INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

UMI University Microfilms International A Bell & Howell Information Company 300 Nortfi Zeeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800.521-0600

Order Number 9120711

The validity of the Luria Nebraska Neuropsychological Battery-Children’s Revision for children with mild mental retardation

Patterson, Carolyn McCreary, Ph.D.

The Ohio State University, 1991

UMI 300 N. Zeeb Rd. Ann Arbor, MI 48106

THE VALIDITY OF THE LURIA NEBRASKA NEUROPSYCHOLOGICAL

BATTERY-CHILOREN'S REVISION FOR CHILDREN WITH

MILD MENTAL RETARDATION

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

By

Carolyn McCreary Patterson, B.S., M.S.

The Ohio State University

1991

Dissertation Committee: Approved by

0. Hammer, Ph.D.

G . G. Berntson, Ph.D. H. L e l a n ^ Ph.D., Adviser Department of Psychology Regina Gunsett, Ph.D. VITA

July 20, 194:2 ...... Born - Painesville, Kentucky

1966 ...... B.9., Ohio Btate University

1964 ...... M .9., Ohio Btate University

1966-1967 ...... Occupational Therapist Orient State Institution for the Mentally Retarded Orient, Ohio

1967-1969 ...... Occupational Therapist Children’s Hospital Columbus, Ohio

1969-1981 ...... Occupational Therapy Consultant Skilled Nursing Facilities Columbus, Ohio

1372-1976 ...... Occupational Therapist ECCQ Family Health Center Columbus, Ohio

1978-1981 ...... Occupational Therapist Springfield City Schools Springfield, Ohio

1982-1984 ...... Occupational Therapy Instructor Columbus Public Schools Columbus, Ohio

1981-1988 ...... Occupational Therapist Nisonger Center for the Mentally Retarded Ohio State University Columbus, Ohio

1987-1990 ...... Occupational Therapy Instructor School of Allied Medical Prof. Ohio State University Columbus, Ohio

ii 1990-present ...... Postdoctoral Fellow, Pediatric Neuropsychology Children's Hospital Columbus, Ohio

FIELD OF STUDY

Major Field: Psychology

111 TABLE OF CONTENTS

VITA ...... ii

LIST OF TABLES ...... v

LIST OF FIGURES ...... vi

CHAPTER PAGE

I. INTRODUCTION ...... 1 The Research Question ...... 6

II. LITERATURE REVIEW ...... 9 Luria’s Theory of the Functional Organization of the B r a i n ...... 9 Developmental Neuropsychology ...... 14 Brain Dysfunction in Persons with Mental Retardation ...... 22 Brain Dysfunction and Spina Bifida .... 23 The Luria Nebraska Neuropsychological Battery-Children’s Revision...... 26 The Wisconsin Card Sorting T e s t ...... 31 Trail Making T e s t ...... 35 AAMD Adaptive Behavior Scale - School E d i t i o n ...... 36

III. METHODOLOGY . 40 P u r p o s e ...... 40 S u b j e c t s ...... 40 Procedure ...... 41

IV. RESULTS AND CONCLUSIONS ...... 43 O v e r v i e w ...... 43 Nature of Neuropsychological Function . . 45 Clinical and Statistical Agreement .... 58 Myelomeningocele Statistical and Clinical M e m b e r s h i p ...... 65 Adaptive Behavior ...... 65 S u m m a r y ...... 75

V. DISCUSSION ...... 78

LIST OF REFERENCES ...... 94

APPENDIX A ...... 101

iv LIST OF TABLES

TABLE PAGE

1. Cluster means, standard deviations, and ranges for Cl through S 3 ...... 47

S. Clusters defined by LNNB-C scale means, plus and minus one standard deviation . . . 51

3. LNNB-C scales contributing to significant differences between cluster pairs ...... 52

4. Cluster frequencies for TUT part A ...... 57

5. Cluster frequencies for TUT part B ...... 57

G . Clinical category frequency ...... 61

7. Typical clinical protocol, B41 ...... 62

8. Typical clinical protocol, N B l ...... 64

3. Mean ABS-SE factor scale scores for clusters ...... 67

10. Cluster frequency of ABS-SE factor scale scores in the deficient r a n g e ...... 69

11. Mean ABS-SE factor scores for clinical c a t e g o r i e s ...... 71

12. Frequency of ABS-SE factor scale scores in the deficient range for clinical c a t e g o r y ...... 72

13. Mean ABS-SE factor scale scores for diagnostic category ...... 73

14. Frequency of ABS-SE factor scale scores in the deficient range for diagnostic c a t e g o r y ...... 74 LIST OF FIGURES

FIGURES PAGE

1. Cluster profiles, based on average LNNB-C I score distance from critical l e v e l ...... IB

VI CHAPTER I

INTRODUCTION

The use of neuropsychological assessment with a devel­ opmental ly disabled school age population, including persons with mild mental retardation, is growing in importance.

This development is based predominantly on a basic theoreti­ cal rationale, with little direct evidence available regard­ ing practical application or the validity of specific neuro­ psychological assessments with these populations (Fischer,

1988; Skoff, 1988). Utilization of neuropsychological assessment with persons with mental retardation or with developmental delay can contribute on two levels: 1) knowl­ edge related to the nature of the processes and the possible amelioration of specific patterns of deficit in intellectual functioning, and 2) knowledge regarding the ontogenetic development of neuropsychological functions (Fischer, 1988;

Skoff, 1988; Fletcher & Taylor, 1984).

Knowledge related to the nature and possible ameliora­ tion of deficits in intellectual functioning can contribute to the development of optimal programming. School age chil­ dren with mild mental retardation are not a homogeneous

group and optimal educational strategies will vary among 2

individuals. Neuropsychological assessment can assist in

identifying relative neuropsychological strengths and weak­ nesses which can serve as a basis For determining optimal

instructional/habilitâtion strategies.

A developmental neuropsychological approach to assess­ ment within these populations has additional importance.

Knowledge regarding the ontogenetic development of neuropsy­ chological functions can contribute to basic information re­ garding normal child neuropsychological development, provide a basis for prediction of "risk" for disability and a basis for long term program planning. Developmental neuropsy­ chology places an emphasis on change as a unit of analysis.

Assuming this emphasis, neuropsychology can focus on how developmental disabilities disrupt processes of systematic change CFletcher & Taylor, 19B4; Fletcher, Miner, & Ewing-

Cobbs, 1987). Fletcher and Taylor C1984) propose a "func­ tional organization approach" to developmental neuropsychol­ ogy . Their approach takes into consideration the manifest form of a disability and correlated abilities/disabilities, which covary with a set of moderator variables including environmental characteristics and social influences. Al­ though Fletcher cites Biegel, Bisanz, and Bisanz C19B3) and

Werner and Kaplan C1956) as examples of theories which view development as a process of increased differentiation and hierarchical organization, and thus focus on change, the earlier theories of Pavlov and Vygotsky provide a stronger 3 basis for current developmental neuropsychology. Vygotsky

(19G2), building on Pavlov’s concept of functional organiza­ tion, CPavlov, 13573 proposed criteria for an approach to the study of neurological function as it relates to behav­ ior. He believed that the approach should be developmental, should address the relationship between higher mental func­ tions and lower, elementary psychological functions, and should take into consideration socially meaningful activity as an explanatory principle (Kozulin, 19BBD. Luria, with a basis in Pavlov and Vygotsky’s theories, proposed that higher cortical functions are "social in origin, mediate in structure, and conscious and voluntary in mode of function"

(Luria, 1956, p. 30). Fletcher and Taylor’s proposed func­ tional organization approach is not in conflict with Luria and Vygotsky’s theoretical approach. In Fletcher’s and

Taylor’s view the central nervous system influences the manifest disability via the limits imposed on the basic competencies, or on functional organization. The contribu­ tion of the environment to functional organization and an emphasis on change over time are inherent in both Luria’s and Vygotsky’s theories and in Fletcher and Taylor’s ap­ proach .

Analysis of change in various neuropsychological func­ tions can contribute to our basic knowledge regarding the development of higher cortical functions as well as the nature of the various intellectual growth problems leading 4 to mental retardation. Analysis of change over time within a population of school age children with mild mental retar­ dation may be applied to hypothetical subgroups, such as children with apparent specific neurological deficits or relatively focal dysfunction, and children with neuro­ psychological ly based maturational delay or lag in rate of developmental learning. These subgroups may be further broken down into types of dysfunction and degree of delay, respectively. Analysis of change may also be applied to various diagnostic groups, such as myelomeningocele with associated hydrocephalus. Fischer (1988) stated:

To the extent that neuropsychological procedures meas­

ure cerebral pathology and a wider range of cognitive

and behavioral functions than routine tests of intelli­

gence or adaptive behaviors, they can be useful for

describing the specific functional capacities of the

mentally retarded as well as the involvement of central

nervous system dysfunction, and in doing so, provide

more comprehensive information for rehabilitation

and/or treatment. Cp. 151)

It is believed that the question of cerebral pathology referred to by Fischer should be applied to this population in a manner other than that traditionally found in the neuropsychological literature. Rather than attempting to establish membership in hypothetical etiological groups of brain dysfunction versus environmental or cultural-familial, 5 or of cerebral pathology versus non-cerebral pathology, one might be better served by initially asking what the nature of neuropsychological function is in this population. This nontraditional question has a basis in the fact that not all neuropathology is clinically identifiable, nor does the clinical approach usually take into consideration the multi­ variant influences on neuropsychological development over time. In order to study change over time in this population one might initially attempt to identify subgroups which reflect differential neuropsychological functioning and/or development. A combination of statistical and clinical analysis might reveal subgroups reflective of specific neuropsychological deficit or specific focal dysfunction versus neuropsychological based maturational delay or lag in rate of developmental learning. In using the terminology

"specific neuropsychological deficit or specific focal dysfunction" and "neuropsychological based maturational delay or lag in rate of developmental learning" an attempt is made to differentiate between children with damage or dysfunction in specific areas resulting in identifiable neuropsychological strengths and deficits versus children with slowed development due to a generalized delay in neuro- developmental processes, general diffuse dysfunction, or to environmental deprivation. It is expected that this initial neuropsychological subtyping may be important to the devel­ opment of a means of identifying groups of children who will 6 respond differently to various intervention strategies in terms of the facilitation of an optimal outcome for each child. Given the imperfect correiation between psycho- metrical ly measured intellectual function and adaptive behavior, especialiy at the upper range of mental retarda­ tion (Grossman, 19B3), it is also of interest to determine the relationship between measures of neuropsychological function and adaptive behavior.

THE RESEARCH QUESTION

Although clinical utility is predicted for the Luria

Nebraska Neuropsychological Battery - Children’s Revision

CLNNB-C) (Golden, 1SB7) with persons having psychometrically measured intellectual function in the 50 to 70 IQ range, it is believed that interpretation based on the standard deci­ sion rules utilized with the LNNB-C will not be useful and that the traditionally initial research question regarding the utility of the battery for differentiating between non­ brain dysfunctional and dysfunctional developmentally handi­ capped children is inappropriate. The belief regarding the decision rules is based on the expectation that children with developmental delay will miss items heavily loaded on educational level and thus will demonstrate a profile indi­ cative of brain dysfunction according to standard decision rules. The belief regarding the inappropriateness of the traditional research question is based on the expectation 7 that not all signs of neuropathology are clinically identi­ fiable, and on the basis of the multivariant influences on neuropsychological development over time. It is expected that statistical analysis of the performance of this popula­ tion on the battery will result in the identification of subtypes and that these statistically identified subtypes may or may not correlate with categories of neuropsychologi­ cal function established by clinical interpretation. It is also expected that the statistical subtypes will demonstrate within group similarities and between group differences in adaptive behavior profiles. A diagnostically homogeneous subgroup of the sample will provide an initial opportunity to determine whether a relationship might exist between neurological diagnosis and statistical subtypes, clinical categories and adaptive behavior profiles.

Research Questions

1. Will the experimental population differ in their clini­ cal neuropsychological manifestations and statistically cluster into subgroups?

2. Will the statistically clustered subgroups demonstrate agreement with clinical diagnostic categories established by clinical analyses?

3. Will the medically diagnostically homogenous subgroup of the sample cluster together on statistical analysis and will they belong to the same clinically established diagnostic 8

category?

4. Will the statistically clustered subgroups demonstrate

within group similarities and between group differences in

adaptive behavior?

5. Will clinical categories demonstrate within group simil­

arities and between group differences in adaptive behavior?

5. Will the medically diagnostically homogenous group demon­

strate similar adaptive behavior?

Definition of Terms

The population: school age children, 3 through 13

years of age, with psychometrically measured intellectual function in the 50 through 70 range.

Clinical neuropsucholooical manifestations: neuropsy­ chological function as measured by the Luria Nebraska Neuro­ psychological Battery-Children’s Revision (Golden, 1SG7).

Statisticallu cluster; form groups with the utiliza­ tion of cluster analysis.

Medicallu diannosticallu homogeneous group : myelome­ ningocele with associated hydrocephalus.

Clinical analusis: clinical interpretation based on item analysis adapted from Luria's original model (Luria,

19BB5.

Adaptive behavior; as measured by the AAMD Adaptive

Behavior Scale - School Edition (Lambert, ISBl). CHAPTER II

LITERATURE REVIEW

LURIA'S THEORY OF THE FUNCTIONAL ORGANIZATION OF THE BRAIN

Luria (13GG) in his historical survey of the develop­ ment of the concept of functional organization describes the development of the study of localization of functions in the cerebral cortex. He characterizes the localization and antilocalization positions as "attempts to identify complex mental processes with the material structure of the brain"

(Luria 19GG, p. 18). He contrasts and describes the two opposing approaches as:

one attempting to relate mental processes to circum­

scribed areas of the brain, with the brain regarded

as an aggregate of separate organs, and the other

assuming that mental activity is a single, indivisible

phenomenon, a function of the whole brain working

as a single entity. Cp. 19)

Luria credits both positions with making the following con­ tributions to the study of the brain: the localizationists,

1) for regarding the brain as a differentiated organ, 2) for utilizing an analytical approach, and 3) for considering the brain as an app%2r6 ntly homogeneous mass but with specific 10 areas especially significant in the performance of various forms of mental activity; the antilocalizationists, 1) for

insisting that the brain, although a highly differentiated organ, always functions as a whole, thus contributing to concepts of plasticity and cortical tone, and 2) for con­ ceiving the vertical organization of functions and thus, the need for hierarchical analysis of relationships.

These two positions both looked upon mental functions

"as phenomena to be directly correlated with brain struc­ ture" Cp. 20). The meaning of the word "function" at this time "was taken to mean the functions of a particular organ"

Cp. 20). In contrast, in the context of Luria's theory, the concept of function refers to an aggregate of complex temp­ orary connections rather than a direct property of a partic­ ular center of the brain. This concept of function has an origin in the work of Pavlov C1957). Luria C1966) states :

Since Pavlov advanced his reflex theories, the word

"function" has come to mean the product of complex

reflex activity comprising: uniting excited and

inhibited areas of the nervous system into a working

mosaic, analyzing and integrating stimuli reaching the

organism, forming a system of temporary connections,

and thereby ensuring the equilibrium of the organism

with its environment. Cp. 21)

Vygotsky proposed that while reflexes provide the founda­ tions of behavior, nothing can be learned from them about 11

the construction erected on the foundation. He believed

that a new scientific psychology had to be developmental,

had to resolve the problem of interrelation between higher

mental functions and the lower elementary psychological

functions, and that it had to take into consideration so­

cially meaningful activity as an explanatory principle

(Kozulin, 1986).

Luria C1SS5) basing his work on Pavlov and Vygotsky, developed a theory which was neither one of localization nor of equipotentiality, but of pluripotentiality reflected in the concepts of functional systems and alternate functional systems. Functional systems are described as mosaic pat­ terns of interacting areas of the brain which coordinate, through arousal and inhibition, to produce a given behavior.

The analogy of a chain with links has been used to describe how these functional systems work CPlaisted, Gustavson,

Wilkening, & Golden, 1983). Each link represents the con­ tribution of a specific area of the brain to a behavior. If one area is dysfunctional, the behavior is impaired. Luria, using the concept of pluripotentiality, proposes that any given specific area of the brain can contribute to several different functional systems. The concept of alternate functional systems proposes that multiple functional systems can be responsible for a given behavior, thus if one system becomes impaired another system may mediate the behavior of the impaired system. IE

In addition to providing a synthesis of localization and antilocalization or equipotential positions, that is, the concept of pluripotentiality, Luria’s theory provides a foundation for developmental neuropsychology. Luria C1566D cites Vygotsky’s studies of the development of the higher forms of mental activity in the course of ontogenesis and concludes:

...at successive stages of their development the struc­

ture of the higher mental functions does not remain

constant but that they perform the same task by means

of different, regularly interchanging systems of con­

nections. Cp. 31)

Luria and Vygotsky postulated a sociohistorical origin of higher cortical functions, phylogenetically and ontogeneti- cally. Luria describes the structure of the functional sys­ tems as consisting of three units: Unit I. mediating arous­ al and attentional processes ; Unit II. sensory reception and integration; and Unit III, motor execution, planning and evaluation. Unit I consists of the reticular activating system and brain stem; Unit II, the primary, secondary and tertiary zones of the temporal, parietal, and occipital lobes; and Unit III, the motor strip, premotor area and prefrontal area of the frontal lobes (Luria, 1973).

Not without controversy, based on Luria"s classifica­ tion of functional units and of primary, secondary and tertiary zones. Golden and Wilkening C19B1) discuss five 13

stages of sequential brain development: stage one, the

development of the reticular activating system, generally

developed at birth and fully operational by 12 months post­

conception; stage two, the primcry zones of Units II and

III, concurrent with stage one; stage three. the secondary

areas of Units II and III, concomittant with stage one and

two but extending through approximately age 5 years; stage

four, development of the tertiary zone of Unit 11, beginning

at age 5 to 0 years and extending to age 12; stage five,

development of the tertiary area of Unit III, beginning at

10 to 12 years of age and continuing into the early 20’s.

Controversy has emerged in respect to Golden’s fifth stage

and the development of prefrontal function. Many authors

propose that prefrontal functions emerge much earlier, and

in fact, Luria C1973) stated that:

... the prefrontal regions of the cortex do not mature

until very late in ontogeny, and not until the child

has reached the age of four to seven years do they

become finally prepared for action. Cp. 87)

Perhaps the controversy has some basis in differing domains of reference. Luria makes reference to maturation of pre­ frontal regions, apparently a reference to physiological changes. Golden and Wilkening (1981) explicitly state:

...times for various developmental periods given here

are based on behavioral rather than physiological

observations. As such, they are subject to change as 14

our understanding increases and are not to be seen

as rigid or essential to the basic theory." Cp. 72)

Other investigators propose that physiological development continues until about age 13 (Reines & Goldman, 1980).

Nevertheless, Investigations of the development of prefron- tal regions and functions continue, with many investigators of behavioral functions believed to be subserved*'by the prefrontal areas referring to the impairments demonstrated as "prefrontai-type" deficits. In clinical practice, the

LNNB-C is being supplemented with the Wisconsin Card Sorting

Test and the Halstead-Reitan Trailmaking Test, in order to measure these "prefrontal-type" functions.

Thus, Luria’s theory of higher cortical function in man regards the brain as having a pluripotential organization which is influenced by social-cultural history, with the structures of functional organization changing with succes­ sive stages of development.

DEVELOPMENTAL NEUROPSYCHOLOGY

Child neuropsychology is defined by Rourke, Bakker,

Fisk, and Strang (1983) as the study of brain-behavior relationships in the developing human organism. Fletcher and Taylor (1984) emphasize that in order for chid neuropsy­ chology to be truly developmental, the processes of change must be addressed. The knowledge base for developmental neuropsychology is drawn from many disciplines including 15

developmental psychology, pediatrics, neurology, biology,

ethology, behavioral embryology, developmental neurobiology

CSpreen, Tupper, Risser, Tuokko, & Edge11, 13841, and behav­

ioral teratology CCoyle, Wayner, & Singer, 13801. Develop­

mental neuropsychology appears to have a theoretical bases

in the integration of multiple developmental and neuropsy­

chological sources. Luria frequently cites Piaget when de­

scribing developmental aspects of function (Luria, 13661.

Fletcher and Taylor (13841 in describing their approach cite

Siegal, Bisanz, and Bisanz (13831 and Werner and Kaplan

(13581 as theories which view development as a process of

increased differentiation and hierarchical organization and address change. They urge that theories such as these be

incorporated into the field of developmental neuropsycholo­ gy .

Luria and Vygotsky’s theories with their basis in

Pavlov are examples of neuropsychological theory which con­ tains a strong developmental element. The theoretical aspect of developmental neuropsychology appears to have its roots in Vygotsky’s theory (Vygotsky 1380, cited in Luria,

1388; Vygotsky, 13851. Based on his investigations Vygotsky

(13851 formulated the following “rule":

In disturbances occurring in early stages of develop­

ment resulting from a local brain lesion it is the

nearest higher center which suffers the most, whereas

the nearest lower center suffers less. In local brain IB

lesions to a mature brain it is the nearest lower

center which functionally depends on the higher zone

which suffers primarily, whereas the nearest higher

center, which became independent in the course of

development and which functions at a regulatory higher

level, suffers less. Cp. 3B4D

Thus, identical Cor similar) syndromes in children and adults can be the result of different lesions, and identical lesions in children and in adults can result in very dif­ ferent disorders CVygotsky, 1965). In the immature brain lower levels are basic for the development of functions subserved by the higher levels; that is, simple functions form the basis of more complex functions. Vygotsky also observed that in the immature brain compensatory functions are performed by lower centers, in the mature brain, by higher centers. In the first stages of development the higher forms of psychological functions are intimately con­ nected with external activities, later these functions become interiorized, taking the shape of inner mental activ­ ities. Exteriorization of functions is one of the most efficient means of compensation. Luria credited Vygotsky with these theoretical formulations stating that a fulfill­ ment of his previsions in a series of empirical investiga­ tions remains a task for the future, that is, "the careful study of the paths of development of higher psychic func­ tions in ontogenesis" (Luria, 19B5, p. 391). It is this 17 task that is beginning to be addressed by empirical studies in developmental neuropsychology. Vygotsky made a further major theoretical contribution to the study of developmental psychology in identifying the importance of social history as an origin of higher cortical/psychological functions

(Luria, (1965). Thus experience becomes a major variable for the study of developmental neuropsychology.

Bolter and Long (1985) credit Margaret Kennard with one of the earliest research efforts addressing age-related effects in brain damage, based on studies with monkeys. Her work contributed to the recognition that injury to an imma­ ture brain differs in effect than similar injury to a mature brain. Bolter and Long also cite the classic work of Teuber and Rudel (1962) on the effects of early damage which led to the postulation of three possible deficit patterns: initial deficits which resolve, initial deficits which persist, and deficits which appear only later in development. General consensus at this time is that a variety of variables inter­ act to determine the effect of brain damage on function: age at insult; locus of the injury; the severity, etiology and chronicity of the damage; the tasks utilized to measure the effects of injury; gender differences; and pre-injury and post-injury experiences.

Empirical studies exploring various aspects of the development of neuropsychological function in children currently abound. Although Spreen at al. (1984) provide a 18 comprehensive overview of the literature in the field, a more recent comprehensive review is not evident. Studies serving to exemplify current research include Rourke, 1982;

Snow, Barnett, Cunningham, and Ernst, 1988; and Harbord,

Finn, Hall-Craggs, Robb, Kendall, and Boyd, 1880.

Rourke (1882) established support for the development of a comprehensive model for studying central processing deficiencies in children. Observing that not enough is known about the relevant parameters of learning in children, he maintains that a coherent theoretical structure is needed in order to explain previous findings and to guide future research. His proposed model encompasses three principal axes of concern in human brain-behavior relationships; the progression from lower to higher centers, from posterior to anterior regions, and from right to left hemispheres.

Rourke addresses only the right to left hemisphere progres­ sion. Drawing on the theoretical position of Goldberg and

Costa (1881) he develops principles which fit with previous findings:

1) there is an ontogenetic progression from the sa­

lience of right-hemisphere functions to that of left-

hemisphere functions; (2) the evident change in child­

ren’s conceptualizations from global to specific is a

reflection of the right- to left-hemisphere ontogenetic

development; (3) the development of right-hemisphere 19

systems is a prerequisite for the adequate develop

ment of left-hemisphere systems; C41 in the normal

course of affairs in the formation of constructs and

concepts, right-hemisphere systems provide the content

for concepts whereas left-hemisphere systems are par­

ticularly geared to the articulation, elaboration, and

stereotypic application; and (5) diminished access to

disordered functioning of right-hemisphere systems is

especially debilitating with respect to the development

of adaptive abilities. (Rourke, 1982, p.4)

Snow et al. (1988) utilized cross-modal (tactile-visu­

al) discrimination and memory tasks to study cross-modal

development in normal and learning disabled children.

Developmental trends identified support a cortical matura­

tion period in the 8 to 9 year range, consistent with Lur­

ie’s theory and with previous research. Differences between

normal and learning disabled groups were concluded to sup­

port a developmental lag position, or a functional reorgani­

zation position.

Harbord et al. (1980) investigated the myelination

patterns, on magnetic resonance, of 30 developmentally

delayed children ages 6 months to 10 years. Four categories were identified; normal myelination, globally delayed

myelination, topographical distribution appropriate for age but white-matter signal intensity more appropriate to a 20

younger child (suggestive of white matter immaturity), and

patchy focal abnormalities.

These studies are representative of some of the current

lines of research. The integration of findings from lines

of research focusing on structural and on functional devel­

opment, and of longitudinal studies combining structural and

functional development can result in a coherent body of

knowledge regarding the development of brain-behavior rela­

tionships. These types of studies can be applied to normal

children as well as to children at risk or with known in­

sults and developmental delay.

Theoretical and empirical literature reflect several

issues specific to developmental and child neuropsychology

versus adult neuropsychology. The application of neuropsy­

chological assessment to children in comparison to adults

demands special considerations. Both Rourke et al. (1983)

and Fletcher and Taylor (1984) discuss areas of caution which must be observed in the application of neuropsycholog­

ical assessment with children. Generally these cautions are related to the fact that the central nervous system of the

child is quite different from that of the adult both physio­

logically and functionally, that the child's nervous system

is in a state of rapid change, and that systematic anomalies may produce atypical and unique brain development in the

individual child (Rourke, et al., 1983). Fletcher and 21 Taylor (198%) discuss these cautions in the form of fal­ lacies concerning the nature and strength of inferences that can be made about a child’s brain based on behavior: dif­ ferential sensitivity fallacy, similar skill fallacy, spe- cial-sign fallacy, and brain-behavior isomorphism fallacy.

These fallacies are described as being based in a failure to recognize differences between child and adult, faulty logic, and inadequate research. Given the short history of the empirical study of brain-behavior relationships in children not a great deal is understood regarding the mechanisms that subserve complex behavior in the developing organism, and what has been learned has apparently not been rendered into a holistic or meta-analysis.

Luria states that different mechanisms will subserve various behaviors or functions during a child’s development as well as during the development of a given skill or func­ tion (Luria, 19BB). Thus a simplistic downward extension to children of the adult model of neuropsychological function is inappropriate. However, although Luria’s model may not be appropriate for the purpose of localizing dysfunction in children, it can be utilized to conceptually organize func­ tion and dysfunction in children. This is feasible because of Luria’s description of functional systems based on sub- servance by primary, secondary, and tertiary areas, with less complex functions being subserved by primary areas, more complex functions subserved by secondary areas, and the 22 most complex functions subserved by the tertiary areas.

Inferences regarding the localization of dysfunction and/or the developmental history of dysfunction, especially with children with early and/or unspecified insults, await the development of an integrated knowledge base regarding behav­ ioral development and the development of the functional systems in normal children and children with neurological and/or neurodevelopmental anomalies.

BRAIN DYSFUNCTION IN PERSONS WITH MENTAL RETARDATION

Spreen, Tupper, Risser, Tuokko and Edgell (1984) in their discussion of neuropsychological issues and neuro- pathological considerations relevant to persons with mental retardation cite studies which suggest that a large propor­ tion of cognitive deficits in the mentally retarded have a basis in neuropathology (approximately 70%) and that a fair amount of persons in the mildly handicapped range also have neuropathology (approximately 30%). It is stated that the brain damaged child with mental retardation is not a diag­ nostic entity and that there is considerable variability from one individual to another. The presence of hetero­ geneity can be supported by a recent study of evoked poten­ tials in persons with mild mental retardation (Gasser,

Pietz, Schellberg, Kohler, 1988). An additional means of identifying subtypes of brain dysfunction is suggested by

Tramontana and Sherrets (1985) in their study of brain 23 impairment in psychiatric disorders. They propose that children with neurological delay, as well as those with diffuse neurological dysfunction will demonstrate increased ventricular-brain ratios on CT scan, while persons with more localized lesions will demonstrate greater density varia­ tions. Identification of subtypes of brain dysfunction within the mild range of mental retardation may be a more appropriate ultimate validation effort than that of merely differentiating the neurologically impaired and nonimpaired, especially if the subtypes can ultimately be demonstrated to benefit from differential educational and habilitative strategies. Strategies related to developmental sequence at a slower rate may be most appropriate for the delayed. The diffusely impaired may benefit from the addition of specific training strategies. Children with relatively specific impairments may benefit from the devleopment of and training in the use of alternate functional systems and/or the use of external compensatory strategies.

BRAIN DYSFUNCTION AND SPINA BIFIDA

Spina bifida, the failure of the bony spinal canal to close prenatally, is sometimes accompanied by meningocele

(herniation of the meninges through the bony defect) or by meningomyelocele (herniation of the meninges and the spinal cord through the bony defect). Meningomyelocele results in paralysis and sensory loss below the level of the protrusion S4 and In bowel and bladder dysfunction. Frequent complica­ tions include infection and hydrocephalus requiring shunting

(Robbins, Cotran, and Kumar, 1984). Other frequently asso­ ciated abnormalities are protrusion of the cerebellum and the medulla into the upper spinal canal CArnold-Chiari syn­ drome), an S-shaped kinking of the medulla, underdevelopment of parts of the visual and auditory systems and some of the cortical gyri, and clusters of nonfunctional neurons in many parts of the brain due to aberrant migration of cells during embryogenesis (Spreen et al., 1984).

The developmental outcome of children with meningomy­ elocele has been widely studied. Past studies, compared to more recent investigations, have generally demonstrated a relatively negative outcome. In the past, as a group, children with meningomyelocele and hydrocephalus have demonstrated mild to moderate retardation, impaired concen­ tration and visual-perceptual difficulties, and impaired fine motor skills (Mazur, Aylward, Colliver, Stacey, and

Menelaus, 1988; Spreen et al., 1984). Typically performance scores were lower than verbal scores, although verbal skills have been characterized as being at a superficial level

(Spreen et al., 1984).

More recent studies report a more positive outcome.

This difference in apparent outcome may be attributable to both improved medical management, and refinement of research design. Simultaneous early repair of myelomeningocele and 25 simultaneous insertion of ventriculoperitonal shunt has been reported to result in an outcome of 7 of 10 subjects with normal or slightly below normal intelligence and psycho­ social development CHubballah and Hoffman, 1387). In regard to improved research design, grouping of patients into diagnostically homogeneous groups yields differential out­ come information. The results of a study correlating both number of shunts and level of lesion with intelligence, in

82 children, were in agreement with previous studies sug­ gesting that children with meningomyelocele tend to have lower intelligence quotients. However, children with less than three shunt procedures had significantly lower verbal, performance, and full scale IQ scores than children with no shunt procedures. Children with lesions below the thoracic level had significantly higher verbal, performance, and full scale IQ scores than those with lesions at the thoracic level CMazur et al., 1388). In children without hydrocepha­ lus intellectual function varied with lesion level. In children with hydrocephalus the impact of lesion level became insignificant, especially when multiple shunt revi­ sions were necessary due to shunt failure from blockage or infection. A previous study finding lower IQ scores in shunted children with meningomyelocele attributed this to the occurrence of CNS infections CMcLone, Czyzewski, Raimon­ di, & Sommers, 1382). Children with thoracic level lesions on the average had IQ scores below normal; children with 2G

upper lumbar, lower lumbar and sacral level lesions fell in

the normal range (80 to 100); sacral level patients had the

highest IQ scores. In addition to organic neurological

complications, sitting balance, joint contractures, socio­

economic factors, and educational opportunities were consid­

ered to be relevant variables influencing outcome (Mazur et

al., 1906).

Although children with spina bifida and meningomyelo­

cele as a group tend to demonstrate mild to moderate retar­

dation and poorer visual-motor perception than normal chil­

dren, the outcome is variable with a basis in both primary

and secondary variables. Continued improvement in medical

techniques and interventions as well as improved social-

educational interventions may result In continued improve­

ments in developmental outcome.

THE LURIA NEBRASKA NEUROPSYCHOLOGICAL BATTERY-CHILDREN'S

REVISION

The Luria Nebraska Neuropsychological Battery (LNNB)

and the Luria Nebraska Neuropsychological Battery - Child­

ren’s Revision (LNNB-C) represent attempts to standardize

the clinical investigative techniques of A. R. Luria. Lur­

ia ’s qualitative clinical investigative technique consisted of selecting simple items based on a patient’s response to

the previous item or items, following missed items with

items which measured skills thought to be mediated by the 27 same area of the brain. This procedure was followed until a diagnosis was confirmed. The items in the Luria-Nebraska batteries represent an attempt to test all the principal alternate functional systems by testing their specific links. An initial attempt to standardize Luria’s technique was made in 1975 by Christensen and resulted in a battery that was considered to be excessively lengthy and not com­ pletely standardized. The battery was subsequently modified by Golden and Colleagues, resulting in a briefer and fully standardized battery, the LNNB CPlaisted, Gustavson, Wil- kening, and Golden, 1983). The LNNB-C is a downward exten­ sion of the adult battery (Golden, 1987). It has been suggested, due to the elimination of items intended to measure behaviors mediated by the prefrontal areas, that the

LNNB-C be routinely supplemented with Halstead Reitan's

Trail Making Test Part A and Part B and the Wisconsin Card

Sorting Test (Greta Wilkening, personal communication, June,

1987).

The LNNB-C consists of 149 items grouped into 11 clini­ cal scales. The normative sample consisted of 125 normal children, 25 at each age level between 8 and 12. The 11 clinical scales do not measure unitary skills which could be inferred from the scale name. However, items on a particu­ lar scale will have a component of the skill named. The 11 clinical scales are Cl, Motor; C2, Rhythm; C3, Tactile; C4,

Visual; C5, Receptive Speech; CB, Expressive Speech; C7, 28

Writing; CB, Reading; C9, Arithmetic; CIO, Memory; and Cll,

Intellectual Processes. Internal consistency reliability

for the LNNB-C is generally high, ranging from the low 7 0 ’s

to the mid 90's, based on a brain-impaired or heterogeneous

sample. Reliability is lower for nonimpaired samples (Gold­

en, 19875.

The manual cites initial studies supporting the valid­

ity of the LNNB-C for differentiating impaired and non­

impaired groups. Gustavson et al. C19B45 established age effects, the critical level formula, the decision rule re­ garding number of scale scores above critical level, and validated these regarding their ability to discriminate between normal and brain-injured children. The authors caution against use of actuarial rules alone and urge that effective diagnoses require detailed analysis of the indivi­ dual subtests and observations of the examiner, including observations of how subjects perform tasks and the type of error made. The assessment of long-term and delayed memory are cited as limitations of the instrument (Gustavson, et al., 19845.

The LNNB-C has been validated to discriminate between normal, psychiatric and neurologically impaired children

(Carr, Sweet, & Rossini, 198B5. Scales with optimal dis­ criminant function were Motor, Rhythm, Visual, Expressive

Speech, Reading, and Arithmetic. No incremental discrim­ inative ability was found for the LNNB-C in comparison to 29

the UIISC-R. UIISC-R subtests with optimal discriminative

ability were Similarities, Arithmetic, Comprehension, Block

Design, Object Assembly and Coding. The authors caution

that these discriminative abilities be considered suggestive

rather than conclusive due to small sample size. Another

validation study found no discriminative ability with con­

duct disordered children, supportive of a hypothesis that

conduct disorder is heterogeneous in etiology CLevandowski,

1985).

Validation studies with learning disabled and non­

leaning disabled children have supported the LNNB-C’s abili­

ty to discriminate between these groups (Geary & Gilger,

1984; Geary, Jennings, Schultz, and Alper, 1984). A study designed to determine the discriminant validity of the LNNB-

C between groups of learning-disabled children found that although the LNNB-C discriminated between normals and read­

ing-spelling disabled children, it did not discriminate between normals and math disabled children. Due to method­ ological difficulties in this study the authors cautioned that the results be considered suggestive rather than con­ clusive (Nolan, Hammeke, and Barkley, 1983). It may be of

interest to note that this study considered only signifi­ cance of quantitative differences between groups, a quali­ tative item analysis, as suggested by Gustavson et al.,

1984, may have been successful in identifying differences.

Other studies finding poor differential diagnoses for 30

subtypes of learning disability CSnoui S Hynd, 1985; Snow,

Hynd, & Hartlage, 15841 also use only actuarial analysis.

It appears that qualitative analysis is relatively absent in research studies. Snow et al. (1984) found significant differences between a group of learning disabled children placed in self contained classes (more severe; mean full scale 10 82.60) and a group of learning disabled children placed in regular classrooms with resource room services

(less severe; full scale IQ 91.70) only on academic scales

(receptive speech, writing, reading, arithmetic) and thus concluded that the LNNB-C does not discriminate between these groups. This conclusion assumes that the "academic scales" are not based on an intent to measure neuropsycho­ logical functions, when in fact a review of Luria (1966) would reveal the intent of the items and their neuropsycho­ logical significance. It is possible to qualitatively analyze an academic scale to determine the hypothetical neuropsychological significance of a subjects performance.

Although validity for the LNNB-C in differential diag­ nosis of normal versus brain dysfunctional, psychiatric versus neurologically impaired, and normal versus learning disabled children has been established it is not without controversy. A failure to heed the suggestion regarding the use of qualitative analysis by Gustavson (1984) is evident in the research designs. 31

THE WISCONSIN CARD SORTING TEST

The Wisconsin Card Sorting Test (WCST) is a test origi­

nally developed by E . Berg to assess abstract reasoning in

adults. After utilization for a number of years in re­

search, the WCST was standardized and formally published as

a clinical neuropsychological instrument in 1381 by R. K.

Heaton CChelune and Baer, 1386). The WCST utilizes stimulus

and response cards displaying figures of varying form, number, and color. Subjects are to sort the cards based on

either color, form, or number, without direct knowledge of the correct sorting principle. Subjects are informed as to

the correctness of each response. The examiner shifts response set after the subject makes 10 consecutive correct responses.

Heaton’s normative sample included 150 normal individu­ als, 123 males and 27 females, with a mean age of 35.3 years, standard deviation 15.3, mean education 13.3 years, standard deviation 3.2. The brain damaged group consisted of 208 individuals with structural cerebral lesions, (1^5 males and 63 females), 31 with diffuse lesions and 111 with focal lesions. Mean age was 42.1 years, standard deviation

16.1, mean years of education was 12.7, standard deviation

3.2. Test score cutoffs are utilized for diagnostic predic­ tions in two categories, focal frontal versus nonfrontal damage and brain damage versus non-brain damage. The best diagnostic predictions based on the normative study were the 32 perseveration response scores (Heaton, (1981). Chelune and

Baer (1986) published developmental norms for the WCST.

Tentative norms for school aged children were developed based on a sample of 105 school aged children, 8 years to 12 years of age for Categories Achieved (concept formation requiring the use of positive and negative feedback to formulate problem-solving strategies, Perseverative Errors

(the inability to suppress ongoing activity despite environ­ mental feedback that it is no longer appropriate), and

Failures to Maintain Set (increased reactivity to extraneous stimuli). The results of this study were interpreted to indicate that children make rapid developmental gains in number of Categories Achieved and significantly reduce the number of Perseverative Errors with advancing age, with

Failure to Maintain Set showing improvement after age nine.

The performance on WCST was indistinguishable from adult performance by age 10, with the growth curves for the three variables showing some correspondence with known stages of neural growth in the brain. Inferences could not be made regarding whether children utilize the same cognitive strat­ egies as adults, or whether the WCST is sensitive to frontal lobe lesions in children. However a set of hypotheses were offered relevant to future research regarding the develop­ ment of frontal lobe functioning. Since performances of six year old children were similar to those of adults with focal frontal lesions, Chelune and Baer (1986) suggested that the 33

frontal regions possibly are not yet functionally mature at

this age. Further, the performance of seven year old chil­

dren was better than the performance of adults with focal

frontal lesions, but not those with focal nonfrontal le­ sions, suggesting that the frontal regions are maturing but

are still not yet functionally mature. Performances of ten year old children were not different than that of normal adults, perhaps indicating that the functional development of the prefrontal regions has matured, at least to the extent they are tapped by this task.

Gorenstein, Mamato, and Sandy C19B9) cite empirical studies which have demonstrated that individuals with a variety of disorders demonstrate significantly more per­ severative errors than normals: patients with prefrontal lesions, antisocial psychiatric patients, alcoholics, and children with a physician’s diagnosis of attention deficit disorder. Gorenstein et al. utilized three tasks incorpo­ rating disruption and baseline features which theoretically are linked with prefrontal dysfunction (The Trailmaking

Test, the Stroop Color-Word Test, and the Sequential Memory

Task) as well as three tasks with an empirical history of differentiating individuals with prefrontal damage (the

WCST, a Sequential Matching Memory Task, and Spontaneous

Reversal of the Meeker Cube) . Results were generally com­ patible with a prefrontal-deficit theory of inattention- overactivity, the inattentive-overactive children performed 3*1

in the direction of prefrontai-type deficit relative to

normal controls. The authors maintain that inferences re­

garding a possible organic defect in inattentive-overactive

children can not be made until this issue is addressed based

on neurophysiological findings or at least knowledge of a

particular form of brain trauma that is likely to have oc­ curred. Their results are stated to only establish a func­

tional resemblance between inattentive-overactive children and individuals with prefrontal damage.

Wiss (1982) compared performance of normally achieving, learning disabled, and mentally retarded children on the modified version of the WCST CMCST). The modification, established by Nelson, essentially consists of eliminating all ambiguous cards. Wiss found a significant difference between normal achieving and mentally retarded subjects but no differences between normals and learning disabled sub­ jects. The differences were in categories achieved, not in perseverative errors nor in failure to maintain set. It is suggested that one reason for finding no significant differ­ ences in perseverative errors may be a lack of sensitivity of the MCST, due to the omission of ambiguous cards, as compared to the WCST. 35

TRAIL MAKING TEST

Trail Making Test (TMT) part A and Part B are subtests

□f the Halstead-Reitan Neuropsychological Test Battery for

Older Children. Part A requires the child to connect with a

pencil line the numbers from 1 to 15. Part B in addition

requires that the child alternate connecting letters and

numbers (Rourke et al., 19633. Performance is timed and

scores are recorded as number of seconds required to com­

plete performance including corrections as a result of

examiner cues. Selz and Reitan (19793 published raw score ranges for children 9-14 years old for four classifications: normal, slightly below normal, below normal but not defini­ tive for brain damage, and definite impairment. Part A and

Part B are generally considered under separate neuropsycho­

logical variables: Part A, visual-spatial and visual-se­ quential ability; Part B, concept formation, reasoning, organization ability and flexibility in applying principles

(Davison, 1974; Rourke & Finlayson, 13753.

A number of investigations support the ability of the

TMT to discriminate between children with and without brain lesions (Davids, Goldenberg, & Lufer, 1957; Davison, 1974;

Reitan, 19713. The TMT has also been found to discriminate between learning-disabled and normal children (Davis, Adams,

Gates, and Cheramie, 1989; Mittelmeier, Rossi, and Berman,

19893. In addition, pattern variation among learning dis­ abled children, on Part A versus Part B, have been 36

correlated with those found in lateralized lesions CRourke &

Finlayson, 1975).

Although not well documented in the literature, TMT is

used to provide information regarding "prefrontal-type"

function (Gorenstein et al., 1989), or "executive function."

This information is derived from the child’s susceptibility

to disruption of cognitive processes by competing responses, or, ability to establish and maintain shifting of set be­ tween number and letter on Part B .

AAMD ADAPTIVE BEHAVIOR SCALE - SCHOOL EDITION

The AAMD Adaptive Behavior Scale - School Edition CABS-

SE) is based on the AAMD Adaptive Behavior Scale, 1974 edition (Lambert, Ulindmiller, Tharinger, & Cole, 1901).

Part One of the scale is designed to measure coping skills and social habits in nine behavioral domains considered to be important to the development of personal independence in daily living: Independent Functioning, Physical Develop­ ment, Economic Activity, Language Development, Numbers and

Time, Prevocational Activity, Self-Direction, Responsibili­ ty, and Socialization. Part Two measures maladaptive behav­ ior in twelve domains related to personality and behavior disorders: Aggressiveness, Antisocial versus Social Behav­ ior, Rebelliousness, Trustworthiness, Withdrawal versus

Involvement, Mannerisms, Interpersonal Manners, Acceptabili­ ty of Vocal Habits, Acceptability of Habits, Activity level. 37

Symptomatic Behavior, and Use of Medications.

The scale is administered by first person assessment or

by third-party assessment. In first-person assessment a

person who is familiar with the child and who has had train­

ing or experience in Judging the appropriateness of the

scale items completes the scale. In third-party assessment

a trained person interviews the person or persons providing

the information. Norms are available for Regular, EMR, and

TMR classroom groups, age three through sixteen with the

exception of age three through six for the EMR groups and age 16 for the Regular group (Lambert, et al., 1981).

Adaptive behavior is defined as "the effectiveness or degree with which the individual meets the standards of personal independence and social responsibility expected for his or her age and cultural group" (Grossman, 1983, p. 157).

9ince 1959 the AAMD definition of mental retardation has included deficits in adaptive behavior in addition to sub­ average intellectual functioning. Although the constructs of intelligence and adaptive behavior overlap they are not identical, especially in the upper ranges of intellectual function included in mental retardation. Intelligence implies an abstract potential while adaptive behavior empha­ sizes the individuals ability to cope with everyday situa­ tions (Grossman, 1983).

An interest in the correlation of neuropsychological measures and adaptive behavior is not surprising given the 38

roots of the term. These roots go back to the work of Pav­

lov, Anokhin and others, relevant to reflexology, classical

conditioning, and adaptive responses CLeland, 1983). In

this regard, adaptive behavior has been found to correlate

closely with autonomic function in developmentally delayed

preschool children CBerntson, Ronca, Tuber, Boysen, & Le-

land, 1385; Tuber, Ronca, Berntson, Boysen, & Leland, 1985;

Barclay & Leland, 1983). Specifically, an inverse corre­

lation was found between the magnitude of cardiac decel­

eration to simple nonsignal stimuli and adaptive behavior

measured by the Adaptive Behavior Scale for Infants and

Early Childhood CABSI). In another investigation suggesting

a relationship between neuropsychological function and adap­

tive behavior, Trites (1986) examined the relationship

between intellectual quotient, Halstead Reitan measures of

neuropsychological functioning (impairment index), and employment history of mildly retarded young adults. He

found little relationship between intellectual quotient and work history, but a strong relationship between impairment

index and work history. Individuals with similar intellec­ tual quotients differed in both the level of their impair­ ment index and in success in the work setting and community.

Individuals with impairment indexes in the mild range were more successful than individuals with impairment indexes in the moderate to severe range.

Thus, adaptive behavior measures are valuable, in both 39 research and clinical contexts. Important additional infor­ mation is provided regarding how a group of subjects Cor in the clinical context, an individual) is meeting the adaptive challenges of the environment, in the presence of varying levels of intellectual and neuropsychological functioning. CHAPTER III

METHODOLOGY

PURPOSE

The purpose of this study was to determine the nature

of neuropsychological function in a sample of school age

children, most of whom had psychometrically measured intel­

lectual function in the range of mild mental retardation.

It was hypothesized that statistical analysis would reveal subgroups of neuropsychological function and that these

subgroups may or may not correlate with categories based on clinical interpretation. It was hypothesized that the statistically identified subgroups and the clinically cate­ gorized subgroups would differ systematically on a measure of adaptive behavior.

SUBJECTS

A group of 52 subjects, age 9 through 13, were recruit­ ed. Forty-two of the subjects were recruited from school systems and were documented to have psychometrically mea­ sured intellectual function in the mild range of mental retardation (50 to 70, inclusive). Ten subjects were re­ cruited through a meningomyelocele clinic, 1 with known

40 41

intellectual quotients in the 50 to 70 range, and 6 without

documented intellectual quotients. School systems included

large and small urban and suburban public. Mean age was

11.5 years; mean intellectual quotient, BE.

Myelomeningocele subjects included 5 with thoraco­

lumbar level lesions, 5 with lumbar-sacral lesions. All

subjects had undergone shunt procedures. One subject had

been shunted only one time, 0, two times, and 1 four times.

PROCEDURE

The following assessments were administered: The Luria

Nebraska Neuropsychological Battery-Children's Revision

CLNNB-C); Trailmaking A and B (Halstead Reitan) CTMT); The

Wisconsin Card Sorting Test CWCST); the AAMD Adaptive Behav­

ior Scale, School Edition CABS-SE). The Adaptive Behavior

Scales were completed by teachers or by parents.

Assessments were administered by examiners who had achieved an interrater reliability of 90% relevant to item scoring on the LNNB-C and to summary scores of the WCST and

TMT A and B. Reliability was established initially and at intervals of at least every 20 subjects. Videotapes randomly obtained during actual data collection were uti­ lized .

Clinical analysis was based on item analysis and was designed to categorize individual protocols as reflective of impairment in sensorimotor and neurocognitive abilities 42 believed, according to Luria’& model CLuria, 1966), to be subserved by primary and secondary, or by tertiary areas of the parietal, temporal, occipital, and frontal systems.

Agglomerative hierarchical cluster analysis was uti­ lized to determine the existence of statistically identifi­ able subgroups. I score distance from critical level was entered for scale scores on Cl through Cll and SI through

53. Squared Euclidean coefficient for within group averages was utilized as a measure of dissimilarity. Number of clusters was determined by identifying the point in the clustering procedure at which the squared Euclidean distance coefficients became relatively large, indicating that clus­ ters containing quite dissimilar members were being com­ bined. Chi square, and one-way analysis of variance CANOVA) plus Scheffe test were utilized to determine the signifi­ cance of relationships of other variables to cluster member­ ship . CHAPTER IV

RESULTS AND CONCLUSIONS

This study was designed to determine the validity of the Luria Nebraska Neuropsychological Battery Children’s-

Revision for identifying subtypes of neuropsychological function in school age children, the largest group of whom had psychometrically measured intellectual function in the range of mild mental retardation. In addition, the rela­ tionship between statistical and clinical analyses, the relationships between neuropsychological functioning and adaptive behavior, as well as medical diagnoses, were exam­ ined .

OVERVIEW

Cluster analysis, utilizing clinical and summary scale

T score distances from critical level, yielded 5 clusters.

All LNNB-C clinical and summary scales contributed to sig­ nificant differences between some cluster pairs, but only a few contributed to differences between others. Some agree­ ment between statistical and clinical categorization was demonstrated when clinical categorization was based on degree of sensorimotor inefficiency. Sample performance on

43 44

the TMT and WCST was generally poor, with significant dif­

ferences between cluster pairs only on TMT. Myelomenin­

gocele subject membership was distributed across clusters

and clinical categories.

One-way ANOVA utilizing adaptive behavior composite

scores, followed by the Scheffe test, documented no signifi­

cant statistical differences between clusters, clinical categories, nor diagnostic groups. A post hoc analysis of

the clinical significance of adaptive behavior factor scale scores was performed. Comparisons across clusters revealed

an increased incidence of deficient scores on factor 1,

Personal Self-Sufficiency, for cluster 5, the cluster re­ flecting minimal neuropsychological inefficiency, and con­ taining three myelomeningocele subjects; in addition, clus­ ter 1, the cluster reflecting mild neuropsychological inef­ ficiency, demonstrated more deficient scores than other clusters.

Comparisons of adaptive behavior factor scale scores across clinical categories revealed an increased frequency of deficient scores for subjects with unilateral sensorimo­ tor inefficiency. Comparison of developmentally handicapped subjects with myelomeningocele subjects demonstrated an in­ creased frequency of deficient scores for those in the developmentally handicapped group, with the exception of factor 1, where all myelomeningocele subjects demonstrated deficient scores. 45

In general, clinical analysis, utilizing item analysis

and based on Luria’s neuropsychological model, indicated

that the subjects in the sample as a whole had difficulty

with tasks believed to be subserved, according to this model, by the secondary areas of the temporal and frontal

lobes, and the posterior tertiary area. Variations from this included a few subjects with only relatively mild difficulty with the more elementary tasks, and many subjects with inefficiency suggested in additional areas.

NATURE OF NEUROPSYCHOLOGICAL FUNCTION

The first research question posed was whether the experimental population would differ in their clinical neuropsychological manifestations and statistically cluster into subgroups. Cluster analysis yielded five clusters.

Clusters 1 through 3 are composed of 21, 13, and 14 sub­ jects, respectively. Cluster 4 consists of only 1 subject, with spastic diplegia, the only subject in the sample with this medical diagnosis. Cluster 5 consists of 3 subjects from the myelomeningocele subsample with unknown IQ scores, and with relatively low profiles, suggesting that their IQ's may fall outside the range of mild mental retardation. On a continuum of relative severity of apparent neuropsychologi­ cal inefficiency, as measured by T score distance from critical level for clinical and summary scales of the LNNB-

C, the clusters are ordered, from least to most severe, as 4B

follows: cluster 5, minimal inefficiency; cluster 1, mild

inefficiency; cluster 3, moderate inefficiency, cluster 2

severe inefficiency. Because cluster 4 contained only one

member, a developmentally handicapped subject with spastic

diplegia, this cluster has been excluded from further analy­

sis. Means, standard deviations, and ranges for T scale

distance from critical level are shown in Table 1.

The averaged profiles for each cluster, and for the

sample as a whole, are shown in Figure 1, with critical

levels indicated along the T scale. In general, the pro­

files of clusters containing predominantly developmentally

handicapped subjects, that is 1, 2, and 3, were similar in

that generally the speech, academic, and intellectual func­

tion scales were relatively highly elevated, although at

significantly different levels for each cluster. These

clusters differed in their overall level of elevation, and

the relative elevation of the more elementary scales (motor,

rhythm, tactile, and visual) and the memory scale (see

Appendix A for a listing of clinical and summary scales by

name and number). The averaged profile for the mildly

inefficient cluster (1) reflects motor, tactile, visual and

sensorimotor scales below the critical level, with rhythm

and memory slightly elevated. Moderate elevations are re­

flected for the academic scales, with the exception of arithmetic which is more highly elevated. Higher elevations are reflected for the speech scales. Although the 17

Table 1

>^W/T w^w» .w < _ ,»^WW «. w*. l^»W

Clusters

LNNB--C Scales Mild Severe Moderate Minimal 1 E 3 5 Cl: mean -Ü.E 1.3 0.3 -0.1 s t d . d e v . 1.1 0.8 0.8 0.7 range 1.3 E.5 E.8 1.3 CE: mean 0.1 E.5 1.7 —1. B std. d e v . 1.0 1.1 1.3 0.1 range 3.8 1.0 l.E 3.E

C3: mean -0.8 3.3 0.7 - 0 . 8 s td. d e v . l.E 1.1 0.3 l.B range 1.0 3.8 3.5 3.3 Cl: mean 0.3 E.l 1.1 -1.0 s t d . d e v . l.E 1.1 1.5 O.E range 1.1 B.l 5.3 0.1 C5: mean 1.3 5.1 3.7 1.0 std. dev. 1.7 O.B l.B 0.3 range 8.3 1.7 1.1 E.3 CB: mean 1.7 1.3 3.B -1.5 st d . d e v . 1.7 1.3 1.1 l.E range B . B 1.1 1.5 E.l C7: mean 1.5 1.3 3.8 -E.O st d . d e v . 0.3 1.5 1.3 -E.O range E.3 5.7 1.0 0.1 CB: mean 1.5 3.0 E.9 -1.7 std. d e v . 1.0 1.1 0.7 0.8 range 3.7 1.0 S.B 1.5 C9: mean 1.3 3.3 3.3 -1.5 s t d . d e v . 0.3 0.8 0.3 0.5 range 3.E 3.E 3.3 0.3 CIO: mean 0.1 E.O 1.7 -0.3 std. dev. 0.3 1.1 1.1 0.3 range 3.E 3.E 3.3 1 .B Cll: mean 1 .B 3.8 E.3 1.1 std. dev. 0.8 l.E 0.3 1 .E range 3.5 3.B 3.1 E.E SI: mean E.3 3.8 3.B -0.1 s td. d e v . 1.1 0.8 0.7 1.1 range 1.B E.3 E.BE.B SE: mean -0.1 1.3 1 .E -0.5 std. dev. 1.3 0.8 1.1 E.5 range 1.8 E.3 1.0 1.7 S3: mean -0.3 1.1 1.3 -1.3 s t d . de v . 1.1 1.1 E.OO.B range 3.3 0.1 3.3 3.0 48

Clinical and Summary Scales

01 02 03 04 05 06 07 08 09010 Oil SI 82 S3

120 120 115 h /\♦ 115 / 110 \ \ 110 ./ \ / o \ 105 f \ / \ \ 105 1 100 Pc-V- 100 0 / /- L 95 / \ A i f A 95 /' 90 /X \ ' v 7 \ % i ' f \\ 90 / ■1 A\ 85 I u 85 . □ X 80 A / \ 1— 80 1 w / D - ■ X 75 1 4 p I 75 X\ \ \ . • . \ 70 \ A'" : m 70 □ ; . \ \ 65 65 • % T P / \ • \ 60 □' *• --P ' A 60 ■ ' , P 55 \ p. P 55 50 '-tk 50 ' . \ 45 ■ 45 P 40 40 o cluster 1 O cluster 2 ■A cluster 3 35 35

30 30 ♦ Cluster 4 ■ cluster 5 X averaged profile (clusters 1,2, 3) 01 02 03 &

Figure 1. Cluster profiles, based on average LNNB-C T score distance from critical level (see Appendix A for listing of clinical and summary scales by name and number). 49

pathognomonic scale is generally highly elevated for the

sample as a whole, a nontraditional interpretation may be

indicated. Traditionally, for subjects in the normal range

of intellectual function, elevations on this scale are

indicative of brain injury, and of compensation for brain

injury. However, items on the scale predominantly consist

of complex educational tasks, therefore, the experimental

population elevations may be as much related to poor educa­

tional progress than as the presence of and compensation for

brain injury.

The averaged profile for the moderately inefficient

cluster C33 is similar to that of the mildly inefficient

cluster (1) in configuration, but higher. The moderately

inefficient profile (cluster 3) reflects slight elevations

on the motor, tactile, visual, and sensorimotor scales, with

moderate elevations on rhythm and memory, and high eleva­

tions for the speech, academic, and intellectual scales.

The averaged profile for the severely inefficient cluster

(S3 reflects similar relative elevations, but differs in

that the sensorimotor scales are highly elevated. The

profile for the minimally inefficient cluster (53, the clus­ ter containing 3 myelomeningocele subjects with suspected

IQ’s above 70, reflects elevations only on receptive speech and intellectual functions, with the pathognomonic and memory scales only slightly elevated.

Although item analysis may be the preferred method of 50 clinical analysis for subjects with psychometrically measured intellectual function in the range of mild mental retardation, for the purpose of matching profiles of future subjects with the best fitting cluster, the clusters may best be characterized by plus and minus one standard devia­ tion from the mean score above critical level for each clinical and summary scale, that is, a range which would

include 68.26% of the cluster. Table 2 shows the mean I score distance from critical level and the range for plus and minus one standard deviation, for each scale, for clus­ ters 1 through 3, and 5. Where these I scale ranges for clusters overlap, two methods of determining best cluster fit for a given profile can be utilized. First, determine which mean the subjects score most closely approximates.

Second, look for scale score ranges which do not overlap with other clusters.

One-way analysis of variance CANOVA) was utilized to determine which LNNB-C scales contributed to the variance between clusters. A significant effect was indicated for all scales CF C3,*17] ranging from ‘±1.8104 to 8.9517, p-

<.001). The Scheffe test was utilized to identify pairs of clusters which were significantly different at the .05 level. Variables which contributed significantly to differ­ ences between pairs of clusters are shown in Table 3.

All scales contributed significantly to differences between cluster pairs 1-2 Cmild and severe inefficiency. 51

Table 2

w ^ — — — — « »ww ,w , _- _ Minus,,^,,ww One*^..w Standard Deviation

Cluster

LNNB--C Scale Mild Severe Moderate Minimal 1 2 3 5 Cl: mean -0.2 1.3 0.3 -0.1 + 1 SO O.B 2.3 1.3 O.B - 1 SO -1.2 0.3 -0.1 -1.1 C2 mean O.'i 2.5 1.7 -l.B + 1 SD 1.1 3.5 2.7 -0 .B - 1 SO -O.B 1.5 0.7 -2.B C3 mean —0. B 3.3 0.7 -O.B + 1 SD 0.2 1.3 1.7 0.2 - 1 SD -l.B 2.3 -0.3 -l.B Cl mean 0.3 2.1 1.1 -1.0 + 1 SD 1.3 3.1 2.1 0.0 - 1 SD -0.7 1.1 0.1 -2.0 C5 mean 1.3 5.1 3.7 1.0 + 1 SD 2.3 B.l 1.7 2.0 - 1 SD 0.3 1.1 2.7 0.0 C6 mean 1.7 1.3 3.B -1.5 + 1 SD 2.7 5.3 l.B -0.5 - 1 SD 0.7 3.3 2.B -2.5 C7 mean 1 .5 1.3 3.3 -2.0 + 1 SD 2.5 5.3 l.B -1.0 - 1 SD -0.5 3.3 2.B -3.0 CB mean 1.5 3.0 2.3 -1.7 + 1 SD 2.5 1.0 3.3 -0.7 - 1 SD 0.5 2.0 1.3 -2.7 C9 mean 1.3 3.3 3.3 -1.5 + 1 SD 2.3 1.3 1.3 -0.5 - 1 SD 0.3 2.3 2.3 —2 .5 CIO mean 0.1 2.0 1.7 -0.3 + 1 SD 1.1 3.0 2.7 0.7 - 1 SD —0.6 1 .0 0.7 -1.3 Cll mean l.B 3.B 2.3 1 .1 + 1 SD 2.B l.B 3.3 2.1 - 1 SD O.B 2.B 1.3 0.1 SI mean 2.3 3.B 3.B -0.1 + 1 SD 3.3 l.B l.B 0.9 - 1 SD 1.3 2.B 2.B -1.1 SB mean -0.1 1.3 1.2 —0.5 + 1 SD O.B 5.3 2.2 0.5 - 1 SD -1.1 3.3 0.2 -1.5 S3 mean 1.1 1.1 2.0 O.B + 1 SD 2.1 2.1 3.0 l.B - 1 SD 0.1 0.1 1.0 -0.1 52

TABLE 3

LNNB-C Scales Contributina to Sianifleant Differences Between Cluster Pairs

Cluster Pairs

Scale 1-2 5-2 1-3 5-3 5-1 2-3

Cl + + + - — —

C2 + + + + — —

C3 + + + - - +

Cl + + -- — —

C5 + + + + — —

C6 + + + + + -

C7 + + + + + -

CB + + + + + -

C9 + + + + + -

CIO + + + + — —

Cll + + + - — —

SI + + + + -

S2 + + + - - +

S3 + + + + — +

Note: + = significant difference - = no significant difference

cluster 1 = mild inefficiency cluster 2 = severe inefficiency cluster 3 = moderate inefficiency cluster 5 = minimal inefficiency

see Appendix A for name and number of LNNB-C scales 53 respectivelyD and 5-2 (minimal and severe inefficiency), and

1-3 Cmild and moderate inefficiency), with the exception of

Visual Functions for pair 1-3. Most scales contributed to differences between cluster 5 and 3, (minimal and moderate

inefficiency) with Motor, Tactile, Visual Functions, Intel­

lectual Processes, and Left Sensorimotor not contributing.

Clusters 5 and 1 (minimal and mild inefficiency) were sig­ nificantly different only on the academic scales. Expressive

Speech, and 51, Pathognomonic. Clusters 2 and 3 (severe and moderate inefficiency) differed significantly only on Tac­ tile, and Left and Right Sensorimotor Scales.

The direction of these differences can be determined by examining Figure 1 which shows the averaged profiles for all subjects, and for each cluster. For cluster pairs with all or most scales contributing to differences (1-2, Cmild and severe inefficiency], 5-2 [minimal and severe inefficiency], and 1-3 Cmild and moderate inefficiency]), the profile for the mildly inefficient cluster (1) is lower than that for the moderately inefficient cluster (3), and even lower than for the severely inefficient cluster (2). Thus, consistent with the labeling of the clusters as minimally through severely inefficient in apparent neuropsychological func­ tion, this suggests greater inefficiency in neuropsycholog­ ical function for clusters 2 and 3 in comparison to cluster

1. Similarly, the profile for cluster 5 is much lower than for cluster 2, suggesting more inefficiency of 54

neuropsychological function for cluster 2.

For cluster pair 5 and 3 (minimal and moderate ineffi­

ciency, respectively), the pair with almost all variables

contributing to difference, the direction of significant

differences was that of less efficient function for cluster

3 on the speech and academic scales, and for Rhythm, Memory,

Pathognomonic, and Right Sensorimotor. No statistically

significant differences were indicated for the Motor, Tac­ tile, Visual Functions, Intellectual Processes, and Right

Sensorimotor Scales. Of interest is the high elevation on

Intellectual Processes for cluster 5, and the significantly higher Left Sensorimotor Scale score, suggesting relative

inefficiency within cluster 5 on functions tapped by these scales.

Regarding direction of significant differences between cluster pairs with only a limited number of scales contrib­ uting, for cluster pair 5 and 1 (minimal and mild ineffi­ ciency) the direction of differences was that of less effi­ cient function for cluster 1 on the academic scales, and on the Expressive Speech and Pathognomonic Scales. Of inter­ est, in terms of similarity reflecting relatively ineffi­ cient function for cluster 5, is the lack of significant difference on Receptive Speech and Intellectual Processes.

Cluster pair 2 and 3 (severe and moderate inefficiency) differed significantly only on Tactile and Sensorimotor

Scales, in the direction of less efficient function for 55

cluster 2.

Although cluster analysis indicated significant statis­

tical within group homogeneity and between group heterogene­

ity, there may be clinically significant differences within

the clusters. The range of I scale differences for the

LNNB-C scales, shown in Table 1 and falling between 0.1 and

6.6 T, support this observation. In addition, the identifi­

cation of 27 clinical categories also lends support to the observation of clinical heterogeneity within clusters. This

highlights the importance of using qualitative analysis and

item analysis when clinically interpreting profiles of children with psychometrically measured intellectual func­ tion in the 50 to 70 IQ range.

One-way ANOVA and the Scheffe test were utilized to determine whether test age and IQ contributed significantly to the variance between clusters. A significant effect was indicated for test age CF C3, 47] - 2.41, p<.005). However, no significant difference was demonstrated between clusters at the .05 level. Means for clusters 1 through 3 and 5 were

139 Cmild inefficiency], 125 Csevere inefficiency), 133

Cmoderate inefficiency), and 146 Cminimal inefficiency) months, respectively. A significant effect was indicated for IQ CF [2, 38] = 6.74, p < .005). A significant difference was demonstrated between clusters 1 and 2 Cmild and severe inefficiency), at the .05 level. Means for dusters 1 through 3 were 65 Cmild inefficiency), 57 Csevere), and 62 56

(moderate), respectively. The lower IQ For cluster 2 is

consistent with their higher LNNB-C profile.

The supplemental tests, Trail Making Test CTMT) and the

Wisconsin Card Sorting Test (WCST) were administered as

hypothetical measures of prefrontal-type function. Because

of the controversial issues regarding inferences about

prefrontal-type functions in children, and in order to

reduce the number of variables in the cluster analysis, they were not included in the cluster analysis, and will be considered only descriptively. Means for the entire sample of 52 subjects were 2.135 and 2.365 for parts A and B, respectively. A frequency table for TMT scores for each cluster is shown in Tables 4 and 5. Scores of 0 through 3 indicate, no impairment, mild, moderate, and severe impair­ ment, respectively. Because of the number of cell frequen­ cies below 5, data was collapsed in order to obtain a valid statistical analysis using chi square. Categories 0 and 1 were collapsed into "1", and clusters 2 (severe inefficien­ cy) and 3 (moderate inefficiency) were collapsed into "2", clusters 4 (one subject with spastic diplegia) and 5 (mini­ mally inefficiency) were excluded. Chi square indicated a significant difference in the distributions across catego­ ries for Part A (chi square C2, N“4B3 = 7.47, p < .05) and

Part B (chi square C2, N-4B] = 12.65, p<.005). Cluster 1

(mild inefficiency) demonstrated significantly less impair­ ment on both parts A and B than clusters 2 and 3 combined 57

Table H.

Cluster Frequencies for TUT Part A

TMT Category

Cluster 0 1 2 3

Mild Cl) 0 11 S S

Severe CE) 0 2 3 8

Moderate C3) 0 3 3 8

Sp. Diplegia C4) 0 0 0 1

Minimal C5) 0 1 0 2

Table 5

Cluster Freauencies for TMT Part B

TMT Category

Cluster 0 1 2 3

Mild Cl) 3 SS 8

Severe CS) 0 0 0 13

Moderate C3) 0 1 3 10

Sp. Diplegia C4) 0 0 0 1

Minimal CS) 1 0 1 1 58

Csevere and moderate inefficiency). This is consistent with the ranking of the LNNB-C profiles for clusters, with the cluster 1 profile lower than that for clusters 2 and 3, reflecting less neuropsychological inefficiency for cluster

1 .

One-way ANOVA and the Scheffe test, utilizing standard deviation scores for Categories Achieved, Perssverative

Errors, and Failure to Maintain Set, on the WCST, indicated no significant differences between clusters at the .05 level. Sample means were -3.8 (range +0.8 to -7.5), +3.2

(range -0.2 to +8.3), and -0.1 (range -1.1 to +3.3) for

Categories Achieved, Perseverative Errors, and Failure to

Maintain Set, respectively. Means for Categories Achieved and Perseverative Errors indicate poor performance. Sample mean number of Categories Achieved was 1.8, with a range from 0 to 6.

CLINICAL AND STATISTICAL AGREEMENT

The second research question addresses the extent of agreement between statistical clustering and clinical cate­ gorization. Clinical categorization of profiles based on item analysis yielded 11 categories. These 11 categories were based on difficulty with tasks believed to be sub­ served, according to Luria’s model, by the sensorimotor, secondary, and tertiary areas. Further refinement, based on severity of difficulty (mild, moderate, or severe) yielded 53

27 categories. Because of sample size, and the limited

number of clusters, statistical agreement could not be

determined. The frequency of cluster members falling in

each of the 11 categories is shown in Table 6. It can be

seen that for each cluster, at least 50% of the members did

fall in two of the categories, suggesting some agreement.

In order to determine the extent of statistical agree­

ment, the 11 categories were collapsed into three, based on

sensorimotor involvement: none CNSM), unilateral CUSH), or

bilateral CBSM). Clusters 4 Cone subject with spastic

diplegia) and 5 Cminimal inefficiency, 3 subjects) were

excluded because of small sample size which would result in

too many cells with frequencies less than 5. Chi square

indicated a significant difference in cluster distribution

across clinical categories Cchi square C4, N = 48] = 31.15,

p<.0001; phi coefficient = .01).

Cluster 1 Cmild neuropsychological inefficiency) is

characterized by no sensorimotor inefficiency, clusters 2

and 3 Csevere and moderate neuropsychological inefficiency,

respectively) by bilateral sensorimotor inefficiency. As

shown in Table 6, the cluster 4 member, with spastic diple­

gia, was categorized as having bilateral sensorimotor im­ pairment, and inefficiency in 4 secondary and 1 tertiary

area Cposterior). Two of the three members of cluster 5

Cminimal neuropsychological inefficiency) were categorized as demonstrating no sensorimotor inefficiency, with 60

inefficiency demonstrated in 2 secondary and 1 tertiary

Cposterior) area.

Highest frequency of clinical category was bimodal.

The most frequent clinical categories were bilateral senso­ rimotor, 4 secondary, 1 tertiary CB41) and no sensorimotor,

2 secondary, 1 tertiary CN21), with 13 and 12 subjects each, respectively. Eight of the 13 subjects in B11 are from cluster 2, ten of the subjects in clinical category N21 are from cluster 1.

The typical profile in the B41 clinical category re­ flected difficulty with tasks believed to be subserved by the sensorimotor strips (bilaterally), the secondary areas of the frontal, parietal, temporal, and occipital systems, and the posterior tertiary area. Performance on the LNNB-C, including clinical observations, did not suggest significant tertiary anterior inefficiency, according to clinical analy­ sis based on Luria’s model. Although performance on WCST and TMT was poor across the sample in general, as well for subjects in this clinical category, inferences were not made regarding prefrental function. This clinical decision was based on the controversial nature of the issue of prefrental function in children, emerging from the empirical litera­ ture. Areas of difficulty characterizing the B41 subject are shown in Table 7.

The clinical category with the second highest frequen­ cy, N21, reflected no difficulty with sensorimotor tasks, B1

Table 6

Clinical Categoru Freauencu

Cluster

Clinical Category 1 2 3 4 5 Total

1. NSM 4S IT 0 0 1 0 0 1

2. NSM 3S IT 4 0 0 0 0 4

3. NSM 2S IT 10 0 0 0 2 12

4. RSM 3S IT 0 0 2 0 0 2

5. RSM 2S IT 2 0 0 0 0 2

6. LSM 4S IT 0 0 2 0 0 2

7. LSM 3S IT 1 0 1 0 1 3

B. BSM 4S 2T 0 3 1 0 0 4

9. BSM IS IT 1 8 3 1 0 13

10. BSM 3S IT 3 1 4 0 0 8

11. BSM 2S IT 0 1 0 0 0 1

Total 21 13 14 1 3 52

Note: NSM = no sensorimotor inefficiency RSn = right sensorimotor inefficiency LSM “ left sensorimotor inefficiency BSM = bilateral sensorimotor inefficiency S = secondary area inefficiency I = tertiary area inefficiency

cluster 1 = mild neuropsychological inefficiency cluster 2 = severe neuropsychological inefficiency cluster 3 = moderate neuropsychological inefficiency cluster 1 = one subject with spastic diplegia cluster 5 = minimal neuropsychological inefficiency G2

Table 7

Tuoical Clinical Protocol B41

Difficulty Demonstrated

Subserved By

Sensorimotor tactile and motor items

Secondary graphesthesia Parietal stereognosis

Secondary visual perception pictures Occipital recognition of words, letters

Secondary perception of pitch relationships Temporal reproduction of pitch relationships perception and evaluation of acoustic signals phonemic hearing, analysis, synthesis word comprehension, effect of repetition understanding simple sentences repetition of sentences; verbal memory

Secondary simple sequential movements Frontal dynamic organization of hands complex organization of movements and actions reproduction of pitch relationships reproduction of musical melodies motor performance of rhythmic groups sequencing - phonemic, words fluency and automatization of speech predicative speech, reproductive speech spontaneous productive speech

Tertiary optic spatial organization of hands Posterior visual spatial orientation selection of grapheme spatial symbolic functions logical grammatical structure categorical structure of numbers internal operations intermediate operations synthesis, phonetic, visual concrete optic spatial

Note; B11 = inefficiency in bilateral sensorimotor, four secondary, and one tertiary areas 63 but did demonstrate difficulty with tasks believed to be subserved by the secondary areas of the temporal and frontal lobes, and the posterior tertiary area. These protocols, in contrast to B41, are characterized by less inefficiency of neuropsychoiogical functions. The areas of difficuity typically reflected on the N21 protocol are shown in Table

8.

Although many protocols varied from the two most fre­ quent, B41 and N21, ail subjects demonstrated difficulty with tasks believed to be subserved by the secondary tempo­ ral and frontal areas. Variations in the direction of iess inefficiency included mild or no difficulty with the more eiementary tasks subserved by the secondary temporal and frontal areas, and in a few cases, by milder difficulty with logical grammatical structure. Variations in the direction of more inefficiency included 4 subjects who qualitatively demonstrated difficulties suggestive of prefrontal-type inefficiency. Of these 4, one demonstrated difficulty with tasks subserved by 3 rather than 4 secondary areas, with only mild difficulties demonstrated on tasks subserved by the secondary temporal and frontal systems. Difficulties demonstrated which are considered to be suggestive of pre­ frontal-type inefficiency inciuded severe perseveratory tendencies, with perseveration across tasks as well as within tasks, and consistent difficulties with the reguiato- ry function of speech, beyond a level expected based on 64

Table 8

Tupical Clinical Protocol NSI

Difficulty Demonstrated

Subserved By

Secondary perception of pitch relationships Temporal evaluation of pitch relationships reproduction of pitch relationships reproduction of musical melodies phonemic hearing, analysis, synthesis word comprehension, effects of repetition repetition of sentences narrative speech verbal memory

Secondary simple sequential movement of hands Frontal dynamic organization of hand movements complex organization of movements reproduction of pitch relationships phonemic synthesis; speech, writing, reading pathological inertia of previously established stereotypes, speech narrative speech

Tertiary optic spatial organization of hand movements Posterior naming, qualitative observation spatial symbolic logical grammatical structure categorical structure intermediate operations visual synthesis abstract categorization

Note; NEl = inefficiency in no sensorimotor, two secondary, and one tertiary areas 65 language ability.

MYELOMENINGOCELE STATISTICAL AND CLINICAL MEMBERSHIP

The third question posed was whether the myelomeningo­ cele subjects would cluster together on statistical analysis and whether they would belong to the same clinically estab­ lished diagnostic group. Subjects with myelomeningocele were not homogeneous for cluster nor clinical category membership. Two each belonged to clusters 1 and 2 Cmild and severe neuropsychological inefficiency, respectively), three each to clusters 3 and 5 Cmoderate and minimal neuropsycho­ logical inefficiency, respectively). Clinically, 3 were in category 3 Cno sensorimotor, 2 secondary, 1 tertiary) one in

7 Cleft sensorimotor, 3 secondary, 1 tertiary), 4 in 3

Cbilateral sensorimotor, 4 secondary, 1 tertiary) , and 2 in

10 Cbilateral sensorimotor, 3 secondary, 1 tertiary). This distribution across clusters and clinical categories sup­ ports a hypothesis of heterogeneity of function in this diagnostic group.

ADAPTIVE BEHAVIOR

Research questions 4, 5, and B addressed the relation­ ship between Adaptive Behavior Scale, School Edition CLam- bert et al, 1981) composite scores and cluster, clinical category, and medical diagnostic group memberships, respec­ tively. Lambert's norms for the Educable Mentally Retarded 66

(EMR) were utilized (Lambert, 1981). The Scheffe test,

following one-way ANOVA, indicated no significant

differences, at the .05 level, between clusters CF [3,39] =

2.86, p<.05), clinical categories CF C2,41D - 0.61, p.5458),

nor diagnostic groups CF [1,41] = 5.20, p.<05) on adaptive

behavior composite scores. Mean composite score, expressed

in percentile, for the entire sample was 43.6, with a mini­ mum of 3.2 and a maximum of 96.2. This finding would sup­ port a hypothesis that elements besides neuropsychological function contribute to the variation in adaptive behavior scores. These other elements are most likely environmentally based.

In order to further explore the relationship between adaptive behavior scale performance and cluster, clinical category, and medical diagnostic category memberships, a post hoc analysis was performed utilizing the factor scale scores, rather than the composite scores. The AB9-SE fac­ tors are : factor 1, Personal 3elf-Sufficiency; factor 2,

Community 9elf-5ufficiency; factor 3, Personal-9ocial Re­ sponsibility; factor 4, 5ocial Adjustment; factor 5, Person­ al Adjustment. Group means and frequency of members with scores of 7 or below were analyzed in terms of clinical significance. 67

Table 3

Mean AB5-SE Factor Scale Scores For Clusters

ABS-SE Factor

Cluster 1 2 3 4 5

Mild Cl) S.SI 10.S*i 10.5S 10.47 10.41

Severe C2) 8.55 10.36 11.82 11.18 10.55

Moderate C3) S.82 S. *45 10.00 10. SO 10.55

Minimal C5D 2.00 12.00 11.00 11.66 11.00

Note: Factor 1 = Personal Self-Sufficiency Factor 2 = Community Self-Sufficiency Factor 3 = Personal-Social Responsibility Factor 1 = Social Adjustment Factor 5 = Personal Adjustment 6 8

ADAPTIVE BEHAVIOR FACTOR SCALE SCORES FOR CLUSTERS

The mean ABS factor scale scores for clusters 1 through

5 are shown in Table 3. When interpreting the ABS-SE

factor scale scores, scores of 7 or below, that is two

standard deviations below the mean, are considered deficient

(Lambert, 1981). As a group, only the minimally inefficient

cluster (5) demonstrated a mean factor scale score in the

deficient range. This is reflected for factor 1, Personal

Self-Sufficiency, and can be attributed to the physical

disability resulting from myelomeningocele. Table 10 shows the frequency of factor scale scores of 7 or below in each cluster. The mildly inefficient cluster Cl) demonstrated the highest total number of factor scale scores of 7 or below, with 3 members each demonstrating scores of 7 or below For factors 1, 3, and 4; Personal Self-Sufficiency,

Personal-Social Responsibility, and Social Adjustment, respectively. All cluster 5 members had scores of 7 or below on factor 1. Four members of the severely inefficient cluster CS) had scores of 7 or below on factor 1.

Finding a higher frequency of deficient scores in the mildly inefficient cluster Cl) may be somewhat unexpected, because this cluster had a lower neuropsychological profile than the moderately and severely inefficient clusters C3 and

2). The increased number of deficient scores may be due to the members of the mildly inefficient cluster having enough general ability to perceive their difference, to experience 69

Table 10

Cluster Freauencu of ABS-SE Factor Scale Scores in the Deficient Ranoe

ABS-SE Factor

Cluster 1 2 3 4 5

Mild Cl) frequency 3 1 3 3 2 percentage 18 6 18 18 12

Severe C2) frequency 4 2 0 0 1 percentage 36 18 0 0 9

Moderate (3) frequency 2 2 1 1 1 percentage 18 18 9 9 9

Minimal CS) frequency 3 0 0 0 0 percentage 100 0 0 0 0

Note: Factor 1 = Personal Self-Sufficiency Factor 2 = Community Self-Sufficiency Factor 3 = Personal-Social Responsibility Factor 4 = Social Adjustment Factor S = Personal Adjustment 70

frustration in coping, and to act out, or demonstrate non-

compliance .

ADAPTIVE BEHAVIOR FACTOR SCORES FOR CLINICAL CATEGORIES

Table 11 shows the mean ABS-SE factor scale scores for the clinical categories, no sensorimotor (NSM), unilateral sensorimotor (USM), and bilateral sensorimotor CBSfl) ineffi­ ciency. No means fell within the deficient range. Table

12 shows the frequency of deficient scores for each clinical category. A higher percentage of the USM members demonstrat­ ed deficient factor scale scores. The USM distribution represents four subjects, all categorized as L31, that is, based on an item analysis, they demonstrated difficulty with tasks believed to be subserved by the left sensorimotor areas, three secondary ares, and the posterior tertiary area.

ADAPTIVE BEHAVIOR FACTOR SCORES FOR DIAGNOSTIC GROUPS

Table 13 shows the mean factor scores for diagnostic categories, that is developmental handicapped, and myelome­ ningocele subjects. The only mean factor score in the deficient range is factor 1 for the myelomeningocele group, again this can be attributed to their physical disability.

Table 14 shows the frequency of deficient factor scale scores for both diagnostic groups, reflecting deficient scores for all myelomeningocele subjects for factor 1, with 71

Table 11

Mean ABS-SE Factor Scores for Clinical Categories

ABS-SE 1Factor

Clinical Category CN) 1 2 3 4 5

NSM (14) 8.6 10.8 10.4 10.7 10.6

USM CS) 8.3 9.8 10.2 10.6 10,7

BSM CEO) 3.1 10.8 11.B 11.2 10.5

Note; Factor 1 = Personal Self-Sufficiency Factor 2 = Community Self-Sufficiency Factor 3 = Personal-Social Responsibility Factor 4 = Social Adjustment Factor 5 = Personal Adjustment

NSM = no sensorimotor inefficiency USM = unilateral sensorimotor inefficiency BSM = bilateral sensorimotor inefficiency 72

Table 12

Ranoe for Clinical Cateooru

ABS-SE Factor

Clinical Category CN) 1 2 3 4 5

NSM (14) frequency 4 1 1 2 1 percentage 23 7 7 14 7

USM CS) frequency 2 2 2 1 0 percentage 22 22 22 11 0

BSM C20) frequency 6 2 1 1 3 percentage 30 10 5 5 15

Note; Factor 1 = Personal Self-Sufficiency Factor 2 = Community Self-Sufficiency Factor 3 = Personal-Social Responsibility Factor 4 = Social Adjustment Factor 5 = Personal Adjustment

NSM “ no sensorimotor inefficiency USM = unilateral sensorimotor inefficiency BSM = bilateral sensorimotor inefficiency 73

Table 13

Heart ABS-SE Factor Scale Scores For Diagnostic Cateooru

ABS-SE Factor

Diagnostic Category (N) 1 2 3 4 5

Developmentally Handicapped (37) 10.0 10.4 10.9 10.7 10.4

Myelomeningocele CS) 2.2 11.3 11.3 12.0 11.3

Note: Factor 1 = Personal Self-Sufficiency Factor 2 = Community Self-Sufficiency Factor 3 = Personal Social Responsibility Factor 4 = Social Adjustment Factor 5 = Personal Adjustment 74

Table 14

Ranoe For Diaonostic Cateooru

ABS-SE F actor

Diagnostic Category (N) 1 2 3 4 5

Developmentally Handicapped (37) frequency 7 5 4 4 4 percentage 19 14 11 11 11

Myelomeningocele (6) frequency 6 0 0 0 0 percentage 100 0 0 0 0

Note; Factor 1 = Personal Self-Sufficiency Factor 2 = Community Self-Sufficiency Factor 3 = Personal-Social Responsibility Factor 4 = Social Adjustment Factor 5 = Personal Adjustment 75

no other factor scores deficient for this group. In con­

trast, the developmentally handicapped group demonstrated

deficient scores in all factors, with 7 (19%) members having

deficient scores on factor 1, Personal Self-Sufficiency.

SUMMARY

The experimental population did differ in their clini­ cal neuropsychological manifestations and statistically cluster into subgroups. All LNNB-C scales contributed to differences between some cluster pairs, and only a few contributed to differences between others. In addition, IQ contributed to difference between the mildly inefficient cluster Cl) and the severely inefficient cluster (5). Range of I scale distance from critical level within clusters, as well as the large number of clinical categories established by item analyses, suggest that, although there is signifi­ cant homogeneity, there may also be significant clinical differences within clusters. Performance on WEST and TMT were poor for the sample as a whole, with differences be­ tween clusters only for TMT. Those differences were consis­ tent with the level of relative neuropsychological ineffi­ ciency reflected by the clusters.

The extent of agreement between statistical clustering and initial clinical categorization could not be determined, because of the small sample size and the high number of clinical categories yielded by item analysis. However, by 7B collapsing clinical categories based on sensorimotor

Inefficiency, and by excluding clusters 4 and 5, significant agreement was documented. Clinically the sample as a whole demonstrated difficulty with tasks believed to be subserved by the secondary areas of the temporal and frontal lobes, and the posterior tertiary area. Variations from this

Included a few subjects with only relatively mild difficulty with the more elementary tasks, and many subjects with

Inefficiency suggested In additional areas.

Myelomeningocele subject membership was distributed across clusters and clinical categories. This finding Is consistent with previous findings regarding the heterogene­ ity of function In Individuals with this medical condition.

Although there was no significant statistical differ­ ences between clusters, clinical categories, nor diagnostic groups for adaptive behavior composite scores, a post hoc analysis based on the clinical significance of factor scale scores did reflect some differences. Comparisons across clusters revealed an Increased Incidence of deficient scores on factor 1, Personal Self-Sufficiency, for the minimally

Inefficient cluster (5). In addition, differences were also documented for the mildly Inefficient cluster Cl) versus the severely and moderately Inefficient clusters CS and 3), with the mildly Inefficient cluster Cl) demonstrating more defi­ cient scores. Comparison of factor scale scores across clinical categories revealed an Increased frequency of 77 deficient scores for subjects with unilateral sensorimotor

inefficiency. Comparison of developmentally handicapped subjects with myelomeningocele subjects demonstrated in­ creased frequency of deficient scores for the developmental-

ly handicapped group, except for factor 1, with all myelome­ ningocele subjects demonstrating deficient scores on this factor only. CHAPTER V

DISCUSSION

The results of this study support the validity of the

LNNB-C for identifying subtypes of neuropsychological func­

tion in school age children with psychometrically measured

intellectual function in the range of mild mental retarda­

tion. This finding has educational, clinical, and research significance. Educationally, the identification of subtypes

provides objective evidence of the heterogeneity within the mildly retarded population. Not all mildly retarded indi­

viduals are alike in their neurocognitive function, and

therefore differential educational strategies are in order.

Further, etiological uniformity does not result in neuro­ cognitive homogeneity. Clinically, the identification of

subtypes may allow for the generation of differential educa­

tional and habilitation strategies, and the subsequent

evaluation of the efficacy of those strategies in yielding

improved outcomes. In terms of research contributions, the

identification of subtypes may provide a structure of heter­

ogeneity useful in serial or longitudinal studies. The

results of longitudinal studies can provide information

regarding the effect of varying levels of neuropsychological

70 79 inefficiency on systematic change over time.

Statistical analysis yielded three clusters with pre­ dominantly developmentally handicapped members. In terms of profile analysis, these clusters were similar in that the speech, academic, and intellectual fundton scales were highly elevated, although at significantly different levels for each cluster. The clusters differed in their overall level of elevation, and the relative elevation of the more elementary scales and the memory scale. It can be hypothe­ sized that the magnitude of neuropsychological inefficiency is the basis of the difference between clusters. That is that the performance of the members of the mildly ineffi­ cient cluster Cl) may document inefficient operations in fewer areas of the brain than that of the moderately and severely inefficient clusters C3 and 2). In fact, tradi­ tional interpretation based on profile analysis might yield the hypothesis that the mildly inefficient cluster Cl) demonstrated deficits with a basis in inefficient operations of the temporal and frontal regions, because of the rela­ tively higher elevations on the receptive and expressive speech scales. Similarly, analysis of the profiles for the moderate C3) and severe C2) inefficient clusters would yield a hypothesis of more diffuse inefficiency, with perhaps more severe involvement of the temporal and frontal regions.

Extreme caution should be utilized before applying these hypothesis clinically and the rationale for this 00 caution will be further discussed under the topic of future research. Nonetheless, it may be stated at this point that the source of the apparent inefficiencies can not be determined within the confines of this study, in the absence of more complete clinical assessment, including the documen­ tation of medical history, and especially of objective evi­ dence of brain dysfunction. Should future research repli­ cate these findings, one might conclude that children with profiles and item analyses corresponding with mild ineffi­ ciency (cluster 1) might have more potential to benefit from programming based on neuropsychological strengths and weak­ nesses, while children with profiles and item analyses corresponding with moderately and severely inefficient clusters (3 and 2) may benefit less.

Clinical categorization, based on item analysis, lends some support to these hypotheses. Recall that the distribu­ tion of clinical categories among clusters, shown in Table

B, documents a predominance of the NSM category (no sensori­ motor inefficiency) for the mildly inefficient cluster (1), and within that context, most of the subjects demonstrated a performance suggesting inefficiency in only 2 secondary areas. Furthermore, the moderately inefficient cluster (3) was predominantly categorized clinically as USM (inefficien­ cy in one sensorimotor area; 5 of 14 subjects) and BSM

(inefficiency in sensorimotor area, bilaterally; 8 of 14 subjects). Finally, clinical categorization for the 81 severely inefficient cluster CE3 was exclusively BSM (bilat­ eral sensorimotor) with the majority of the subjects having performances suggestive of inefficient operations in 1 secondary areas, and some with performances suggestive of inefficiency of operations of both tertiary areas.

Because findings should be viewed as inconclusive until replicated, preferably using other neuropsychological bat­ teries, including selected batteries, as will be discussed under future research, it is strongly recommended that clinical analysis of LNNB-C profiles for children in the range of mild mental retardation be based on extensive item analysis, rather than profile analysis.

Clinically, based on item analysis, the typical profile reflected difficulty with tasks believed to be subserved, according to Luria’s model, by the secondary temporal and frontal areas, and by the posterior tertiary area. Varia­ tions in this profile were seen in the direction of milder inefficiency in some of the more elementary functions sub­ served by these areas. In the direction of more inefficien­ cy, difficulties were demonstrated in other secondary areas, and for four subjects, in the tertiary anterior area.

Statistical analysis documented significant agreement be­ tween statistical and clinical subtypes, based on extent of sensorimotor inefficiency, with the clinical subtypes sug­ gesting more neuropsychological inefficiency predominantly from clusters 2 and 3, and those reflecting less 82 inefficiency from cluster 1. Although extent of sensorimo­ tor inefficiency was used in the statistical analysis, ex­ amination of Table 6 reveals an association between extent of sensorimotor inefficiency and extent of secondary ares inefficiency; less sensorimotor inefficiency is associated with less secondary area inefficiency. More subjects CIO of

17) categorized as NSM Cperformance reflecting no senso­ rimotor inefficiency) demonstrated performance suggesting inefficiency in only 2 secondary areas, and these subjects were members of the mildly inefficient cluster. More sub­ jects C5 of 9) categorized as USM (unilateral sensorimotor inefficiency) demonstrated performances suggestive of inef­ ficiency in 3 secondary areas, and these subjects were all members of the moderately inefficient cluster (30. Similar­ ly, subjects categorized as BSM were predominantly also classified as demonstrating inefficiency in 4 secondary areas, and 11 of these subjects belonged to the severely inefficient cluster 92) and 4 to the moderately inefficient cluster (3). This association suggests that, in the mildly mental retarded population, there may be an relationship between extent of inefficiency in the areas of the brain subserving, according to Luria’s model, the more elementary neuropsychological functions, and level of performance, as reflected by cluster membership (i.e. overall level of elevation of profile). This potential relationship, if replicated in future research, lends support to the early 83 identification of neuropsychological inefficiency, with more extensive efforts at remediation and compensation, than traditionally provided, and with efforts more based in a neuropsychological model.

In addition to the finding that the LNNB-C is useful for identifying subtypes, a significant finding was yielded regarding the myelomeningocele subsample. This subsample demonstrated heterogeneity across clusters and clinical categories, with 3 of the 10 subjects demonstrating the mildest degree of neuropsychological inefficiency, albeit these three are believed to fall outside the intended IQ range. This finding is consistent with literature cited in chapter S, supporting the presence of a wide range of func­ tional capacity in this diagnostic category. Clearly, the medical diagnosis of myelomeningocele does not in and of itself effect neuropsychological function in a consistent manner, and therefore, this diagnostic group may not consti­ tute an neuropsychological group. In this regard, past investigations have suggested that there are variables within this population which contribute to heterogeneity in terms of psychological outcome (Mazur et al, 1988). Lesion level, number of shunt procedures, socio-economic factors, and educational opportunities were considered to be relevant variables influencing outcome. Although information rele­ vant to all these variables was not gathered within the present study, simply considering lesion level and number of 8% shunt procedures did not predict level of neuropsychological inefficiency for this small subsample. Based on Mazur's findings, in the present study unexpected cluster membership included one child with a thoraco-lumbar level lesion in the minimally inefficient cluster (5), two in the mildly ineffi­ cient cluster CID, and three with lumbar-sacral level le­ sions in the moderately inefficient cluster C3D, albeit one with a history of meningitis. Given the small number of subjects in the subsample, the limited amount of medical information available, and the lack of information regarding socio-economic status and educational opportunity, one cannot discount Mazur’s findings. However, these unexpected cluster memberships do point to the importance of not making assumptions regarding psychological, nor neuropsychological, function merely from lesion level and number of shunt proce­ dures alone. Information regarding the extent of medical complications surrounding shunt procedures, and information relevant to socio-economic and educational variables must be considered. Of course, clinically, in all fairness to the individual child, full assessment is indicated before making assumptions regrading psychological and neuropsychological function.

The lack of statistically significant differences between various groupings of the experimental population on the composite adaptive behavior scores is somewhat unexpect­ ed and requires some explanation. First, this finding is 85 not consistent with previous findings suggesting a rela­ tionship between autonomic function and adaptive behavior

CBerntson et al., 1385; Tuber et al., 1985; Barclay &

Leland, 1983). Two differences between this previous inves­ tigation and the present investigation are noteworthy.

First, the previous investigation utilized direct psycho- physiological measures and the present, indirect behavioral measures. The use of indirect behavioral measures of neuro­ psychological function, in comparison to direct physiologi­ cal measures, allows for the influence of many non-neurolog- ically based variables, such as quality and quantity of education and reinforcement history with a basis in both home and school experiences. Second, the population studied in the previous investigation may have been more representa­ tive of the developmentally disabled population in general than that of the present study. The subjects in the previ­ ous investigation were all from a community program for the mentally retarded and developmentally delayed, and included children from a variety of medical diagnostic groups, and most likely, with a wider range of intellectual function.

Subjects in the present investigation were selected within a limited range of intellectual function, but more signifi­ cantly, from a more restricted range of educational place­ ments. Most subjects were in public school classes for the developmentally handicapped, excluding, not intentionally, subjects placed in classes for children with severe 8 6

behavioral handicaps, and multiple handicaps. Most public

school settings include training relevant to adaptive behav­

ior in their programs, and this training may be very effec­

tive, thus the children in the sample are somewhat homoge­ neous in this respect. When the training does not work, the

child is most likely placed in another type of class, such as a class for the behaviorally disordered. Thus, although

educational placement practices may vary somewhat across communities, it is likely that children with less adequate

levels of adaptive behavior were excluded from this study, unintentionally, and represent an 'untapped' subsample of

the mildly retarded population. This untapped subsample

should be systematically included in future investigations.

The lack of significant differences between the various groupings of the sample on composite adaptive behavior

scores is also not consistent with Trites (1986) findings regarding a relationship between neuropsychological dysfunc­

tion and adaptive behavior as measured by work history. Two

differences are worthy of comment. First, Trites’ sample

was not complicated by restriction based on educational

placement, and second, the measure of adaptive behavior also

differs. Future studies including a measure of adaptive

behavior which is restricted to performance in a selected

specific situation may yield different results, as compared

to that of the present study where measure of adaptive 87 behavior is based on the teracher or parent’s perception of the child’s behavior in broader contexts.

A second point of explanation relevant to the lack of significant differences between the various groupings of the experimental population on composite adaptive behavior score concerns the nature of the norms utilized. The standardiza­ tion sample for the ABS-SE was drawn from the original

California sample utilized for the Adaptive Behavior Scale-

Public School Version CABS-PSVl (Lambert, Ulindmiller, &

Cole, 1975) plus some additional samples from California and

Florida. The samples included individuals from a variety of ethnic groups and levels of socio-economic status. The total sample consisted of 6500 individuals, ages three through sixteen, with 2600 of those from the ABS-PSV, gath­ ered in 1972. Because expectations may vary across geo­ graphical regions, these norms may not be optimal for use with the experimental sample. In addition, educational placement practices have changed since 1972, and the EMR class of 1972 may differ drastically in complexion from the

DH class of 1990. Analysis based on geographically relevant and more recently established norms may yield a different result.

A final point relevant to the unexpected finding of no significant differences between the various groupings of the experimental sample on composite adaptive behavior score is related to the issue of qualitative versus quantitative 8 8 analysis. The statistical analysis based on composite adap­ tive behavior scores is entirely quantitative. More exten­ sive clinical analysis of an individual child's functioning may reveal not Just whether a child copes, but how, and at what, or whose, expense. Coping style may differ between the various groupings of the sample. Further data would need to be gathered in order to determine the presence of stylistic differences.

Whatever the basis, although there were no significant statistical differences between clusters, clinical catego­ ries, nor diagnostic groups for adaptive behavior compos­ ite scores, a post hoc analysis based on the clinical sig­ nificance of factor scale scores did reflect some differ­ ences. This suggests that, although other variables may contribute more variance to adaptive behavior, neuropsycho­ logical function may make a significant contribution to variance. Again, although more research is indicated, partially because of limitations of this study, which will be discussed later, this finding, especially if replicated using a larger sample, has both clinical and research sig­ nificance. Given that neuropsychological function may be a variable contributing significantly to adaptive behavior, prototypical problems can be identified and differential management strategies can be generated. Although neuropsy­ chological variables and environmental variable always interact to produce given neurocognitive and behavioral 89

constellations, adaptive behavior may be an area where the

interaction between neuropsychological variables and envi­

ronmental variables is extremely salient, in regard to the

manner in which the individual copes with the

neuropsychological variables in the face of environmental

demands. It may be that this interaction contributes more

variance than either variable independently.

THEORETICAL SIGNIFICANCE

Although it was not the intent of this study to test

Luria’s theory, some comment seems in order. Tarter and

Slomka (1983) describe Luria as espousing a defect model of

impaired language capacity for mentally retarded individua­

ls, as well as a dysfunction of incompletely developed ter­

tiary association networks. Other experts have discussed

Luria’s hypothesis that the major problematic area for mentally retarded individuals is related to the development

of functions inherent to the second signal system (personal

communication, Henry Leland, 1987). Findings regarding the

clinical profiles of the sample as a whole, reflecting inef­

ficiency in the secondary areas of the temporal-frontal sys­

tem, as well as in the posterior tertiary area are consis­

tent with these hypothesis. Future research could further

explore the validity of Luria’s theory, as well as other

theories. so

LIMITATIONS AND FUTURE RESEARCH

There are several limitations to the present study, including those related to sample selection, those inherent to the LNNB-C, the nature of and extent of supplemental subtests, and the absence of clinically relevant informa­ tion. Sample selection was influenced by willingness to volunteer, both on the part of school systems and individual parents within those systems. One major result was that children were predominantly volunteered from developmentally handicapped classes, with children from classes for those with severe behavioral handicaps CSBHD infrequently volun­ teered. This may have effected the type of neuropsychologi­ cal profiles obtained, and also the type of adaptive behav­ ior profiles obtained. It could be assumed that the school systems select the more behaviorally disordered children out of developmentally handicapped classes and into SBH classes.

Limitations inherent to the LNNB-C include its left- hemisphere bias. Although right-hemisphere inefficiencies can be identified, left-hemisphere inefficiencies are more strongly reflected. This has resulted in a common clinical practice of supplementing the LNNB-C with tests which tap right-hemisphere functions. Subtests such as these were not administered in the present study. In addition, given the 91 k complexity of the psychological construct of memory, the

LNNB-C memory scale contains a limited number of items tapping the various components of verbal memory, with even more limitations relevant to figurai memory. Finally, the supplemental subtests, administered with the intention of tapping executive or prefrontal functions, although appro­ priate for chronological age, may not have been developmen- tally appropriate for the population. Developmental appro­ priateness of neuropsychological instruments, and especially those intending to measure prefrontal-type functions, is a major issue in developmental neuropsychology. Assessment methods which tap these functions at a earlier developmental stage, and are not a downward extension of adult tests, are being studied (Welsh & Pennington, 1980; Welsh, Pennington,

& Groisser, in press), and may be available for future research with delayed populations.

Finally, a full clinical evaluation was not performed.

Only the neuropsychological battery and the Adaptive Behav­ ior Beale, Bchool Edition were administered, and IQ was based on administrative records. This excluded relevant information such as medical, family, developmental, and school history, qualitative information based on observa­ tions during assessment of intellectual function, as well as intraindividual comparison of verbal and performacne scores, and spread of subtest scores. A full clinical evaluation should result in more refined clinical subtyping, as well as 92 contribute to more extensive description of the cluster members.

Both the findings and the limitations of the study assist in developing a line of research based on the present investigation. First, a larger sample, stratified for age, cognitive funcitoning, and school placement is indicated before results can be generalized. Second, the optimal neuropsychological battery needs to be defined. This should involve the supplementation of the LNNB-C with developmen­ tal ly appropriate subtests which tap nonverbal information- processing abilities, verbal and figurai memory functions, and executive functions at an early stage of development.

Concurrently, studies comparing the LNNB-C with other pedi­ atric batteries, not exclusively based on rate of learning, need to be carried out, in order to define a research bat­ tery which best identifies neuropsychological strengths and weaknesses in this population.

Third, future research should include full clinical evaluations. In addition to the administration of the optimal neuropsychological battery, information should be collected regarding family and social, developmental and school history, and assessments of intellectual ability and academic achievement should be carried out. Also, full clinical evaluaitons will allow for an optimal balance between qualitative and quantitative information regarding psychological, neuropsychological, and behavioral funciton. 93

Once this has been accomplished, two further directions for future research remain. First, the efficacy of the educational and habilitative strategies generated from neuropsychological evaluation will need to be determined.

Although a number of studies have reported findings that children with subtypes of neuropsychological function re­ spond differently to various types of interventions CLyon,

Moats, & Flynn, 1988), these studies have not included children with psychometrically measured intellectual func­ tion in the range of mild mental retardation, nor have they utilized the LNNB-C. Second, serial assessments should be carried out in order to determine the longitudinal course of the different neuropsychological subtypes, espeically under the condition of opitmal intervention.

SUMMARY

Findings from the current investigation support validi­ ty of the LNNB-C for identifying subtypes of neuropsycholog­ ical functioning in school age children with psychometrical­ ly measured intellectual function in the range of mild mental retardation. Although the findings of this study are informative, they should not be generalized without further investigation. This includes the investigation of the relative efficacy of other neuropsychological batteries for defining neuropsychological subtypes, utilization of a large stratified sample, and the utilization of a full clinical 94 evaluation along with the neuropsychological battery. References

Barclay, L. & Leland, H. (1983, August5. Behavioral corre­ lates of autonomic Function in developmentallu disabled infants. Paper presented at the annual meeting of the American Psychological Association, Anaheim, California.

Berntson, G. E ., Ronca, A. E ., Tuber, D. S., Boysen, S. T., & Leland, H. (1985). Cardiac reactivity and adaptive behavior. American Journal on Mental Deficiencu. 89(40), 415-419.

Bolter, J. F ., & Long, C. J. (1985). Methodological issues in research in developmental neuropsychology. In L. C. Hartlage and C . F . Tslzrow (Eds.), The neuropsucholoou of individual differences (pp. 41-59). New York: Plenum Press.

Carr, M . A., Sweet, J. J., & Rossini, E. (1986). Diagnos­ tic validity of the Luria-Nebraska Neuropsychological Battery-Children's Revision. Journal of Consulting and Clinical Psucholoou. 54(3). 354-358.

Chelune, G. J., & Baer, R. A. (1986). Developmental norms for the Wisconsin Card Gorting Test. Journal of Clini­ cal and Experimental Neuropsucholoou. 8(3), 219-228.

Coyle, I., Wayner, M . J., and Singer, G. (1980). Behavior al teratogenesis: A critical evaluation. In T . V . N . Persaud, (Ed.), Advances in the studu of birth defects: Vol. 4. Neural and behavioral teratoloou. Baltimore: University Park Press.

Davids, A., Goldenberg, L ., & Laufer, M . W . (1957). The relation of the Archimedes Spiral Aftereffect and the Trail Making Test to brain damage in children. Journal of Consulting Psucholoou. 21(5). 429-433.

Davis, R . D . , Adams, R . E ., Gates, D . 0., & Cheramie, G . M . (1989). Screening for learning disabilities: A neuropsychological approach. Journal of Clinical Psucholoou■ 45(3), 423-428.

95 36

Davison, L. A. (1974). Current status of clinical neuro­ psychology. In R. M. Reitan and L. A. Davison (Eds.) Clinical neuropsucholoou: Current status and applica­ tions (pp. 325-362). Washington, D.C.: V. H. Winston & Sons.

Fischer, R. S. (1908). The role of neuropsychological assessment in behavioral medicine with the developmen- tally disabled. In D. C. Russo and J. H. Kedesdy, (Eds.) Behavioral medicine with the developmentallu disabled (pp. 143-159). New York; Plenum Press.

Fletcher, J. M . , Milner, M . E ., & Ewing-Cobbs, L. (1967). Age and recovery from head injury in children: Devel­ opmental issues. In H. 5. Levin, J. Grafman, & H. M . Eisenberg (Eds.) Neurobehavioral recoveru from head in.luru (pp. 279-291). New York: Oxford University Press.

Fletcher, J. M., & Taylor, H. G . (1984). Neuropsychologi­ cal approaches to children: Toward a developmental neuropsychology. Journal of Clinical Neuropsucholonu. 6(1), 39-56.

Gasser, T., Pietz, J., Schellberg, D ., & Kohler, W. (1988). Visual evoked potentials of mildly retarded and control chiIdren. Developmental Medicine and Child Neuroloou. 30, 638-645.

Geary, D. C . , & Gilger, J. W. (1984). The Luria-Nebraska Neuropsychological Battery-Children’s Revision: Com­ parison of learning disabled and normal children matched on full scale IQ. Perceptual and Motor Skills. 58, 115.118.

Geary, D . C . , Jennings, 5. M ., 5chultz, D. D ., & Alper, T. 6 . (1984). The diagnostic accuracy of the Luria-Ne­ braska Neuropsychological Battery-Children’s Revision for 9 to 12 year old learning disabled children. 5chool Psucholoou Review. 13(3). 375-380.

Goldberg, E ., & Costa, L. D. (1981). Hemisphere differenc­ es in the acquisition and use of descriptive systems. Brain and Language. 14. 144-173.

Golden, C. J. (1987). Luria-Nebraska Neuropsucholooical Batteru: Children’s Revision, manual. Los Angeles: Western Psychological Services. 37

Golden, C. J., & Ullkening, G. N. (1981). Neuropsychologi­ cal bases of exceptionality. In R . T. Broun and C . R . Reynolds CEds.), Psucholooical perspectives on child­ hood exceotionalitu (pp. 61-30). New York; John Uliley & Sons, Inc.

Gorenstein, E ., Mammato, C . A., S Sandy, J. M. (1383). Performance of inattentive-overactive children on selected measures of prefrontal-type function. Journal of Clinical Psucholoou. 45(40. 813-632.

Grossman, H. J. (Ed.) (1383). Classification in mental retardation. (rev. ed.). Washington, D. C .: American Association on mental Deficiency.

Gustavson, J. L . , Golden, C . J., Wilkening, G. N ., Hermann, 6 . P., Plaisted, J. R ., Maclnnes, W . D ., & Leark, R. A. (1384). The Luria-Nebraska Neuropsychological Battery- Children’s Revision; Validation with brain-damaged and normal children. Journal of Psuchoeducational Assess­ ment. 2, 133-208.

Harbord, M . 6., Finn, J. P., Hall-Craggs, M . A., