cruickshank and D.P. Hallahan (Eds.), Per-
ceptual and learning disabilities in children, Volume
I. New York: Grune and Stratton, 1975, 29-60.

Huelsman, C.B. The WISC subtest syndrome for disabled
readers. Perceptual and Motor Skills, 1970, 30,
535-550.

Jastak, J. and Jastak, S. The Wide Range Achievement Test.
Wilmington, Delaware: Guidance Associates of Delaware,
Incorporated, 1965.

Kass, C.E. Identification of learning disability (dyssym-
bolia). Journal of Learning Disabilities, 1977, 10,
425-432.

Kershner, J. Cerebral dominance in disabled readers, good
readers and gifted children: Search for a valid model.
Child Development, 1977, 48, 61-67.

Kinsbourne, M. Models of learning disability: Their
relevance to remediation. Canadian Medical Associa-
tion Journal, 1975, 113, 1066--1068.

Klasen, E. The syndrome of specific dyslexia. Baltimore:
University Park Press, 1972.

Lenneberg, E.H. Biological foundations of language. New
York: John Wiley and Sons, 1967.

Luria, A.R. The working brain: An introduction to neuro-
psychology. London: Penguin Press, 1973.

Lyle, J.G. Reading retardation and reversal tendency: A
factorial study. Child Development, 1969, 40, 833-843.

Makita, K. The rarity of reading disability in Japanese
children. American Journal of Orthopsychiatry, 1968,
38, 599-614.

Mattis, S., French, J.H. and Rapin, I. Dyslexia in chil-
dren and young adults: Three independent neuropsycho-
logical syndromes. Developmental Medicine and Child
Neurology, 1975, 17, 150-167.

Milberg, W. and Whitman, R.D. Selective attention in good
and poor readers: Hope for the left hemisphere. Paper
presented at the International Neuropsychology Society
Convention, Minneapolis, Minnesota, February 1978.

Money, J. Reading disability. Baltimore: The Johns Hop-
kins Press, 1962.








Money, J. On learning and not learning to read. In J.
Money (Ed.), The disabled reader: Education of the
dyslexic child. Baltimore: The Johns Hopkins Press,
1966, 21-40.

Money, J. Developmental dyslexia. In P.J. Vinken and G.W.
Bruyn (Eds.), Handbook of clinical neurology. New
York: Wiley Interscience, 1969, 377-391.

Muehl, S. and DiNello, M. Early first grade skills related
to subsequent reading performance: A seven year follow
up study. Journal of Reading Behavior, 1976, VIII,
67-81.

Natchez, J. Children with learning problems. New York:
Basic Books, 1968.

Orton, S.P. Specific reading disability strephosymbolia.
Journal of the American Medical Association, 1928, 90,
10T95-1099.

Park, C.E. and Linden, J.D. The etiology of reading dis-
abilities: An historical perspective. Journal of
Learning Disabilities, 1968, 1, 38-50.

Peters, J.E., Dykman, R.A. and Romaine, J.S. A special
neurological examination of children with learning
disabilities. Developmental Medicine and Child
Neurology, 1975, 17, 63-78.

Rabinovitch, R.D. and Ingram, W. Neuropsychiatric con-
siderations in reading retardation. In G. Natchez
(Ed.), Children with reading problems. New York:
Basic Books, 1968, 144-150.

Rosenthal, J.H. The neuropsychopathology of written lan-
guage. Chicago: Nelson Hall, 1977.

Ross, A.O. Psychological aspects of learning disabilities
and reading disorders. New York: McGraw-Hill, 1976.

Rourke, B.P. Reading retardation in children: Develop-
mental lag or deficit. In R.M. Knights and D.J. Bakker
(Eds.), The Neuropsychology of learning disorders.
Baltimore: University Park Press, 1976, 125-138.

Rutter, M. and Yule, W. The concept of specific reading
retardation. Journal of Child Psychology and Psy-
chiatry, 1975, 16, 181-197.

Sabatino, D.A. The information processing behaviors asso-
ciated with learning disabilities. Journal of Learn-
ing Disabilities, 1968, 1, 6-16.








Sabatino, D.A. and Ysseldyke, J. Effects of extraneous
background on visual-perceptual performance of readers
and non-readers. Perceptual and Motor Skills, 1972,
35, 323-328.

