Processing functions of very low birthweight children at eight years of age

MISSING IMAGE

Material Information

Title:
Processing functions of very low birthweight children at eight years of age
Physical Description:
viii, 121 leaves, : ; 28 cm.
Language:
English
Creator:
Davidson, Debra B
Publication Date:

Subjects

Subjects / Keywords:
Birth weight, Low   ( lcsh )
Cognition in children   ( lcsh )
Learning disabilities   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1990.
Bibliography:
Includes bibliographical references (leaves 112-120).
Statement of Responsibility:
by Debra B. Davidson.
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:
oclc - 23296055
ocm23296055
System ID:
AA00011195:00001

Full Text










PROCESSING FUNCTIONS OF VERY LOW BIRTHWEIGHT
CHILDREN AT EIGHT YEARS OF AGE










By

DEBRA B. DAVIDSON


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


UNIVERSITY OF FLORIDA


1990


*UNjIVERTY OF FLORIDA LISRA7S





















Copyright 1990

by

Debra B. Davidson















ACKNOWLEDGMENTS

I would like to thank all of the parents who allowed

their children to participate in this study and all of the

children who participated. Thanks are also extended to Dr.

Deborah Goldberg and the staff at Sacred Heart Hospital for

their assistance in locating children, providing a testing

location, and gathering literature.

I also wish to thank Dr. Jeff Sugarman and Dr. Jack

Naglieri for their help in securing a copy of the Cognitive

Assessment System. Thanks are also extended to Dr. Jack

McAfee and Dr. Charles Dziuban for their assistance with

data analysis. Special thanks go to Dr. Mary K. Dykes for

her expert advice and continuous encouragement.

Finally, I want to express my gratitude to my parents,

Adrian and Gloria Bruininks, for the sacrifices they made

to provide me with a college education, for their support

and encouragement, and for teaching me the importance of

perseverance. And most importantly, I want to thank my

husband and best friend, Jim, for his endless patience,

love, and support and for all of the sacrifices he made in

order for me to pursue a doctorate degree.

iii















TABLE OF CONTENTS


PAGE


ACKNOWLEDGMENTS..................................... .. iii

ABSTRACT ............................................ vi

CHAPTERS

I INTRODUCTION..................................... 1

Theoretical Framework.......................... 2
Statement of the Problem..................... 4
Purpose of the Study......................... 5
Assessment Model............................. 6
Research Questions........................... 7
Delimitations of the Study.................... 7
Limitations of the Study....................... 8
Definition of Terms .......................... 7
Summary...................................... 10

II REVIEW OF RELATED LITERATURE.................... 13

Selection of Relevant Literature............. 13
Overview..................................... 15
Developmental Outcomes of VLBW Children...... 17
Synactive Theory of Development................ 28
Physiological Brain Development in VLBW
Children..................................... 32
Educational Needs of VLBW Children........... 35
The PASS Model of Assessment................... 38
Summary..................................... .. 51

III METHODOLOGY.................................. 55

Research Hypotheses.......................... 56
Subjects.................................... .. 56












Research Instrument.......... ................. 66
Data Analysis................................ .. 77
Methodological Limitations ................... 78

IV RESULTS...................................... 83

Experimental Results ........................... 83
Summary...................................... 88

V DISCUSSION, CONCLUSIONS, AND
RECOMMENDATIONS............................... 89

Generalizability Limitations.................. 89
Evaluation of Hypotheses ..................... 93
Conclusions.................................. .. 95
Implications................................. .. 97
Recommendations for Future Research............ 101
Summary...................................... .. 103

APPENDICES

A INTRODUCTORY LETTER ............ .............. 106

B DEMOGRAPHIC INFORMATION FORM................... 107

C PERMISSION FORM............................... 108

D APPOINTMENT FORM ............................. 109

E CAS RAW DATA................................... 110

REFERENCES .......................................... 112

BIOGRAPHICAL SKETCH ............ ................... .. 121












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

PROCESSING FUNCTIONS OF VERY LOW BIRTHWEIGHT CHILDREN
AT EIGHT YEARS OF AGE

By

Debra B. Davidson

August 1990

Chairman: Dr. Mary K. Dykes
Major Department: Counselor Education


Advances in medical technology have resulted in growing

numbers of very low birthweight (<1501 grams) babies

surviving and reaching school age. There is increasing

concern that these children may be at risk for later

learning problems. Although researchers have found a higher

rate of learning disabilities in very low birthweight (VLBW)

children than in their full-term peers, few researchers have

examined the cognitive processing functions of VLBW

children.

According to synactive developmental theory, babies

born prematurely are placed in environments to which they

are poorly matched. The result of this discrepant organism-

environmental fit is that premature babies often develop

vi












faulty adaptation patterns that may cause physiological

changes within the developing brain. These changes can lead

to later information processing difficulties. Few

researchers have followed this population longitudinally to

determine whether processing problems exist in VLBW

children.

The purpose of this study was to examine cognitive

processing functions of 8-year-old children born with very

low birthweights. The processing functions examined were

those proposed by Luria and later described by Naglieri and

Das; specifically, organization, simultaneous processing,

successive processing, and attention. These processes were

examined using the Cognitive Assessment System. The

Cognitive Assessment System was individually administered to

51 children, 21 VLBW children and 30 full-term children.

Very low birthweight and full-term children were similar

demographically on variables of age, race, gender, and

socioeconomic status.

The four area scores of the Cognitive Assessment System

were compared between VLBW and full-term children using a

multiple analysis of covariance. Covariates included

mother's age and educational level at the time of birth and

child's age at the time of testing. Results of the analysis

revealed significantly lower scores for the VLBW children in

vii










the areas of attention, simultaneous processing, and

successive processing. The most significant difference

emerged in the area of attention. No difference was found

between the two groups in the area of organization. Results

of this study enabled to investigator to support synactive

developmental theory and conclude that VLBW children are at

risk for later cognitive processing difficulties.


viii















CHAPTER 1

INTRODUCTION

The developmental outcome of very low birthweight

(VLBW) babies has been a topic of interest to educational

and medical researchers over the past two decades. The

advent of intensive care nurseries and advances in medical

technology have resulted in growing numbers of VLBW

survivors and intensified interest in their developmental

outcomes. There has been increasing concern that these

babies may be at risk for later developmental difficulties.

Recent studies have allowed for the conclusion that

very low birthweight and full-term babies have similar IQs

throughout early and later childhood (Als, Duffy, &

McAnulty, 1988; Bennett, Robinson, & Sells, 1982; Field,

Dempsey, & Shuman, 1982; Greenberg & Crnic, 1988; Kitchen et

al., 1980; Marlow, D'Souza, & Chiswick, 1987; Nickel,

Bennett, & Lamson, 1982). However, subtle differences

between these two groups have been found to exist that may

be indicative of neurological dysfunction (Als et al., 1988;

Hertzig, 1981; Hunt, Cooper, & Tooley, 1988; Sigman, 1982).

1












Children born with low birthweights have demonstrated a

higher incidence of attentional deficits (Als et al., 1988;

Blennow, Pleven, Lindroth, & Johansson, 1986; Hunt et al.,

1988; Hunt, Tooley, & Harvin, 1982), visual-motor

integration problems (Blennow et al., 1986; Francis-Williams

& Davies, 1974; Hunt et al., 1982, 1988; Siegel, 1983), and

motor deficits (Blennow et al., 1986; Greenberg & Crnic,

1988; Mazer, Piper, & Ramsay, 1988) than full-term

counterparts. Very Low Birthweight newborns have been

determined to have a higher risk for organizational

difficulties in infancy (Als et al., 1988). Neurological

"soft-signs" have been identified more often in VLBW

children than in their full-term peers (Driscoll et al.,

1982; Francis-Williams & Davies, 1974; Hertzig, 1981;

McCormick, 1985) as have reading problems (Francis-Williams

& Davies, 1974; Nickel et al., 1982).

Theoretical Framework

Possible differences between children born of very low

birthweights and those born at term can be interpreted

within the framework of synactive developmental theory (Als,

1978, 1982, 1985, 1986; Als et al., 1988; Linton, 1986).

According to this theory, environmental factors influence

brain development through an interplay of sensory

information and experience. Between 26 and 40 weeks'

gestation, areas of the brain, especially the association











areas, are rapidly maturing and differentiating. Results of

animal studies have led researchers to conclude that there

are sensitive periods of brain development during which

specific environmental input is necessary for development to

proceed normally (Hubel, Wiesel, & LeVay, 1977). Preterm

birth may have a major impact on the developing brain

because the experiences of premature babies are not those

that are biologically expected (i.e., those that would

normally occur).

The result of preterm birth is a discrepant organism-

environment fit. Premature babies are placed in

environments to which they are poorly matched and often have

problems adapting to these new environments. Consequently,

such babies are at risk for developing faulty adaptation

patterns. In an attempt to protect themselves from

offensive environmental input, preterm babies may develop

defense behaviors such as gaze aversion, closing eyes, or

flaccidity. Preterm babies may not be able to attend or

respond to environmental input without causing stress to

autonomic, motor, and physiological systems. The

experiences of such babies are often distorted; therefore,

such babies may develop defense behaviors and disorganized

approaches for dealing with environmental input. According

to synactive theory, the resulting distortions and

disorganizations can cause changes in the developing brain











and place the premature baby at risk for later processing

and organizational difficulties.

Statement of the Problem

Improved neonatal care and decreasing mortality among

VLBW populations have resulted in increasing numbers of

these children reaching school age (Levene & Dubowitz, 1982;

Saigal, Rosenbaum, Stoskopf, & Sinclair, 1984). Although

the decrease in mortality has not been accompanied by an

increased incidence of severely handicapping conditions

among VLBW children (Chamberlin, 1987; Drillien, Thomson, &

Burgoyne, 1980), such children continue to be at risk for

learning difficulties despite IQs that are typically in the

average range (Hunt et al., 1982; 1988). The increased risk

status of VLBW children for later learning difficulties has

been shown to exist even when socioeconomic status is

accounted for, further suggesting the presence of subtle

neurological differences in VLBW children (Hertzig, 1981;

Lasky et al., 1987; O'Reilly, O'Reilly, & Furono, 1986;

Sigman, 1982).

The extent of differences between VLBW children and

children born at term has not been observed until school age

(Hirata et al., 1983; Hunt et al., 1982; Hunt et al., 1988;

Pederson, Evans, Chance, Bento, & Fox, 1988) and few studies

have been devoted to the longitudinal study of VLBW

children. Although some researchers have found VLBW











children to be at higher risk for learning disabilities

(Nickel et al., 1982; Hunt et al., 1988), knowledge of

cognitive processing styles, organizational abilities, and

attentional processes in these children is lacking. Only

very recently have researchers identified attentional and

organizational difficulties in preterm neonates that

continue to exist during infancy (Als et al., 1988). The

degree to which these difficulties continue to exist into

later childhood is unknown.

There is a need to determine why VLBW children are at

risk for later learning difficulties. Perhaps these

children have subtle processing deficiencies similar to

those that have been observed during the neonatal period;

specifically, difficulty organizing and processing

environmental input. Knowledge of information processing,

organizational, and attentional processes of VLBW children

may facilitate educational programming for these children

and perhaps ultimately reduce their high rate of educational

difficulties. Additionally, examination of these processes

in VLBW children at school age is necessary to support the

synactive theory of development in terms of long range,

neurologically-based effects of premature birth.

Purpose of the Study

The purpose of this study was to investigate

information processing, organizational, and attentional












processes in 8-year-old children who were born with very low

birthweights. Global intellectual scores have generally

failed to differentiate between VLBW and full-term

children. Sole use of such scores has been criticized

because of their tendency to obscure subtle or discrete

abilities (Lezak, 1988; Naglieri & Das, 1988). In this

study, the subtle processes that may contribute to learning

difficulties in VLBW children were investigated.

Assessment Model

Luria (1980) described three functions of the brain:

arousal, coding, and planning/organization. Problems with

organization and arousal have been found to exist at higher

rates in VLBW neonates than in those born at term (Als,

1985; 1986; Als et al., 1988; Ruff, 1986). Additionally,

organizational difficulties have been found to exist at

higher rates in VLBW children during infancy (Als et al.,

1988). Information about coding processes in VLBW children

is currently lacking, but synactive theory would suggest

that these children have deficient coding processes relative

to full-term peers.

Most traditional intelligence tests do not adequately

measure planning and arousal, and few can successfully

measure coding processes (Naglieri & Das, 1988). The PASS

(Planning-Arousal-Successive-Simultaneous) model of

assessment proposed by Das (1972) and recently described by












Naglieri and Das (1988) is an effective measurement system

for examining these processes. The Cognitive Assessment

System (CAS), based on the PASS model and developed by

Naglieri and Das (1987), was used in this study to examine

organizational, attentional, and coding processes in 8-year-

old children who were born of VLBW. Eight-year-old children

were chosen because of the need for longitudinal follow-up

and the tendency for subtle learning problems to first

manifest themselves during this time (Pederson et al.,

1988).

Research Questions

The following research questions are examined in this

study in relation to children at 8 years of age:

1) Do VLBW children have lower successive

processing scores than full-term peers?

2) Do VLBW children have lower simultaneous

processing scores than full-term peers?

3) Do VLBW children have lower organizational

processing scores than full-term peers?

4) Do VLBW children have lower attentional

processing scores than full-term peers?

Delimitations of the Study

This study is delimited geographically to the northwest

Florida area referred to as "the Panhandle"and nearby

states. Subjects were obtained from northwest areas of












Florida ranging geographically from Pensacola, a medium-

sized city located in Escambia county, to Tallahassee, a

medium-sized city located in Leon County. Additionally,

subjects were obtained from various small towns and rural

areas of southern Alabama and southern Mississippi.

Subjects for the study were attending public schools and

most subjects were 8 years of age. Two subjects, one full-

term and one VLBW, were 9 years, 2 months of age. No

procedures for randomizing or matching were applied.

However, attempts were made to obtain similar samples of

VLBW and full-term children on the demographic variables of

age, gender, race, and socioeconomic status. Socioeconomic

status was defined by the mother's need for federal

assistance in paying hospital bills at the time of the

child's birth (i.e. medicaid). The independent variables

examined in this study were those measured by the Cognitive

Assessment System; specifically, organization, simultaneous

processing, successive processing, and attention.

Limitations of the Study

Since this study included only 8- and 9-year-old

children, the findings should not be generalized to children

of other ages. Additionally, because this study contained

fewer VLBW children of lower socioeconomic status than has

been reported nationally, caution should be used when

generalizing results of this study to the entire population











of VLBW children. Caution should also be used when

generalizing results of this study to VLBW children outside

of the Florida panhandle area, particularly to those VLBW

children residing in large metropolitan areas.

Another limitation of this study involved the

instrumentation. The Cognitive Assessment System was not

normed and there was a lack of reliability and validity

information on the instrument at the time of the study. The

CAS was chosen over other existing instruments because it

provided a better measure of the variables of interest.

However, the lack of normative data and questions about the

reliability and validity of the CAS were limitations in this

study.

Relatedly, the fact that one examiner tested all of the

subjects was a limitation. The possibility of experimenter

bias existed, but was minimized by the fact that the CAS is

an objective instrument. Additionally, subjects were

scheduled several weeks in advance so as to maintain some

degree of anonymity regarding group membership.

Definition of Terms

A number of terms are used that require further

definition.

Arousal. Arousal is a unit identified in Luria's

(1973) model, it is the prerequisite for all mental

activity. Arousal allows an individual to direct attention











toward relevant stimuli, direct attention away from

irrelevant stimuli, and divide attention between activities

without decreased efficiency.

Coding. Coding is a unit identified in Luria's (1973)

model, coding is involved in receiving, processing, and

storing information. Luria (1966) identified two coding

processes--successive and simultaneous.

Cognitive Assessment System. The Cognitive Assessment

System is an instrument developed by Naglieri and Das (1987)

to examine four areas of cognitive functioning: attention,

successive processing, simultaneous processing, and

organization.

Full-Term. Full-term children refer to those children

born with birthweights of at least 2500 grams.

PASS Model of Assessment. The PASS model of assessment

is an information-integration model first proposed by Das

(1972) that explains cognitive functioning in terms of three

brain units--arousal, coding, and planning.

Planning. Planning is a unit identified in Luria's

model (1973) that is involved in formulating plans of

action, regulating behavior so that it corresponds to plans,

comparing results of actions to intentions, and correcting

errors.












Simultaneous Processing. Simultaneous processing is

one of the coding processes described by Luria (1973) in

which each element of a stimulus is related to all other

elements. Simultaneous activities are said to be surveyable

because all aspects of the stimulus are accessible during

inspection.

Successive Processing. Successive processing is one of

the coding processes described by Luria (1973) in which each

element of a stimulus is related only to adjacent elements.

Therefore, stimuli are integrated into series and

synthesized into a chain-like progression of elements.

Synactive Developmental Theory. Synactive

developmental theory is the theory proposed by Als (1978) to

explain the difficulties that preterm babies often

experience when exposed to environmental stimulation. The

organism-environmental mismatch that exists with these

babies may lead to long range problems with information

processing according to synactive theory.

Very Low Birthweight. Very low birthweight children

refer to those children born with birthweights less than

1501 grams.

Summary

There is a need to examine long-term cognitive

processing functions of VLBW children. Although VLBW

children have been found to have cognitive processing












deficits as infants and young children, studies that examine

processing functions of older VLBW children are lacking. It

was the intent of this study to provide information about

cognitive processing functions of VLBW children at 8 years

of age. Specifically, organization, simultaneous

processing, successive processing, and attention were

examined. The results of this study have implications for

educators and others who work with VLBW babies and children.

A review of related literature is presented in Chapter

II. Chapter III presents a description of methodology.

Results are then presented, discussed, and related to

previous research findings in Chapter IV. Summary of

findings, conclusions, and recommendations for future

research are presented in Chapter V.















CHAPTER II

REVIEW OF LITERATURE

In Chapter II, an analysis of the professional

literature involving outcomes of VLBW children, synactive

developmental theory, educational needs of VLBW children,

and the Cognitive Assessment System are presented. The

chapter is divided into eight major sections. Selection

criteria for the literature that was reviewed and an

overview of the importance of longitudinal follow-up of VLBW

children are presented. Included in the overview is a

discussion of mortality and morbidity of VLBW children. In

the subsequent sections, literature related to

developmental outcomes, physiological development, and

educational needs of VLBW children, synactive developmental

theory, and the Cognitive Assessment System is reviewed.

The chapter concludes with a summary and implications of

previous research as it relates to the present study.

Selection of Relevant Literature

An initial step in the review of the literature was

that of determining the criteria for the inclusion of

references. All relevant studies completed in the last 10

years (1980-1990) were examined. In addition, any notable

13












research cited in the literature earlier than the 1980 year

time period was also considered.

Professional literature concerning outcomes of VLBW

children and the Cognitive Assessment System was required to

meet the following criteria to be included in the review:

1. The subjects and the settings in which the

experimentation took place had to be thoroughly

described.

2. The treatment conditions and experimental

procedures were detailed enough to permit

replication.

3. The experimental design and data analysis

procedures were presented without significant

losses of information.

4. The interpretations of the experimenter had to be

consistent with the results displayed.

In order to exhaustively review the literature related

to outcomes of VLBW children, synactive developmental

theory, and the Cognitive Assessment System, the following

sources were used for the literature review: Dissertation

Abstracts International, Educational Resources Information

Clearinghouse (ERIC), Psychological Abstracts, and Current

Index to Journals in Education (CIJE). References initially

selected were located through the libraries at the

University of Florida, the University of West Florida, and












Sacred Heart Hospital, through the interlibrary loan system,

or through other professionals in the field. Descriptors

used in this literature search included VLBW, processing,

premature, coding, attention, and organization.

The references that were selected were critically

reviewed and those that described empirical investigations

were chosen based on the investigator's judgment that the

references presented a clear description of subject

selection, methodology, and results. Professional

literature other than empirical investigations were also

included if, in the author's judgment, the information that

was included provided a valuable contribution to the

knowledge base about or an understanding of VLBW children,

synactive developmental theory, or the Cognitive Assessment

System.

Overview

Very low birthweight (VLBW) infants account for

approximately 1.5% of all births (Peterson, 1988; Slater,

Naqvi, Andrew, & Haynes, 1987), but constitute the largest

group of newborns at risk for later handicaps (Levene &

Dubowitz, 1982). A meta-analysis of nine studies revealed

that there was a 58% survival rate for VLBW infants during

the 1970s (Levene & Dubowitz, 1982). Survival rates of

infants weighing less than 1000 grams at birth have doubled

since the middle 1970s (Saigal et al., 1984). The large












number of VLBW survivors has led to increasing concern about

the developmental outcomes for this population.

Although the decreasing mortality rate among VLBW

children has not been accompanied by a proportionate

increase in major handicapping conditions (Chamberlin, 1987;

Drillien et al., 1980; Hunt et al., 1982; Kraybill, Kennedy,

Teplin, & Campbell, 1984; Saigal et al., 1984; Stewart,

Reynolds, & Lipscomb, 1981), these infants contribute

disproportionately to the number of children with

neurological and developmental delays (Hunt et al., 1988;

Siegel, 1983; Slater et al., 1987). VLBW babies are 3 times

as likely as full-term babies to develop later

neurodevelopmental handicaps (McCormick, 1985).

The differences between VLBW and full-term children are

subtle and difficult to identify. However, as many as one-

third of VLBW children encounter academic or social

adjustment problems at school age (Blennow et al., 1986). A

follow-up study of 25 children who weighed less than 1000

grams at birth, revealed that 64% had been or were currently

receiving special education services at 6 to 16 years of age

(Nickel et al., 1982). Only 28% of VLBW children were rated

by teachers as achieving on or above grade level. Hunt et

al., (1988) found that 16.7% of VLBW children had learning

disabilities at age 8. It appears from results of these

studies that underlying differences exist between children











born with very low birthweights and those born at term.

These differences may have been masked by global

intellectual measures often used in research (Aylward, 1988;

Pederson et al., 1988). However, subtle problems have been

found to exist at higher rates in VLBW children than in

children born at term.

Developmental Outcomes of VLBW Children

Studies examining intellectual functioning of VLBW

children have generally found them to be within average

ranges on global measures. Although some researchers have

found statistically different intelligence quotients (IQs)

in these two populations, differences have generally been

small. Very low birthweight children as a group tend to

have average IQs. A general problem with much of the

research in the area of developmental outcomes of VLBW

children concerns the definition of VLBW. Although VLBW has

traditionally been defined as weighing less than 1501 grams

at birth, many researchers have included children weighing

more than 1501 grams in their samples while others have

limited their samples to those weighing less than 1000 grams

at birth. Birthweights of samples of VLBW children are

specified in the following literature review when they

deviate from the traditional definition of VLBW. Unless

otherwise stated, VLBW children in the studies reviewed

included only those weighing less than 1501 grams at birth.












Intellectual Functioning of VLBW Children

Kitchen and colleagues (1980), in a follow-up study of

158 VLBW survivors, found that VLBW children had IQs in the

average range at age 8, although their scores were slightly

lower than those of children in the full-term control

group. Similarly, Field, Dempsey, and Shuman (1982)

followed 56 preterm children weighing less than 1600 grams

at birth. At age 5, preterm children scored slightly lower

than full-term children on the McCarthy Scales. However,

mean scores for the preterm children were above the 50th

percentile. In a study by Nickel and colleagues (1982) of

25 children weighing less than 1360 grams at birth, full

scale intelligence quotients as determined on the Wechsler

Intelligence Scale for Children- Revised (WISC-R) were in

the average range at 6.1 to 18.7 years of age with Verbal,

Performance, and Full Scale scores of 94.1, 91.5, and 90.5,

respectively.

Short term follow-up studies have shown also that VLBW

and term-birth children have similar IQs. In a study of 16

VLBW children, Bennett et al., (1982) found that Stanford-

Binet IQs were in the average range at age 3 with a mean of

106. Bayley Mental Development indices were also average

for children in this study at the 6 to 30 month age range.

Als et al., (1988) followed 112 preterm and 48 full-term

babies. These investigators found that Bayley scores did











not differ between the two groups at 9 months of age, with

all mean scores above 100. Marlow et al. (1987) followed

654 children weighing less than 2000 grams at birth to a

median age of 3 years, 3 months. The mean Griffith Mental

Development Scale and Wechsler Preschool and Primary Scale

of Intelligence (WPPSI) overall quotients were 101.6 and

101.8, respectively. Greenberg and Crnic (1988) followed 30

preterm infants weighing less than 1801 grams at birth and

found language and cognitive development to be adequate at 2

years of age. Mean Bayley scores for the VLBW sample of

children were 99.4 and 103.8, respectively, for the mental

and motor scales. Although results of these studies allow

for the conclusion that VLBW children are comparable

intellectually to full-term peers, other investigators have

uncovered subtle differences between these two groups of

children.

Learning Disabilities in VLBW Children

VLBW children have a higher incidence of problems than

full-term children in the areas of visual-motor integration,

attentional abilities, motor skills, language, and

behavior. A follow-up study of 102 VLBW children at ages 4

to 8 revealed that 20.6% had possible learning disabilities

as evidenced by average IQs and deficits in such areas as

language comprehension and visual-motor integration (Hunt et

al., 1982).












Hunt et al. (1988) examined 108 VLBW children at age 8

and evaluated 57 of these children again at age 11 using the

Wechsler Intelligence Scale for Children-Revised (WISC-R),

Bender-Gestalt Test, and Wide Range Achievement Test

(WRAT). At age 8, 36.1% were classified as normal with no

indications of learning difficulties. Visual-motor deficits

were the most common difficulties exhibited by 21.4% of the

VLBW children. Language and performance disabilities were

each manifested by 12% of these children. Learning

disabilities, defined as a 15 point discrepancy between Full

Scale WISC-R score and one WRAT achievement subtest score,

were found in 16.7% of all children at age 8. However, it

should be noted that the children classified as learning

disabled had a mean WISC-R score of 110.3 and regression to

the mean could have accounted for some of the discrepancies

observed (Reynolds, 1984). At age 11, similar results were

obtained. Using the same criteria, approximately two-thirds

of the VLBW children were classified as having mild to

severe disabilities with 14.6 exhibiting signs of learning

disabilities.

Siegel (1983) conducted a longitudinal study of 42 full-

term children and 44 preterm children who had a mean

birthweight of 1236 grams. At 5 years of age, children were

evaluated using the Satz Battery (Fletcher & Satz, 1980).

The Satz Battery consists of the Peabody Picture Vocabulary












Test (PPVT), Recognition Discrimination Task, Beery Visual

Motor Integration Test (VMI), Alphabet Recitation, and

Finger Localization. It has been shown to be predictive of

reading problems when administered at age 5 (Fletcher &

Satz, 1980). Siegel concluded that VLBW children are at

risk for learning disabilities as evidenced by significantly

lower scores on three of the five areas of the Satz

Battery. The VLBW group of children obtained a VMI score of

89.6 while the full-term group of children obtained a score

of 102. The group of VLBW children also obtained lower

scores than the full-term group of children in the areas of

Recognition Discrimination and Alphabet Recitation.

Language comprehension as measured by the PPVT did not

differ between the two groups.

Blennow and colleagues (1986) followed 45 ventilator-

treated babies weighing less than 2500 grams at birth.

Neurological examination and the Griffith Mental Development

Scale administered at ages 6 and 7 revealed that 33% had

attentional problems and 40% had locomotor and performance

deficits. These deficits were not correlated with

birthweight or duration of ventilator treatment. Visual-

perceptual and visual-motor deficits were also apparent in

this sample. The authors concluded that approximately one-

third of low birthweight, ventilator-treated babies are at

risk for later school difficulties.












Mazer and colleagues (1988) recently followed 78

infants weighing less than 1500 grams at birth to 3 years of

age. Locomotor quotients on the Griffith Scales decreased

over time from a mean score of 116 at 6 months to 98 at 36

months. The degree to which this decreasing pattern of

motor scores continues into later childhood is unknown and

is indicative of a need for longitudinal follow-up of VLBW

children.

The above studies suggest that children born of very

low birthweights are at risk for later learning

disabilities. This risk appears to be greater for VLBW

children than it is for their full-term peers and may

involve problems in the areas of motor functioning, visual

perceptual skills, language comprehension, and academic

skills.

Early Processing Difficulties of VLBW Children

A recent study that included 160 newborns investigated

112 preterm and 48 full-term children to 9 months of age

(Als et al., 1988). At 42 weeks postconception, the

Assessment of Preterm Infant's Behavior (APIB) was

administered. The APIB was developed to measure primarily

the ". infants reactivity and threshold of

disorganization and stress to environmental input" (Als et

al., 1988, p.9). At 42 weeks postconception, preterm babies

as a group were more disorganized and had lower stress












thresholds than full-term counterparts. This

disorganization and vulnerability to stress increased with

degree of prematurity. These differences existed in the

preterm group of children even when social class was

controlled statistically.

Infants in this study were later administered the

Kangaroo Box (K-Box) paradigm at 9 months of age. The K-Box

is a plexiglass box containing a wind-up kangaroo that is

accessible only through a transparent mobile porthole latch-

door (Als et al., 1988). In attempting to retrieve the

kangaroo, the infant draws upon cognitive, motor, social,

and affective capacities. The paradigm yields information

about organizational abilities in children across several

behavioral dimensions and is believed to correspond to

parameters measured by the APIB.

At 9 months of age, preterm children as a group were

significantly poorer than full-term children in fine motor

organization, cognitive appreciation of the situation,

affective response, attentional organization, pleasure and

pride, combining social and object play, and overall

competence. Earlier preterm children also had more

difficulty with autonomic organization. Children who were

disorganized on the APIB at 42 weeks postconception

continued to be disorganized at 9 months on the K-Box

paradigm. Although Bayley scores did not differ between the












groups of preterm and full-term children at 9 months of age,

more preterm children were perseverative in their play,

distractible, and disorganized in their approach to problem

solving on the K-Box. The authors, following a synactive

theory of development, concluded that the problems exhibited

by many of the VLBW children were neurologically based and

originated shortly after birth.

O'Reilly and colleagues (1986) followed 102 infants

with birthweights less than 1500 grams to 9 months of age at

which time the Bayley Scales, Gesell Developmental

Schedules, and a range of motion test were administered.

The infant's ability to habituate to environmental input

during the neonatal period, as measured by the Brazelton

Neonatal Assessment Scale, was correlated with the mental

scale of the Bayley at 9 months of age (r= -.73).

Additionally, habituation during the neonatal period was

significantly correlated with the overall Gesell score (r=

-.52). Ability to orient to incoming stimuli on the

Brazelton during the neonatal period was also significantly

related to the Bayley motor and mental scale scores and the

Gesell gross motor scale score at 9 months of age (r= .31,

.44, and .28, respectively). Family income was not an

important predictor of outcome when variables were examined

through multiple regression analysis.











Ruff (1986) found that preterm infants had less

organized behavior than full-term infants in their approach

to novel objects. She examined 41 children, 17 of whom had

birthweights less than 1500 grams and 24 who were born full-

term, at 30 to 32 weeks adjusted age. She found that the

group of preterm children spent an average of 19.3 seconds

examining novel objects compared to 55.4 seconds of

examination by the full-term group of children. Latency

time before examination of objects also differed

significantly (p<.01) between the two groups with mean

latency times of 10.6 seconds and 21.9 seconds,

respectively, for the full-term and preterm groups of

children. Additionally, full-term children generally

approached objects in an organized manner by examining, then

mouthing, and finally banging objects. Conversely, many of

the preterm children did not display this organized

approach. Rather, their behavior was observed to be less

differentiated than that of the full-term group. Ruff

concluded that preterm children have difficulties related to

reactivity, sustained attention, and organization at

approximately 7 months of age. The degree to which

behaviors observed in these preterm children foreshadow

later attentional and organizational problems has yet to be

explored.










In a recent study, Cohen and colleagues (1988) examined

89 prematurely born children who were 8 years of age. The

22 of these children who were found to have learning

problems were compared to the 67 without apparent learning

problems. All children weighed less than 2500 grams at

birth. Learning problems were defined as having a full-

scale IQ score of 80 or above and either achievement below

the 25th percentile on standardized mathematics or reading

tests, retention or recommended retention, or special class

placement. The overall rate of learning problems was 25%

for this preterm group of children.

Although larger birthweight preterm children were

included in this study, the increased risk status of preterm

children for later school difficulties was apparent. The

authors did not report the proportion of smaller preterm

children who experienced learning problems at age 8.

Therefore, it is unknown whether smaller preterm children in

the study had learning problems at the same magnitude as

larger preterm children. Demographic data components such as

socioeconomic status were not predictive of learning

problems in this study. Neonatal self-regulation behavior

as measured by state organization (i.e. amount of active

sleep) was predictive of outcome at age 8. Because active

sleep has a distinct neurophysiological component in terms

of organization, the authors suggest that the neonate's











abilities to organize states can provide information about

possible later neurological difficulties in the areas of

learning, attention, and organization.

The results of these studies indicate that VLBW

children as a group have a higher rate of early processing

difficulties than children born at term. Difficulties have

been observed in the areas of organization, attention,

habituation, orientation, and neonatal self-regulation.

Researchers have suggested that these early processing

difficulties may be predictive of later neurological

problems.

Neurological Problems of VLBW Children

Other researchers have provided additional evidence of

neurologically based differences in VLBW children, lending

further support to Al's theory. Francis-Williams and Davies

(1974) examined visual-motor integrative functioning in 65

VLBW children at ages 5 to 12 using the Bender Gestalt

Test. Immaturity associated with neurological dysfunction

(Koppitz, 1964) was apparent in their design reproductions,

with 36 of the children scoring at least one standard

deviation below the mean. Separations, rotations, and

distortions were common in their designs.

In a study of 23 infants weighing less than 1001 grams

at birth, it was determined that a longitudinal

complications rate of 30% existed for children aged 18










months to 3 years (Driscoll et al., 1982). Neurological

deficits were apparent in 17% of these children, although

the authors were not specific about the nature of these

deficits. Hertzig (1981) conducted an 8 year follow-up

study of 66 children whose birthweights were between 1000

and 1750 grams. Neurological examination revealed that 50%

of the sample had signs of neurological dysfunction at

follow-up. Localized signs (i.e. signs associated with

central nervous system dysfunction such as pathological

reflexes) were found in 13 of the subjects. Two or more non-

focal or "soft" signs (i.e. dysfunction in speech, balance,

coordination, gait, sequential finger-thumb opposition, and

muscle tone) were found in 20 of the subjects. Dysfunction

was not associated with socioeconomic status.

Based on the above studies, it appears that differences

between VLBW and full-term children may exist at a

neurological level. These differences may emerge shortly

after birth and continue throughout life. The synactive

theory of development proposed by Als (1986) provides a

basis for understanding the process by which these

developmental differences might evolve.

Synactive Theory of Development

The central themes underlying the synactive theory of

development are consistency of environmental interactions

throughout life and predictability of behavior (Als, 1986).










Some researchers have maintained that newborns do not have

the capacities or capabilities that are necessary in later

childhood, thereby making prediction impossible except in

extreme cases (Prechtl, 1984). Discontinuity of development

is emphasized in this line of thought. Conversely, the

synactive theory holds that the manner in which an infant

interacts with the environment is predictive of later

behavior. Als et al. (1988) state that ". an organism

draws on all its capacities at all times in its

developmental progression [and] how well it does so is

consistent over time" (p.4).

There are four principles that are incorporated into

synactive developmental theory (Als et al., 1988):

(1) The first principle, species adaptedness, is

based on the thesis that an organism at any

stage of its development has become competent

at that stage through evolution (Hinde, 1970);

(2) The principle of continuous organism-

environmental interaction (Piaget,1952)

is based on the premise that development occurs

only through interaction with the environment;

(3) The underlying premise of the third principle,

orthogenesis and syncresis, is that development

proceeds toward increasing differentiation

(Bruner, 1968; Piaget, 1952); and










(4) The principle of dual antagonistic

integration is based on the thesis that an

organism constantly strives for smooth

integration by balancing approach and avoidance

behaviors (Schneirla, 1965).

A central idea of synactive theory is that at each

developmental stage, various subsystems exist side by side,

occasionally interacting, but often in a sort of ". .

holding pattern as if providing a steady substratum for one

of the system's differentiation process" (Als, 1982;

p.230). At each developmental level, adaptation occurs only

through an interplay between the environment and the primary

developmental agendum that is determined by the subsystem in

differentiation. The subsystems or developmental agenda

emerge hierarchically in the order of autonomic, motor,

state-organization, attention, and interaction. Development

is sequential in that the infant must successfully master

one level before entering the next. Movement to higher

levels must occur in the background of well-integrated

functioning at lower levels. Otherwise, the result will be

distorted defense behaviors (e.g. averting gaze or closing

eyes) caused by a discrepant organism-environment fit. A

poor fit can lead to a cycle of increasing disorganization

and distortion that may result in physiological changes in

the developing brain (Als, 1986).










As an infant moves to a higher step of differentiation,

synactive theory maintains that previous subsystems are

temporarily disorganized. However, after the new

developmental agendum has been mastered, subsystems realign

at a higher level of differentiation, once again supporting

each other. A higher level of differentiation cannot be

attained if too much stress is present. Rather, a

maladaptive realignment occurs that is costly to the

infant's development. The result in this situation is

rigid, less differentiated functioning and continued use of

maladaptive strategies for dealing with environmental

input.

Because development occurs through an interplay of

sensory information and experience, the environment can

either facilitate or hinder an infant's development (Als,

1985; Als et al., 1988). The result of this interplay is

either "species appropriate ontogenetic integration patterns

. [or] deleterious adaptation patterns" (Als,

1986, p.4). A prematurely born baby is not programmed

biologically to handle the extrauterine environment and is

easily overloaded and stressed (Lawhon & Mazer, 1988;

Linton, 1986). The premature infant's brain is overly

sensitive and lacks necessary inhibitory mechanisms (Als,

1986). These infants should have 13 to 16 additional weeks

in utero where physiological functions, diurnal rhythms,











muscle tone and movement are largely controlled by maternal

factors. Instead, they are placed in environments to which

they are poorly matched and where they are at increased risk

for developing distortions and disorganizations that may

lead to physiological changes in the developing brain.

The baby who is born prematurely is therefore at risk

for developing later processing problems according to

synactive developmental theory. The stresses that often

result from a poor organism-environmental fit may lead to

less differentiated functioning and changes in the

developing brain of the prematurely born neonate. The

result of these physiological brain changes may be long

range processing difficulties.

Physiological Brain Development in VLBW Babies

Physiological development of the brain during the last

trimester of pregnancy contributes to the premature

neonate's problems with environmental adaptation. Brain

cells, especially in the association areas, are rapidly

emerging and differentiating from 26 to 40 weeks gestation.

During the prenatal period, neurons in the brain emerge at a

rate of 250,000 per minute (Thompson,1985). Preterm birth

can have a major impact upon the developing brain.

Developmental distortions may result from active

suppression or inhibition of normal brain pathways during

sensitive periods as has been suggested by animal models











(Huble, Wiesel, & LeVay, 1977; Spinelli, Jensen, & DePrisco,

1980). These suppressions appear to be caused by endorphine

mechanisms which are highly concentrated in association

cortical areas of the brain, areas that have been associated

with attentional, learning, and behavioral problems in

school-aged children (Als, 1986). The premature neonate's

brain lacks inhibitory mechanisms believed to be associated

with differentiation of cortical association areas.

Consequently, these babies may be unable to protect

themselves from sensory input. While normal pathways are

being suppressed, current functional pathways in the

premature neonate's brain are being overactivated. The

result of both processes, according to synactive theory, is

less differentiated and less modulated behavior on the part

of the baby.

The premature neonate's difficulties in processing

information may also be influenced by myelination and

neurotransmitters. The peak myelination time occurs at

around 40 weeks gestation, the time of term birth (Volpe,

1981). Myelin speeds conduction of neural impulses and

accommodates neuronal track growth. Lack of myelin in the

premature neonate's brain could impact information

processing and perhaps result in long term changes in

processing functions of the brain.











The release of neurotransmitters is dependent upon

other regulatory systems functioning properly in the baby

and receptors for neurotransmitters are dependent upon

experience. Other regulatory systems in the premature baby

are often not functioning properly. Additionally, the

experiences of the premature neonate are often not optimal

because of the poor organism-environment fit. In summary,

". [T]he brain and sensory organs are continuously

dependent on each other for normal structural and functional

development [I]t is potentially quite dangerous to be

born before term" (Als, 1986, p.6). Researchers are just

beginning to identify some of the subtle dangers to which

Als was referring, dangers that could have an impact not

only on a child's early development, but on later learning

processes as well.

Premature birth can have an impact on the developing

brain in several ways. Physiological changes could include

inhibition of normal pathways and overactivation of

functional pathways in the premature neonate's brain.

Neurotransmitters and myelin also influence the premature

neonate's ability to process information. These











physiological factors could result in long range processing

difficulties.

Educational Needs of VLBW Children

Increasing numbers of VLBW children are now reaching

school age. While these babies constitute only

approximately 1.5% of all births, their disproportionate

contribution to the number of children with

neurodevelopmental handicaps causes them to be of concern to

educators.

The chances of a VLBW baby surviving as a healthy,

interactive infant tripled between 1960 and 1980 (Stewart,

Reynolds, & Lipscomb, 1981). Major handicapping conditions

such as cerebral palsy and mental retardation have not

increased proportionately in VLBW survivors during this

period (Drillien et al., 1980; Hunt et al., 1982). The

incidence of such handicaps among VLBW infants has remained

relatively stable at 6-8% (Stewart et al., 1981). However,

VLBW children are 3 times more likely than full-term peers

to develop later neurodevelopmental handicaps (McCormick,

1985) and one-third of VLBW children encounter academic

difficulties by 7 years of age (Blennow et al., 1986).

The increasing number of VLBW children reaching school

age and meeting with academic difficulties is a concern to

educators. It becomes an issue of even greater concern when

currently high drop-out rates are considered. Large











expenditures of money are being appropriated for drop-out

prevention programs across the country. Because VLBW

children are at risk for learning problems, they may

consequently be at risk for dropping out of school.

Although subtle learning difficulties have been identified

in VLBW children, few studies have followed these children

into elementary school. Organizational and attentional

difficulties have been identified in infants born VLBW, but

the extent to which these difficulties continue into school

years is still unknown.

There is ample evidence to support the high risk status

of VLBW children for later learning difficulties. The

question of what can be done to reduce academic failure in

these children is still unclear. Perhaps VLBW children have

different processing functions than full-term counterparts

because of, as Als suggested, underlying neurological

differences that occurred shortly after birth. Knowledge of

such processing functions could assist educators in

developing appropriate teaching strategies to enhance

learning in VLBW children. In order to plan effective

strategies for these children, there is a need to determine

whether weaknesses exist in coding processes, attentional

processes, and/or organizational processes (Naglieri, 1988).










If children born with very low birthweights are found

to have later organizational problems, this finding will not

only have an impact on delivery of educational services, but

will also have an impact on service delivery by

professionals in the medical sector as well. Intervention

strategies for promoting organization in neonates have been

outlined by several researchers (Als, 1988; Lawhon & Melzar,

1988) and will have increasing relevance for nursing staff

and early intervention specialists who work with VLBW babies

and young children. If no differences are found between

VLBW and full-term peers at 8 years of age in terms of

organizational abilities, then perhaps some of the

interventions being recommended for use in intensive care

nurseries with these babies will need to be reexamined.

Identification of processing functions of VLBW children

at school age may assist school psychologists in planning

assessments for these children. Use of instruments

sensitive to processing functions such as attentional

processes, coding processes, and organizational processes

may have increasing relevance for use with VLBW children if

they are found to differ as a group from full-term children

in these processing areas.

In this study, attempts were made to look at subtle

cognitive processes rather than global intellectual

functioning of VLBW children. Global IQs tend to mask











discrete differences in intellectual functioning,

differences that can have a profound impact on learning

ability (Aylward, 1988; Kaufman, 1988; Naglieri & Das,

1988). A broader assessment of cognitive functioning can be

obtained by examining successive, simultaneous, attentional,

and organizational processes (Nagieri & Das, 1988). The

Cognitive Assessment System (CAS) developed by Naglieri and

Das (1987) was used in this study to measure these four

processes. In addition to providing a broader measure of

cognitive processes, the CAS was developed to reflect a

theory of information processing that is neurologically

based. Therefore, performance on this instrument should

provide information about neurological aspects underlying

processing functions of VLBW children.

The PASS Model of Assessment

The PASS model of processing is an information-

integration model first proposed by Das (1972) and later

described by Naglieri and Das (1988). The Cognitive

Assessment System (CAS) developed by Naglieri and Das (1987)

follows the PASS model and is intended to measure planning,

arousal, and coding functions. The authors based the model

on the work of Luria (1973) who identified three functional

brain units or blocks.










Luria's Model of Information Processing

The first of Luria's identified units is arousal.

Arousal is the prerequisite for all mental processing

because an optimal state of arousal is necessary in order

for effective cognitive processing to occur (Das, 1984). An

appropriate level of arousal allows an individual to direct

attention toward relevant stimuli, direct attention away

from irrelevant stimuli, and divide attention between

activities without decreased efficiency (Naglieri, 1988).

The second functional unit proposed by Luria is

coding. Coding is the cognitive function involved in

receiving, processing, and storing information. Luria (1966)

found strong evidence of simultaneous and successive

processes within the coding unit of the brain. In

simultaneous processing, each element o\f a stimulus is

related to all other elements of the stimulus (Das, Kirby,

and Jarman, 1975). Because all aspects of the stimulus are

accessible during inspection, simultaneous activities are

said to be surveyable. Luria stressed the importance of

simultaneous processing in understanding relationships of

any kind. Simultaneous processing is also required for

solving matrices (Raven, 1956) and copying designs

(Naglieri, 1988).











The other aspect of coding is successive processing

whereby each element of a stimulus is related only to

adjacent elements. In successive processing, stimuli are

integrated into series and synthesized into a chain-like

progression of elements (Das, 1973). Stimuli cannot be

surveyed in successive processing tasks. Successive

processing is especially important in learning skilled

movements such as writing (Luria, 1966) and is involved in

recalling series of words (Naglieri & Das, 1988).

Simultaneous and successive processing are both involved in

understanding statements and in reading comprehension (Das,

Cummins, Kirby, & Jarman, 1979).

The last unit described by Luria is planning. Planning

is the process that is responsible for ". programming,

regulation, and verification of activity (Naglieri &

Das, 1988, p. 38). It is described in Luria's approach as a

process or function rather than an ability (Das, 1980). A

planning test ". should indicate how the individual

approaches a task, the strategies he uses to reach a

solution" (Das, 1980, p.142). The planning function is used

after information has been received, coded, and stored. The

planning function is involved in formulating plans of

action, regulating behavior so that it corresponds to plans,

comparing results of actions to intentions, and correcting

errors. The planning unit has been determined to be











responsible for complex functioning such as problem solving

(Naglieri, 1988).

Although some amount of planning is necessary to

execute coding processes, a separate planning factor is

possible because variance could not be completely explained

by coding tasks (Das, 1980). The independence of the three

processes is assumed by Luria's theoretical model that

suggests that independent areas of the cortex are

responsible for coding and planning. The independence of

coding and planning is not only supported by factor analytic

studies, but by clinical work as well (Luria, 1973).

The three functional units of the brain have different

underlying anatomical correlates (Luria, 1980). Arousal is

located in the brainstem, diencephalon and medial regions

while coding originates in lateral areas of the neocortex.

Planning, the most complex form of human behavior, is

primarily associated with the frontal lobes. Clinical

studies of patients identified brain lesions have shown

planning and coding to be clearly separate (Das, 1980).

Individuals with frontal lobe damage have been found to have

minimal problem with coding information, but have deficient

planning processes. Those with damage to parietal,

occipital, or temporal lobes typically have little problem

with planning, but display deficits in coding processes.

Popper and Eccles (1977) found prefrontal lobe damage to











result in problems performing tasks requiring flexibility

and insight. Das (1980) suggests that good measures of

planning will be able to discriminate between patients with

and without frontal lobe lesions.

The three functional units, according to Luria (1973),

are responsive to experiences of the individual and,

therefore, are subject to developmental changes. Although

distinct, the units continually interact with each other.

Most traditional intellectual measures are focused only on

one of the functional units proposed by Luria, the coding

process. In other words, intelligence tests generally

measure the ability to process information through sorting,

storing, and retrieving information (Naglieri & Das, 1988).

These instruments fail to measure one of the highest forms

of human behavior- planning (Bracken, 1985; Naglieri,

1988). The Cognitive Assessment System is designed to

provide analysis of all aspects of the PASS model and

measures planning and arousal in addition to coding

processes. The theoretical model underlying the instrument

is complex, but this is considered by the authors to be one

of the major strengths of the instrument (Naglieri, 1988).

The Cognitive Assessment System

Although the CAS is still in the developmental stage,

its characteristics and technical qualities are sufficient

to warrant inclusion in the present study. Extensive











efforts have been directed at developing the CAS and several

studies have supported its validity. Naglieri (1988)

evaluated the PASS model of assessment using six criteria

proposed by Dillon (1986) and concluded that the model had

the necessary characteristics to be considered potentially

useful for assessing cognitive processing functions. The

six criteria included validity, diagnosis, prescription,

comparability, replicability/standardization, and

psychodiagnostic utility.

Validity. The work of Luria (1966) provides evidence

for three functional brain units that have underlying

physiological correlates as has been previously discussed.

Das (1973) later provided additional evidence for the

existence of simultaneous and successive processes in a

factor analytic study of cognitive tasks administered to

groups of Canadian and Indian boys between the ages of 9 and

11.

In a factor analytic study of several cognitive tasks

administered to 60 retarded and 60 nonretarded children, the

successive and simultaneous factors emerged for both groups

(Das, 1972). However, certain tests loaded differently

between the two samples. For example, the Memory for

Designs test was loaded on the simultaneous factor for the

retarded group whereas it was loaded on the sequential

factor for the nonretarded group.











Sequential and simultaneous processes were found to

exist across age levels and socioeconomic levels in a study

of 60 first grade and 60 fourth grade students (Molloy,

1973). A varimax rotation of nine tasks yielded three

factors that were similar for both samples- successive,

simultaneous, and speed. Similarly, the three factors

emerged when the tasks were administered to 60 low

socioeconomic status children and 60 middle-to-high

socioeconomic status children, although slight disparities

were noted. Visual short-term memory loaded on the

successive factor for fourth grade children while it loaded

on the speed factor for first grade children. However,

varimax rotation assumes that underlying factors are

independent and its use with mental tests may be

inappropriate.

Successive and simultaneous processes have also been

found to exist across cultures. Das (1973) administered six

tasks to a group of 90 children from Orissa, India and found

that the same three factors emerged. Likewise, similar

factor structure resulted when 10 cognitive tests were

administered to 40 native Canadian children in the third and

fourth grades (Krywaniuk & Das, 1976). Simultaneous and

successive coding processes also were identified in a group

of 60 Indian children between the ages of 7 and 9 (Dash,

Puhan, & Mahapatra, 1985). Simultaneous, successive, and











planning factors emerged in a factor analytic study of

several cognitive tasks administered to a group of Chinese

students, and each process was significantly related to

reading achievement (Leong, Cheng, & Das, 1985).

The successive, simultaneous, and planning factors of

the PASS model have been found to exist across ages.

Naglieri and Das (1988) administered nine cognitive tasks to

149 second-grade students, 160 sixth-grade students, and 125

tenth-grade students. Orthogonal varimax and promax oblique

rotations revealed that successive, simultaneous, and

planning factors emerged at all three grade levels.

Using the same sample, Naglieri and Das (1987) found

that successive, simultaneous, and planning factors changed

with developmental level and chronological age and were

related to achievement, lending support to the construct

validity of the PASS model. The developmental nature of the

model was apparent as raw scores significantly increased

from grades 2 to 10 for those tasks where the score was the

number of correct responses. Likewise, significant

decreases were noted between the two grades for timed

tasks. Evidence for criterion-related validity was provided

through correlational analyses of processing composite

scores and achievement scores on the Multilevel Academic

Survey Test (MAST). The processing composite scores

correlated significantly with scores on the MAST at each











grade level (r= .207 to .555), and planning-achievement

correlations increased with age. Similar significant

relationships and developmental trends were apparent in the

coding-achievement correlations. Results allowed for the

conclusion that both coding and planning functions were

related to achievement and planning became more important

with increasing age.

Successive and simultaneous processes were identified

in additional studies and were found to be related to

achievement. In a study of 99 fourth grade students, Kirby

and Das (1977) found these processes to be correlated with

reading achievement. Correlation coefficients ranged from

r=.316 between successive processing and vocabulary to

r=.507 between simultaneous processing and reading

comprehension. Das and Cummins (1978), in a study of 52

educable mentally retarded adolescents, found that

simultaneous processes were related to performance IQs on

the WISC-R (r=.58) and WRAT arithmetic subtest scores

r=.28). Successive processing scores were related to a

reasoning task (r=.35) as well as WRAT spelling and reading

subtest scores (r=.32 and r=.33, respectively). Early

reading skills such as decoding have been shown to involve

primary use of successive processing strategies while higher

level skills such as reading comprehension draw largely on

simultaneous processes (Das et al., 1979).










Ashman (1978), in a study of 104 eighth grade students,

found two planning tasks, (i.e. trail-making and visual

search), to load with other activities requiring planning

such as planned composition and verbal fluency (Das, 1980).

These two tasks did not load with successive processing

tasks, (i.e. auditory serial recall, visual short-term

memory, and digit span), or simultaneous processing tasks,

(i.e. figure copying and memory for designs), suggesting

that planning processes can be measured separately from

coding processes. Similar results were obtained in a sample

of 46 mildly retarded adults. Additionally, the same three

factors emerged with a group of trainable mentally retarded

subjects ranging from 12 to 22 years in age, although only

two subtests from each of the three areas were administered.

The attentional component was only recently included in

the PASS model and research was still being conducted on the

validity and reliability of attentional tasks at the time of

this study (Naglieri, 1988). Tasks similar to those

described by Posner and Boles (1971) and Stroop (1935) have

been explored for inclusion in the CAS and seem to be valid

for assessing the attentional component (Naglieri, 1988).

Diagnosis. The inclusion of the planning and

attentional components in the PASS model has resulted in its

being a more sensitive and efficient diagnostic tool than

traditional instruments (Naglieri, 1988). Use of successive











and simultaneous measures alone has been shown by

investigators to be ineffective for identifying learning

disabilities (Naglieri, 1985a; Naglieri & Haddad, 1984).

Although reading disabled students perform similarly to non-

disabled children on successive, simultaneous, and attention

tasks, they perform similarly to retarded children in the

area of planning (Naglieri, 1987). The omission of planning

tasks from traditional measures such as the Kaufman

Assessment Battery for Children (K-ABC) and the WISC-R

causes these instruments to be insensitive to cognitive

weaknesses (i.e. planning deficiencies) that may be

associated with subtle learning problems (Bracken, 1985).

Because the PASS model incorporates planning and attentional

components, it provides a broader conceptualization of

cognitive functioning and may therefore be more effective

for diagnosing exceptional children (Naglieri & Das, 1988).

Because the PASS model has a broader view of intellectual

competence, it may also be likely to be more effective as a

predictor of success in occupational or educational programs

(Naglieri, 1988).

Prescription. The PASS model is a process model rather

than an abilities model. While abilities stress capacities

of the individual, a process model emphasizes the strategies

that an individual uses. Examining processes rather than

abilities can result in a better understanding of how











individuals perform tasks, knowledge of how to train them

more successfully, and information about how to design more

effective educational programs (Das et al., 1979). Training

in successive and simultaneous coding processes has been

shown to result not only in improvement in these processes,

but in improved academic performance as well (Brailsford,

Snart, & Das, 1984, Krywaniuk & Das, 1976). Although

planning and attentional measures have not been included in

these studies, the extent to which training in planning

processes carries over into academic performance can be

inferred from studies of metacognition training (Brown &

DeLoache, 1978; Wellman, Fabricius, & Sophian, 1985).

Comparability. Processing measures must be related to

a target task, and that is the case with the PASS model. By

analyzing the structure of a task, the underlying cognitive

processing component can be ascertained. The differences

between the processing tasks of the PASS are obvious when

components required for performance of each task are

examined. For example, simultaneous processing tasks

require the individual to relate parts of a task into groups

where each element of a group must be considered in relation

to all other elements of a group. In other words, all

aspects of the task will be surveyable (Das et al., 1979).

Successive processing tasks require the individual to

observe the linear nature of stimuli and all aspects of the











task will not be surveyable (Naglieri, 1988). Attentional

tasks require the individual to not only respond to a

stimulus, but also to suppress reactions to irrelevant

stimuli (Stroop, 1935). Finally, planning tasks require the

individual to determine and use the most efficient way to

solve a problem. Planning processes will underlie a task

when ". the individual is required to analyze a task,

develop a means of solving the problem, evaluate the

effectiveness of the solution, modify the plan as needed,

and demonstrate some efficient and systematic approach to

problem solving" (Naglieri, 1988, p.18).

Replicability/Standardization. The value of a

measurement system is increased when it can be organized

into a consistent method and applied across examiners so as

to allow for replication (Naglieri, 1988). Additionally,

normative data for comparative purposes increases the value

of a measurement system by reducing interpretation errors.

Although the PASS model has not yet been standardized, the

operationalization of the model has shown that replication

is possible (Naglieri & Das, 1987; 1988). The Cognitive

Assessment System, developed by Naglieri and Das (1987) and

scheduled for publication in 1990, was designed to provide a

standardized measure of planning, attention, and coding

processes (i.e. successive and simultaneous processing).











Psychodiagnostic Utility. The psychodiagnostic utility

of the PASS model has already been previously discussed.

The utility of the model can be seen in the numerous studies

that have used it across age ranges, exceptionalities, and

cultures. Additionally, identification of children with

learning disabilities may be accomplished more successfully

with the PASS model than with traditional instruments

because of its sensitivity to intellectual variability

across processing functions. The degree to which diagnostic

information obtained from the PASS model can be used for

developing intervention strategies will require additional

research. However, several studies have suggested that

training in processing strategies can result in increased

academic performance and improved use of cognitive

processes (Brailsford et al., 1984).

Summary

Although VLBW babies constitute a small number of total

births, these babies contribute disproportionately to the

number of children with later neurological and developmental

delays. While major handicapping conditions have not been

demonstrated to have increased porportionately in the

population of VLBW children with the increased survival

rate, VLBW children appear to be at risk for later learning

problems. Subtle handicapping conditions such as visual-

motor delays, attentional deficits, organizational problems,











and motor delays have been found to occur more often in

children born of very low birthweight.

Several researchers have suggested that the subtle

problems exhibited by many VLBW children are caused by

underlying neurological differences in this group of

children. According to synactive developmental theory, VLBW

babies are at risk for developing ". deleterious

adaptation patterns (Als, 1986, p.4) because they

are placed in environments that are poorly matched to their

adaptation capabilities. This mismatch could lead to

inhibition of normal pathways and overactivation of current

functional pathways in the premature neonate's overly

sensitive, immature brain, resulting in less differentiated

and less modulated behavior on the part of the preterm

infant. According to synactive theory, premature birth can

be dangerous and lead to long lasting changes in the anatomy

and function of the brain.

Increasing numbers of children born with very low

birthweights are now reaching school age where a high

percentage are encountering academic difficulties. The

reasons for these learning problems have been difficult for

researchers to identify. Although scores on global

intellectual measures do not typically differ between VLBW

and full-term children, a high percentage of VLBW children

display indications of learning disabilities and subtle











developmental differences. Very low birthweight children

tend to be less organized than full-term counterparts as

neonates, with the disorganization continuing into infancy.

The extent to which these organizational differences

continue to persist at later ages has not yet been

examined. The lack of longitudinal follow-up of VLBW

children has led to difficulties in identifying the

variables related to learning problems in this population.

Knowledge of these variables could lead to more effective

educational programming for VLBW children, more relevent

psychological evaluations, and increased information about

the risk status of VLBW children as suggested by synactive

developmental theory.

In an attempt to identify underlying psychological

processes that could affect academic functioning of VLBW

children, the Cognitive Assessment System (CAS), based on

the PASS model, was used in this study. The CAS was based

in Luria's work and provides information on three functional

processes of the brain: arousal or attention, coding (i.e.

successive and simultaneous processing), and planning or

organization. The CAS provides information on planning and

attention, psychological processes that have traditionally

been excluded from measures of intelligence. The CAS was

administered to samples of VLBW and full-term children at 8

years of age. This age was selected for the purpose of







54



longitudinal data collection since at this age subtle

learning difficulties usually become apparent to educators.














CHAPTER III

METHODOLOGY

It was the purpose of this study to examine the

cognitive functions of VLBW children at 8 years of age.

Rather than examining global intellectual functioning, it

was the purpose of this study to investigate cognitive

processing variables in VLBW children. It was proposed that

identification of these variables would lead to more

effective educational programming for VLBW children and,

consequently, a decrease in their currently high risk status

for educational difficulties.

The dependent variables in this study were those

identified by Luria (1980)--specifically, attention

processes, coding processes (i.e., simultaneous and

successive processes), and planning processes. Data were

collected on these variables using the Cognitive Assessment

System developed by Naglieri and Das (1987). The

independent variable in this study was birthweight.

For the purposes of presentation, this chapter has been

divided into five sections. The sections are the

description of the null-hypotheses, description of the











subjects, description of the research instrumentation,

description of the procedures, and treatment of the data.

Research Hypotheses

The following null hypotheses were tested at the .05

level of confidence:

HI There will be no statistically significant

difference between the VLBW children and the full-

term children in the mean organizational

processing scores on the CAS.

H2 There will be no statistically significant

difference between the VLGW children and the full-

term children in the mean simultaneous

processing scores on the CAS.

H3 There will be no statistically significant

difference between the VLBW children and the full-

term children in the mean successive processing

scores on the CAS.

H4 There will be no statistically significant

difference between the VLBW children and the full-

term children in the mean attentional processing

scores on the CAS.

Subjects

There were a total of 51 subjects selected for this

study. The 21 experimental subjects included for the

purposes of this study were all VLBW survivors treated at












Sacred Heart Hospital between the years 1979 and 1981 and

who were 8 years of age at the time of the study. Sacred

Heart is a regional Intensive Care Center (ICC) serving

Pensacola and 17 rural counties in the northwest section of

Florida referred to as the panhandle. Additionally,

children from surrounding states (e.g., Alabama,

Mississippi, and Georgia) are served by the center.

Demographic characteristics of children born between the

years 1979 and 1981 were unavailable. However, such

information was available for all VLBW children treated at

Sacred Heart's ICN between the years 1983 and 1988. It is

likely that the demographic characteristics of children born

between 1979 and 1981 were similar to children born during

the 5 year period following 1983 (D. Goldberg, personal

communication, April 3, 1989).

Between the years 1983 and 1988, a total of 669 VLBW

babies were admitted to the Intensive Care Nursery (ICN) at

Sacred Heart. Of these, 513 babies or 77% were discharged

from the ICN. There was a 23% mortality rate for these

infants during the neonatal period. The survival rate among

these infants was similar to that reported by Hunt et al.

(1982). The number of male and female VLBW survivors was

approximately the same (i.e. 49% male and 51% female).

There was a larger percentage of black VLBW survivors than

was reported by McCormick (1985) with black VLBW babies











constituting 42% of all VLBW babies in the group of VLBW

survivors. The percentage of Hispanic and other races was

less than 1%. Approximately half of the group of VLBW

survivors were of low socioeconomic status, as 55% received

some medicaid services at the time of birth. Medicaid

eligibility was unknown for 6% of the group and the

remaining 39% had private insurance.

Experimental Subjects

The 21 VLBW subjects selected for the purposes of the

study were demographically similar to the population of VLBW

children treated and discharged from Sacred Heart's ICN

between the years 1983 and 1988. There was a slightly

higher percentage of males in the subjects selected than was

found in the population of VLBW children treated at Sacred

Heart during these years. The subjects selected consisted

of 57% males and 43% females while the population consisted

of 49% males and 51% females. The percentages of black

children were similar between the two groups- 43% for the

subjects selected and 42% for the population. Children of

races other than white or black constituted less than 1% of

the population and were not included in the sample of

subjects selected. The mean age of the VLBW group of

children was 8 years, 8 months.











The VLBW subjects selected contained fewer low

socioeconomic children than were found in the population.

Attempts to locate children who were in this category were

often unsuccessful. Many parents of those VLBW children who

could be located either did not respond or refused to

participate. The subjects selected consisted of 33% of the

children whose mothers were of lower socioeconomic status at

the time of birth. Lower socioeconomic children constituted

55% of the population. Thus there was 22%

underrepresentation of low SES children in the sample for

this study. The remaining 67% of the VLBW sample were

considered to be from middle to upper income families at the

time of birth.

Control Subjects

A control group of 30 children was selected for this

study. The mean age of the control group of children was 8

years, 6 months. The control group consisted of 63% white

children and 37% black children. The percentages of male

and female children in the control group were approximately

the same as 53% of the group were male and 47% were female.

Approximately one-third of the sample of children were of

lower socioeconomic status at the time of birth while the

remaining 67% were of middle to upper socioeconomic status

at the time of birth.











Sampling Procedures

A sample of 21 VLBW children were selected from all

VLBW babies treated in the Intensive Care Nursery at Sacred

Heart Hospital between July of 1979 and July of 1981.

Attempts were made to obtain a sample of VLBW children that

was representative of all VLBW children admitted and

discharged from Sacred Heart's ICN during the years 1979-

1981 in terms of gender, race, and socioeconomic status.

All VLBW children selected were 8 years of age at the time

of the study with the exception of one child who was 9

years, 2 months of age. Children born with obvious

anomalies or chromosomal abnormalities were excluded from

the study. None of the children included in the study had

been diagnosed with fetal alcohol syndrome at birth.

Additionally, children who were attending classes for

mentally handicapped students and children with severe

behavioral problems were excluded from the study.

Because norms were not yet available for the Cognitive

Assessment System and in order to provide a group of full-

term children for comparison, 30 full-term children

currently residing in the Pensacola area were selected for

participation in the study. A stratified sample of full-

term children was selected so as to match the VLBW sample of

children in terms of age, gender, race, and socioeconomic

status. The sample was stratified in that specific numbers











of children were selected in each of the demographic

categories for inclusion in the control group of children.

The first step in obtaining a stratified sample of

control subjects involved determining the percentages of

VLBW children constituting each demographic category. These

percentages were then used to determine the number of full-

term children that would be needed in each demographic

category. Attempts were then made to select full-term

children in a manner so that percentages of children in each

demographic category were similar to those of the

experimental group. Children in the control group had

birthweights of at least 2500 grams, were free from obvious

anomalies and chromosomal abnormalities, and were not

attending classes for mentally handicapped students.

The VLBW sample of children was selected by examining

hospital records from the years 1979 to 1981. Addresses and

phone numbers from these records were cross-checked with

data in developmental follow-up records and phone books.

Parents were contacted by mail, informed of the nature of

the study, and asked to participate (Appendix A).

Additionally, they were asked to complete a form requesting

demographic information (i.e. maternal education, birthdate

of child, and race of child) (Appendix B). Finally, they

were asked to complete permission forms allowing their child

to participate in the study (Appendix C). Materials were












sent to parents of all VLBW children for whom current

addresses could be determined. All children whose parents

granted permission for participation in the study were

included in the VLBW sample of children.

Children in the control group were drawn from three

sources--siblings of children enrolled in the developmental

follow-up program at Sacred Heart, children of hospital

employees, and children enrolled in the public schools.

These children were selected so as to resemble the VLBW

group of children in terms of age, race, gender, and

socioeconomic status. Children were located by examining

developmental follow-up records and through a notice posted

in the hospital newspaper requesting volunteers for the

study. Principals of local public schools assisted in

locating some of the children for the study.

When a child was located who had the necessary

demographic characteristics for inclusion in the study,

parents were contacted by phone, informed of the nature of

the study and asked if they would participate. Parents who

expressed interest were then mailed letters explaining the

nature of the study (Appendix A). They were also asked to

complete demographic information and permission forms

(Appendices B and C). All children whose parents granted

permission were included in the study. Children were

selected in this manner until 30 children were found who












closely resembled the VLBW group of children in terms of

age, race, gender, and socioeconomic status.

Parents who granted permission for their child to

participate in the study were mailed a letter thanking them

for participating and notifying them of their appointment

time (Appendix D). They were asked to return the letter

indicating whether they could participate during that time

or whether other arrangements needed to be made. Parents

who needed to reschedule were contacted by phone to arrange

a more convenient time.

Comparison of Groups

The control group of full-term children was quite

similar to the VLBW sample of children in terms of age,

race, gender, and socioeconomic status. Demographic

characteristics of the two groups of children are depicted

in Table 3-1. The VLBW group of children contained slightly

higher percentages of black children and male children than

the full-term group of children. However, these differences

were minimal. The mean ages of the two groups were

similar. The mean age of the term-birth and VLBW groups of

children were 8 years, 6 months and 8 years, 8 months,

respectively. The incidence of learning disabilities was

also similar in both groups of children, but the control

group had a slightly higher incidence. The incidences of

identified learning disabilities in the term-birth and VLBW












groups of children were 6.5% and 4.5% respectively. There

were two children identified as learning disabled in the

control group and one in the experimental group. The

incidence of identified gifted children was also similar

between the two groups of children. There was one child

identified as gifted in the experimental group while there

were two in the control group. Corresponding percentages of

gifted children in the experimental and control groups were

5% and 6%, respectively.


Table 3-1.


Demographic Characteristics of VLBW and Full-
Term Samples of Children


White

Black

Male

Female

Low SES

Middle-High SES


VLBW Group

57% (N=12)

43% (N=9)

57% (N=12)

43% (N=9)

33% (N=7)

67% (N=14)


Term

63%

37%

53%

47%

33%

67%


Group

(N=19)

(N=11)

(N=14)

(N=16)

(N=10)

(N=20)

30


Because slight differences were found to exist between

the VLBW and full-term groups of children, a decision was

made to covary on three variables--child's age at the time


--











of testing, maternal age at the time of birth, and maternal

education at the time of birth. Although the differences

were not statistically significant, the use of these three

variables as covariates strengthened the study by providing

greater statistical power. The VLBW group of children was

slightly older than the full-term group of children.

Maternal age and educational level at the time of birth were

slightly higher for the full-term group. Mother's

educational levels were divided into four levels. Level one

included those with less than a high school education while

level two mothers had graduated from high school. Mothers

with some college education or technical training fell into

level three while level four mothers were college

graduates. Means and standard deviations of these variables

may be located in Table 3-2.

Adequacy of Sample Size

The sample sizes used in the study were small and may

be construed to serve as a limitation of the study. Small

sample sizes have often been a problem with studies of VLBW

children, particularly those studies that are longitudinal

in nature. The present study was an improvement over

previous studies of VLBW children that used even smaller

sample sizes (Bennett et al., 1982; Ruff, 1986). In order

to minimize problems created by small sample sizes, a

control group was selected that was demographically similar












to the experimental group in this study. Additionally, the

use of covariates in the data analysis further minimized any

group differences, thereby strengthening the design of the

study.



Table 3-2. Distributional Characteristics of Covaried
Variables for Full-Term and VLBW Samples of
Children


Variable Group Mean Standard Deviation

Child's Age Term 101.97mos. 7.06mos.
VLBW 104.29mos. 6.40mos.

Maternal Age Term 27.47yrs. 5.22yrs.
VLBW 27.33yrs. 5.69yrs.

Maternal Term 2.87 .78
Education VLBW 2.81 .87


Research Instrument

Children were tested individually using the Cognitive

Assessment System (CAS). The CAS is an individually

administered test of cognitive processes for children

between the ages of 5 and 18. At the time of the study,

the CAS consisted of 16 subtests, four each for the areas

of attention, successive processing, simultaneous

processing, and planning. The test required approximately

2 hours to administer. Directions for administration were

included in the manual.











Validity

Although the CAS was still being developed at the time

of the study, the 16 subtests had been shown to be

adequately valid for research purposes. Therefore,

permission was granted by the Psychological Corporation for

use of the CAS in the current study. The construct

validity of the CAS subtests has been repeatedly shown

(Ashman, 1970; Das, 1980; Leong et al., 1985; Naglieri &

Das, 1987; 1988). Ashman (1978), in a study of 104 eighth

grade students, established construct validity for two of

the planning tasks of the CAS. Naglieri and Das (1988) and

Leong et al., (1985) established the construct validity of

the successive, simultaneous, and planning tasks through

factor analytic studies.

The developmental nature and criterion-related

validity of the CAS subtests were shown in a study by

Naglieri and Das (1987). Cognitive Assessment System

subtests were correlated with scores on the Multilevel

Academic Survey Test (MAST). Correlation coefficients

ranged from r=.207 to .555 with planning-achievement

correlations increasing with age. Other studies have

allowed for the conclusion that the CAS coding subtests are

correlated with achievement (Das & Cummins, 1978; Kirby &

Das, 1977).










Studies were being conducted by the Psychological

Corporation on the attentional component of the CAS at the

time of this study (J. Sugarman, personal communication,

February 3, 1989). However, construct validity for some of

the CAS attention items has been previously established

(Posner & Boles, 1971; Stroop, 1935). Preliminary research

findings involving the attention subtests have allowed for

the conclusion that they are valid (Naglieri, 1988).

Reliability

Studies on the reliability of the CAS were being

conducted through the Psychological Corporation at the time

of the study (J. Sugarman, personal communication, February

3, 1989). However, preliminary research findings have

shown that administration of the CAS subtests across

examiners is possible (Naglieri & Das; 1987; 1988).

Because interrater reliability had not been established at

the time of the study, three interrater reliability checks

were made during the study. Two of these reliability

checks were made with VLBW subjects and one was made with a

control subject. The checks were made at 2-week

intervals. While the examiner administered and scored the

CAS, a second trained examiner observed the administration

and scored the protocol. The two protocols were then

examined for interrater agreement. There was 99% or better

agreement on all three interrater checks, suggesting that











interrater reliability exists for the CAS. Test-retest

reliability of the CAS had not been established at the time

of the study.

Subtests of the Cognitive Assessment System

The 16 subtests, four each for the areas of planning,

successive processing, simultaneous processing, and

attention were as follows:

1)Visual Search. In performing this subtest, the

individual was required to point to an object or

letter surrounded by numerous objects or letters.

The score was the time required from initial

exposure of the stimulus to the point where the

individual identified the target object or letter.

This task was shown to load on a planning factor

(Das, 1984; Naglieri & Das, 1987; 1988).

2)Planned Codes. The individual was required to code

a series of boxes labelled with the letters A, B, C,

or D using sequences of X's and O's given in the

example (i.e. A=OX, B=XX, etc.). There were two

items in this section, each having different codes

and arrangements. The child was asked to code the

boxes as quickly as possible. The score was the

number of boxes correctly coded within the given

time limits. This task was shown to load on a

planning factor (Naglieri & Das, 1988).











3)Planned Connections. In this subtest, the subject

was required to connect numbers located in boxes in

correct numerical order. The numbers were randomly

distributed on the page and the score was the time

required to complete the task. In performing more

advanced items on this subtest, the individual was

required to shift between connecting numbers and

letters in ascending or alphabetical order. Tasks

similar to those on this subtest have been used by

Reitan (1955) and Spreen and Gaddes (1969) and were

originally part of the Army Individual Test of

General Ability (1944). Armitage (1946) found that

this task measured planning because it required the

individual to organize a double relationship, avoid

perseveration, and shift between rules. Das (1984)

and Naglieri and Das (1987; 1988) found that

trailmaking tasks loaded on a planning factor.

4)Matching Numbers. The subject was required to

locate and circle two numbers that were

identical. The length of the numbers ranged from

one to six digits and the score was the correct

number of pairs circled in three minutes. This











subtest was found to load on a planning factor

(Naglieri & Das, 1987; 1988).

5)Design Construction. The individual used blue

and white chips to construct abstract designs

located in a stimulus book. The score was the

number of designs successfully constructed within

the specified time limits. This task was shown

to load on a simultaneous factor (Das & Naglieri,

1989).

6)Simultaneous Verbal. The individual's task was

to listen to a question read by the examiner and

then choose the corresponding picture among six

options that answered the question. The

questions were printed at the bottom of each page

and involved logical-grammatical relationships.

This task was shown to load on a simultaneous

factor (Das & Naglieri, 1989).

7)Figure Memory. Designs ranging in difficulty

from a simple square to an open book design were

embedded in a complex background. The individual

was required to find a specific design and trace

it with a marker. If all components of a design

were traced by the individual, the item was

passed. This subtest was shown to load on a

simultaneous factor (Naglieri & Das, 1987; 1988).










8)Matrices. This subtest consisted of the Matrix

Analogies Test- Short Form (Naglieri, 1985b).

The score was the number of matrices successfully

solved. This subtest was composed of 34 items

similar to those on the Raven's Coloured

Progressive Matrices Test (Raven, 1956). The

Raven's Test was shown to load on a simultaneous

factor (Das et al., 1979). Similarly, this

subtest of the CAS was shown to load on a

simultaneous factor (Naglieri & Das, 1987;

1988).

9)Word Series. This subtest included nine single

syllable words that were presented orally. The

words were of high familiarity and varied from

two to nine words in length. The individual was

required to repeat the words in the same order

presented. The score was the number of items

successfully repeated. This subtest was shown to

load on a successive processing factor (Das et

al., 1979; Naglieri & Das, 1987; 1988).

10)Sentence Repetition and Questions. This subtest

contained two parts. In the first part, the

individual was required to repeat sentences

spoken by the examiner. In the second part, the

individual was required to answer a question











about each sentence as it was again spoken by the

examiner. Color names were substituted for

content words in the sentences in order to

minimize contexual cues (e.g. Blue gave the red a

black). The score was the number of sentences

correctly repeated and the number of questions

successfully answered. This task was shown to load

on a successive factor (Das & Naglieri, 1989).

11)Color Ordering. This subtest requires the

individual to turn a series of colored chips in

the same order demonstrated by the examiner. The

chips were arranged linearly on a small board and

were arranged in the sequence indicated in the

manual. The subtest was designed to measure the

individual's ability to reproduce the order of an

event and was found to load on a successive

factor (Naglieri & Das, 1987; 1988). The score

was the number of items correctly performed.

12)Successive Hand Movements. This subtest was

similar to that used on the K-ABC (Kaufman &

Kaufman, 1983). However, it differed in that six

hand movements were used rather than the three

used on the K-ABC. In performing items on this

subtest, the individual was required to reproduce

series of hand movements demonstrated by the











examiner. The score was the number of items

successfully performed. This subtest was found

to load on a successive factor (Naglieri & Das,

1987; 1988).

13)Expressive Attention. This subtest consisted of

three parts. In the first part, the child read

color words as quickly as possible. In the

second part, the child named colors as quickly as

possible. In performing the third part, the

child was required to attend to relevant stimuli

while suppressing irrelevant stimuli by stating

the inconsistent color that each color word was

printed in. The score was the time required to

complete the third part of this subtest. This

task was shown to load on an attention factor

(Das & Naglieri, 1989).

14)Receptive Attention. This subtest consisted of two

parts. In the first part, the individual was

required to find and underline pairs of letters

that were the same physically. In the second

part, the individual was required to find and

underline pairs of letters that had the same

name. The score was the number of pairs

correctly located on both sections within the











given time limits. This subtest was shown to

load on an attention factor (Naglieri, 1988).

15)Auditory Selective Attention. The individual was

required to listen to a 5 minute tape and

identify target words from a group of stimulus

words. The individual was asked to tap the table

each time a man said certain stimulus words and a

woman said other stimulus words. This task

required the individual to attend to relevant

auditory stimuli while suppressing irrelevant

auditory stimuli and was shown to load on an

attention factor (Das & Naglieri, 1989). The

score was the number of stimulus words

incorrectly identified subtracted from the number

correctly identified. If a child incorrectly

identified more words than were correctly

identified, a zero score was given.

16)Number Finding. The individual was required to

locate and underline specific target numbers

within a page of numbers that contained both the

targets and distractor stimuli. The score was

the number of numbers correctly located. This

task was shown to load on an attention factor

(Das & Naglieri, 1989).










Rationale

The CAS was selected over other instruments for use in

this study because it provided a broader measure of

cognitive functioning than traditional intellectual

measures could provide. Because global intellectual

measures have typically not been effective for examining

differences between VLBW and full-term children, an

instrument more sensitive to distinct cognitive processes

was needed. The CAS was chosen because it provided

measures in four cognitive processing areas (i.e. planning,

simultaneous processing, successive processing, and

attention). Traditional intellectual assessments have not

included the planning and attention components; components

that are crucial to examine in VLBW children in light of

synactive developmental theory.

Assessment Procedures

Each child was tested individually and required

approximately two hours to complete the CAS. Breaks were

provided as needed to participants, usually half way

through the evaluation (i.e. after the Matrices subtest).

Most testing was conducted in a room free from distractions

at the Children's Developmental Clinic located at Sacred

Heart Hospital. Parents were asked to transport their

children to the clinic for testing, but children were

examined without their parents present. Parents who could











not transport their children to the clinic were asked to

meet the examiner at a more convenient location, usually a

nearby school. All testing was conducted by a trained

examiner who attended a one-day training workshop conducted

by the Psychological Corporation. The examiner was a

certified, experienced, and licensed school psychologist

who demonstrated proficient ability in administering the

CAS prior to evaluating children in this study. Subtests

were administered in the order specified in the CAS manual.

Additionally, subtests were administered according to

directions specified in the manual. Parents were given

generally feedback about their child's performance after

testing (i.e. attention span and cooperation).

Data Analysis

Distributional characteristics of all variables were

examined first. These included the mean, standard

deviation, mode, median, kurtosis, and skewness of each

variable. All variables were found to have the necessary

characteristics for the chosen data analyses (i.e. normal

distributions and homogeneous variances). In examining the

four null-hypotheses, mean subtest scores were compared

between groups across the four areas of the CAS using

multiple analysis of covariance (MANCOVA). Because

normative data were not yet available for the CAS at the

time of the study, raw scores were converted to z-scores











across the 16 subtests for the full-term group. Raw scores

were then converted to z-scores for the VLBW group using

this z-distribution. Area aggregate z-scores were

calculated by summing the scores of the four subtests

within each of the four areas of the CAS. Multiple

analysis of covariance was conducted on the resulting four

aggregate cluster scores to determine whether significant

differences existed between groups (p<.05).

In order to control for initial group differences,

three covariates were entered into the analysis: age of

child at the time of testing, maternal age at the time of

birth, and maternal education at the time of birth.

Although attempts were made to equate the two groups on

these three variables, it was found that slight differences

existed. The use of these variables as covariates

decreased the influences that they might have had on the

results of the analysis.

Results of the MANCOVA yielded a significant F-value.

Therefore, subsequent univariate F-tests were performed to

determine which of the CAS area scores differed

significantly between the two groups of children (p<.05).

Methodological Limitations

There were a few methodological limitations in this

study that should be mentioned. First, the CAS was not

completed at the time of the study. The instrument that











was used in this study will, in all probability, differ

from the final version of the instrument. For example, the

final version will include only 12 subtests rather than the

16 subtests that it contained at the time of the study.

Additionally, individual items could change before the

final version of the CAS is published. While it would have

been better to have the final version of the instrument,

the items and subtests in the experimental edition of the

CAS had been shown to measure the processes that the test

was designed to assess (Naglieri & Das, 1987; 1989).

Therefore, the test was considered to be adequate for

examining the processing functions of interest.

Additionally, it was thought to be an improvement over

traditional instruments that have been used in studies of

VLBW children because it provided a broader view of

cognitive functioning (Naglieri, 1988).

Related to this limitation was the fact that all

children were tested by a single trained examiner.

Therefore, the possiblility of experimenter bias effect was

present. To minimize this problem, children were scheduled

several weeks in advance in hopes of reducing any

association of child's name with group membership.

However, the experimenter was sometimes aware of the group

membership of the children because of the extreme

difficulties encountered in locating VLBW children. Also,











parents accompanied their children to the test situation

and often talked spontaneously about their nursery

experiences. The fact that the CAS is an objective

instrument also served to minimize experimenter bias.

Another potential limitation of having a single

examiner involved the reliability of scoring. Because the

CAS is a new instrument, the degree to which reliable

ratings could be given to responses was uncertain.

However, the three interrater reliability checks that were

performed throughout the course of the study should have

minimized this limitation.

Another limitation related to the use of the CAS in

this study was the lack of normative data. Although it

would have been better to have large-scale normative data

against which to compare the performance of VLBW children,

the provision of a demographically similar control group

of full-term children should have minimized this

limitation (Borg & Gall, 1983, Kitchen et al., 1980).

The fourth limitation of this study was related to

medical advances that have occurred over the past decade.

The degree to which results of this study can be

generalized to VLBW children born today is questionable due

to rapid advances in medical technology. Advances in

medical technology are always a problem with follow-up

studies of VLBW babies. Nevertheless, the importance of











studies of VLBW babies. Nevertheless, the importance of

following VLBW babies has been repeatedly stressed in the

literature and the fact that medical technology is

continually changing does not subsume the need for

continued longitudinal studies (Als et al., 1988; Hunt et

al., 1982).

A fifth limitation involved the lack of early relevant

data on VLBW children that might be predictive of outcome.

Although data such as days on oxygen, length of hospital

stay, Apgar scores, and socioeconomic status were

available, such variables have not typically been

predictive of outcome. Although some studies have enabled

the determination to be made that maternal education and

other indicators of SES predict outcome of VLBW children

(Chamberlin,1987; DeHirsch, Jansky, & Langford, 1966; Lasky

et al., 1987), others have failed to substantiate such

findings (Als et al., 1988; Cohen et al., 1988; Drillien

et al., 1980; Hertzig, 1981; O'Reilly et al., 1986).

Rather, it appears that more subtle, neurological early

indicators (i.e. habituation to novel stimuli, state-

organization, attentional processes) may be the best

predictors of outcome (Als et al. 1988; Cohen et al.,

1988). Unfortunately, such information is lacking in the

current study.










Finally, the population of VLBW children from Sacred

Heart's ICN was not entirely representive of VLBW children

nationwide. There was a lack of children from major

metropolitan areas as well as Hispanic children.

Additionally, the sample of VLBW children selected for this

study had fewer low socioeconomic children than had been

found in the population of VLBW children nationwide. The

degree to which results obtained in this study are

generalizable to VLBW children in other samples will depend

partly on the demographic make-up of VLBW children in other

areas of the country. However, the use of neurologically-

based assessment procedures in this study rather than

global intellectual measures should have minimized the

effects of socioeconomic status and race on test

performance because cognitive processes rather than

abilities were examined.














CHAPTER IV

RESULTS

The purpose of this study was to examine cognitive

processing functions of very low birthweight children at 8

years of age. The processing functions examined were those

measured by the Cognitive Assessment System--organization,

simultaneous processing, successive processing, and

attention. Results of statistical analyses are presented

in this chapter.

Experimental Results

Mean raw scores for each of the four areas of the CAS

could not be compared because of differing subtest scales.

For example, some subtest scores were reported in terms of

time taken to complete a task while others were determined

by number of correct responses. Mean raw scores of the two

groups of children for the 16 subtests are reported in

Table 4-1. Inspection of subtest raw scores revealed that

the VLBW sample of children obtained lower scores than the

term-birth group of children in all 16 subtests of the

CAS. Subtest raw scores could not be combined into area

scores without first converting them to z-scores. After

the z-score transformations were performed, subtest z-

scores were added to obtain aggregate area scores for

83










Table 4-1. Comparison of CAS Subtest Raw Scores of Term-

Birth and VLBW Children.


Subtest
1. Visual Search


2. Planned Codes


3. Planned Connections


4. Matching Numbers


5. Design Construction


6. Simultaneous Verbal


7. Figure Memory


8. Matrices


9. Word Order


10.Sentence Repetition


11.Color Ordering


12.Hand Movements


13.Expressive Attention


14.Receptive Attention


15.Auditory Selective
Attention

16.Number Finding


--


Group
Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW

Term
VLBW


Mean Raw Score
121.17 seconds
150.80 seconds

42.57 correct
35.52 correct

264.67 seconds
346.62 seconds

16.87 correct
15.43 correct

6.97 correct
5.86 correct

16.50 correct
16.33 correct

8.67 correct
7.95 correct

9.68 correct
9.48 correct

10.60 correct
10.24 correct

15.93 correct
13.62 correct

9.10 correct
8.29 correct

5.13 correct
4.48 correct

85.63 seconds
94.38 seconds

59.10 correct
54.38 correct

23.77 correct
11.05 correct

50.03 correct
44.19 correct












statistical comparison. Area scores were compared using a

multiple analysis of covariance (MANCOVA).

In this study, four null-hypotheses were proposed to

examine processing functions of VLBW children. The

processing functions of interest were organization,

simultaneous processing, successive processing, and

attention. The organization cluster was composed of the

first four subtests of the CAS (i.e. visual search, planned

codes, planned connections, and matching numbers). The

simultaneous cluster was composed of the next four subtests

of the CAS (i.e. design construction, simultaneous verbal,

figure memory, and matrices). The next four subtests of

the CAS, word order, sentence repetition, color ordering,

and hand movements, comprised the successive cluster

score. The attention cluster score was determined by the

last four subtests (i.e. expressive attention, receptive

attention, auditory selective attention, and number

finding). A multiple analysis of covariance was used to

compare area z-scores of the full-term and VLBW groups of

children across the four areas of the CAS. Three

variables, child's age at the time of testing, maternal age

at the time of birth, and maternal education at the time of

birth, were entered as covariates in the analysis to

control for pre-experimental differences between groups.










Multivariate analysis revealed the presence of

significant differences (F=1.9, p=.041). Therefore,

subsequent univariate F-tests were performed to determine

which variables the two groups differed on. Results

revealed significant group differences in three areas of

the CAS. Distributional characteristics of z-scores for

the four areas of the CAS and results of the multiple

analysis of covariance may be located in Table 4-2.

The MANCOVA procedure yielded an F-value of .36

(p=.78) for the organization cluster. Because the F-value

of .36 was not significant at the .05 level, the first null-

hypothesis could not be rejected. Organization scores of

the two groups of children did not differ at the .05 level

of significance.



Table 4-2. Comparison of CAS Area Scores of Term-Birth
and VLBW Children and Results of MANCOVA.


Standard Univariate
Variable Group Mean Deviation F-Value p
Organization Term .001 1.58
VLBW .574 1.56 .36 .780

Simultaneous Term .000 2.86
VLBW -1.105 3.32 2.90 .045*

Successive Term .001 2.62
VLBW -1.480 2.78 3.05 .039*

Attention Term -0.069 2.015
VLBW -1.550 2.47 4.47 .008*

*p<.05












In the area of simultaneous processing, significant

differences were found to exist between groups. The

MANCOVA was performed with an obtained F-value of 2.9

(p=.045). Consequently, the second null-hypothesis was

rejected because the obtained F-value of 2.9 was

significant at the .05 level. Inspection of group means

revealed that the VLBW group scored significantly lower

than the full-term group in the area of simultaneous

processing.

Scores in the area of successive processing also

differed between the two groups of children. Multiple

analysis of covariance yielded an F-value of 3.05

(p=.039). Because the F-value of 3.05 was significant at

the .05 level, the third null-hypothesis was rejected. The

two groups differed in the area of successive processing.

Inspection of group means enabled the investigator to

determine that the VLBW group scored significantly lower

than the full-term group of children in this area.

Finally, scores in the area of attention were found to

differ between the two groups of children. Multiple

analysis of covariance resulted in a significant F-value of

4.47 (p=.008). Because this F-value was significant at the

.05 level, the fourth null-hypothesis was rejected.

Inspection of group means revealed that the VLBW group











scored significantly lower than the full-term group in the

area of attention.

Summary

Four hypotheses were tested using multiple analysis of

covariance. The results of the MANCOVA procedure were

discussed in this chapter for each of the four dependent

variables investigated in this study. Significant group

differences were found between the full-term and VLBW

groups of children in the areas of simultaneous processing,

successive processing, and attention, with the VLBW

children scoring significantly lower as a group in each of

these three areas. Scores in the area of organization did

not differ significantly between the two groups of

children. Therefore, children born of very low

birthweights were found to have more difficulties than full-

term children at 8 years of age in three of the four

processing functions examined in this study.













CHAPTER V

DISCUSSION, CONCLUSIONS, AND RECOMMENDATIONS

The purpose of this study was to examine cognitive

processing functions of VLBW children at 8 years of age.

The processing functions examined were those proposed by

Luria (1973); specifically, organization, simultaneous

processing, successive processing, and attention. The

Cognitive Assessment System was used to measure these four

dependent variables. Area scores were compared between

groups of VLBW and full-term children using a multiple

analysis of covariance procedure. This chapter includes a

discussion of generalizability limitations, evaluation of

research hypotheses, conclusions, implications, and

recommendations for future research.

Generalizability Limitations

There were several limitations to this study that

could affect the generalizability of results. One

limitation was related to the small sample sizes used in

this study. Although larger samples would have been

desirable, the sizes of those obtained were adequate for

the number of dependent variables examined in the study and

the statistical procedures used. Attempts were made to

strengthen the design by selecting a control group that was










demographically similar to the VLBW group of children in

terms of age, race, gender, and socioeconomic status.

Consequently, the problems created by small sample sizes

were partially controlled by selecting demographically

similar groups of subjects. Additionally, covariates were

used in the data analysis to control for any pre-

experimental group differences between groups. However,

other variables were not controlled and could have been a

source of variance in the present study. Information was

not available on the learning experiences of the children

(i.e. enrollment in early intervention programs, teachers,

schools). Knowledge of these variables would have

strengthened the design of this study. It should also be

mentioned that small sample sizes are frequently a problem

in studies involving long-term follow-up of VLBW children.

The current study was an improvement over previous follow-

up studies that used even smaller sample sizes (Bennett et

al., 1982; Ruff, 1986).

A second limitation to the present study that affects

generalizability of results involved the nature of the

samples obtained. The VLBW group of children obtained in

the current study was not representative of the population

of VLBW children. The VLBW sample of children contained

fewer low socioeconomic children than would be found in the

general population of VLBW children. It was difficult to











locate these children and many of the parents of children

who could be located either did not respond or refused to

participate. It is unknown if parents who agreed to

participate differed from those who could not be located,

did not respond, or refused to participate. Knowledge of

the degree to which these groups differed would have

strengthened this study. Because a large number of VLBW

children are born into lower socioeconomic families, the

degree to which results of the current study can be

generalized to VLBW children in general is questionable.

However, approximately one-third of both the VLBW and full-

term groups of children were from lower socioeconomic

families. Therefore, this group was represented at least

to some degree in both samples. Nevertheless, caution

should be used when relating results of this study to VLBW

children from lower socioeconomic backgrounds. It should

also be mentioned that samples in this study lacked

children from hispanic backgrounds as well as children from

large metropolitan areas. Therefore, caution should be

used in generalizing results of this study to children with

these demographic characteristics.

A third limitation to this study was related to

medical advances that have occurred in the past decade and

those that will occur in the future. Rapid advances in

medical technology will most likely have a profound impact











on long range effects of very low birthweight. Therefore,

the degree to which results of this study will be

generalizable to VLBW children born today is questionable.

Finally, the instrumentation used in this study could

potentially influence generalizability of results. The

significant results obtained in this study were related to

the test used in the study. The test of choice, the

Cognitive Assessment System, was not fully developed at the

time of the study. The final selection of items had not

been accomplished, normative data were not yet available,

and the reliability of scoring procedures had not been

entirely established. The Cognitive Assessment System was

chosen over existing instruments because it had been shown

to provide a better measure of the dependent variables of

interest. However, the efficiency with which the Cognitive

Assessment System measures these variables was still being

investigated at the time of the study. Sufficient research

had been conducted on the Cognitive Assessment System to

establish adequate reliability and validity of test items.

Therefore, the instrument was considered to be acceptable

for the current study.

The provision of a demographically similar control

group in this study should have minimized limitations

resulting from lack of normative data. However, the

control group used in this study was quite small and this