Satz, P. Cerebral dominance and reading disability: An
old problem revisited. In R.M. Knights and D.J. Bakker
(Eds.), The Neuropsychology of learning disorders.
Baltimore: University Park Press, 1976, 273-296.

Satz, P. Specific learning disabilities: A brief state
of the art appraisal. Publication of the Research
Task Panel. President's Commission on Mental Health,
1977.

Satz, P. and Sparrow, S. Specific developmental dyslexia:
A theoretical formulation. In D.J. Bakker and P. Satz
(Eds.), Specific reading disability. Rotterdam:
Rotterdam University Press, 1970, 41-60.

Satz, P., Rardin, D. and Ross, J. An evaluation of a
theory of specific developmental dyslexia. Child
Development, 1971, 42, 2009-2021.

Satz, P. and Friel, J. Some predictive antecedents of
specific learning disability: A preliminary one-
year follow up. In P. Satz and J. Ross (Eds.), The
disabled learner: Early detection and intervention.
Rotterdam, The Netherlands: Rotterdam University
Press, 1973, 79-98.

Satz, P. and Van Nostrand, G. Developmental dyslexia:
An evaluation of a theory. In P. Satz and J. Ross
(Eds.), The disabled learner: Early detection and
intervention. Rotterdam, The Netherlands: Rotterdam
University Press, 1973, 121-148.

Satz, P., Friel, J. and Rudegeair, F. Some predictive
antecedents of specific reading disability: A two-,
three- and four-year follow-up. In J.T. Guthrie (Ed.),
Aspects of reading acquisition. Baltimore: The Johns
Hopkins Press, 1976, 111-140.

Satz, P., Taylor, H.G., Friel, J. and Fletcher, J. Some
developmental and predictive precursors of reading
disabilities: A six year follow-up. In D. Pearl
and A. Benton (Eds.), Dyslexia: An appraisal of
current research. In preparation, 1977.

Schain, R.J. Neurology of childhood learning disorders.
Baltimore: William and Wilkins, 1977.

Searle, S. Linear models. Baltimore: John Wiley and
Sons, 1971.








Sheer, D.E. Focused arousal and 40 Hz EEG. In R. Knights
and D.J. Bakker (Eds.), The Neuropsychology of
learning disorders. Baltimore: University Park
Press, 1976, 71-88.

Sobotka, K.R., Black, F.W., Hill, S.D. and Porter, R.J.
Some psychological correlates of developmental
dyslexia. Journal of Learning Disabilities, 1977,
10(6), 42-48.

Springer, S. and Eisenson, J. Hemispheric specialization
for speech in language disordered children. Neuro-
psychologia, 1977, 15, 287-293.

Strete, C., Stamm, J.S., Kreder, S.V. and Lovrich, D.
Visual perception and memory in children with learning
disabilities. Submitted for publication, 1978.

Symmes, J.S. and Rapoport, J.L. Unexpected reading failure.
American Journal of Orthopsychiatry, 1972, 42, 82-91.

Tarver, S., Hallahan, D.P., Cohen, S.B. and Kauffman, J.M.
The development of visual selective attention and
verbal rehearsal in learning disabled boys. Journal
of Learning Disabilities, 1977, 10(8), 26-35.

Taylor, G.H., Satz, P. and Friel, J. Developmental dys-
lexia in relation to other childhood reading dis-
orders: Significance and clinical utility. Submitted
for publication, 1978.

Thompson, L.J. Learning disabilities: An overview.
American Journal of Psychiatry, 1973, 130(4), 393-399.

Torgeson, J. Problems and perspectives in the study of
learning disabilities. In E.14. Hetherington (Ed.),
Review of child development research, volume 5.
Chicago: University of Chicago Press, 1975, 385-440.

Usprich, C. The study of dyslexia: Two nascent trends
and a neuropsychological model. Bulletin of the Orton
Society, 1976, XXVI, 34-48.

Vellutino, F.R. Alternative conceptualizations of dys-
lexia: Evidence in support of a verbal-deficit
hypothesis. Harvard Educational Review, 1977, 47,
334-354.

Vellutino, F.R., Steger, J.A. and Kandel, G. Reading
disability: An investigation of the perceptual
deficit hypothesis. Cortex, 1972, 8, 106-118.







Vellutino, F.R., Smith, H., Steger, J.A. and Kaman, M.
Reading disability: Age differences and the per-
ceptual deficit hypothesis. Child Development, 1975,
45, 487-493.

Vellutino, F.R., Bently, W.L. and Phillips, F. Inter-
versus intra-hemispheric learning in dyslexic and
normal readers. Developmental Medicine and Child
Neurology, 1978, 20, 71-80.

Weber, D.P. Sex differences in mental abilities, hemis-
pheric lateralization and rate of physical growth at
adolescence. Developmental Psychology, 1978, 13,
29-38.

Wechsler, D. Wechsler Intelligence Scale for Children.
New York: The Psychological Corporation, 1949.

Wechsler, D. Wechsler Preschool and Primary Scale of
Intelligence. New York: The Psychological Corpor-
ation, 1967.

Weiner, M. and Cromer, W. Reading and reading difficulty:
A conceptual analysis. Harvard Educational Review,
1967, 37, 620-643.

Winer, B.J. Statistical principles in experimental design
Second edition. New York: McGraw-Hill, 1971.

Witelson, S. Neural and cognitive correlates of develop-
mental dyslexia: Age and sex differences. In C.
Shagass, S. Gershon and A.J. Friedhoff (Eds.),
Psychopathology and brain dysfunction. New York:
Raven Press, 1977, 15-49.

Zangwill, O.L. Cerebral dominance and its relation to
psychological function. London: Oliver and Boyd,


Zangwill, O.L. Dyslexia in relation to cerebral dominance.
In J. Money (Ed.), Reading disability. Baltimore:
The Johns Hopkins Press, 1962, T03-1E 4.

Zigler, E. Developmental versus difference theories of
mental retardation and the problem of motivation.
American Journal of Mental Deficiency, 1969, 73,
5-3--556.













BIOGRAPHICAL SKETCH

Eileen B. Fennell was born in Massillon, Ohio on

August 5, 1942. When she was two, her family returned to

New York City where she attended public and parochial

schools. Graduating from high school in 1960, she moved

to Gainesville, Florida and attended the University of

Florida on an academic scholarship. She graduated with

Honors in Psychology in 1964, having also been elected to

Phi Beta Kappa. Marrying her husband Bob in 1964, she

was employed by Dr. Paul Satz in the Neuropsychology La-

boratory of the Department of Clinical Psychology and

began a continuing professional interest in human behavior.

During the next six years, her two children were born

and her husband completed his medical education. Follow-

ing a two-year hiatus as a "Navy-wife" in Memphis, she

returned to Gainesville with her family in 1972. While

her husband completed post-doctoral training in pediatric

nephrology, she again found employment as a research assist-

ant in the Departments of Psychiatry and Clinical Psychol-

ogy. Before deciding to enter graduate school in 1974,

she was employed as a Project Coordinator for the Florida

Regional Medical Program in the Health Systems Research

Division of the University. The Master's degree was re-

ceived in 1975 under the chairmanship of Paul Satz. The




81


Clinical Internship and Doctoral Program was completed in

August, 1978. Upon graduation, she expects to continue

a university affiliation. She is a member of a number

of scientific and professional organizations.








I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate in scope and quality,
as a dissertation for the Degree of Doctor of Philosophy.




Paul Satz, Chairman
Professor of Clinical Psychology


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate in scope and quality,
as a dissertation for the Degree of Doctor of Philosophy.




Jacquelin R. Goldman, Cochairman
Professor of Clinical Psychology


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate in scope and Quality,
as a dissertation for the Degree of Doctor of Philosophy.


/ //"
Hugh C Davi
Professor of Clinical Psychology


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate in scope and quality,
as a dissertation for the Degree of Doctor of Philosophy.


If /
,Kenneth M. Heilman
:Professor of Medicine








I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate in scope and quality,
as a dissertation for the Degree of Doctor of Philosophy.


/ '/ --" ,
W. Keith Berg
Associate Professor of Psychology


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate in scope and quality,
as a dissertation for the Degree of Doctor of Philosophy.




Randy L. Garter
Assistant Professor of Statistics



This dissertation was submitted to the Graduate Faculty
of the Department of Psychology in the College of Arts and
Sciences and to the Graduate Council, and was accepted as
partial fulfillment of the requirements for the degree of
Doctor of Philosophy.

August 1978


Dean, Graduate School








































UNIVERSITY OF FLORIDA
I 12il 1 5I I 4III IIIII i iII I11111111111 111111
3 1262 08553 9434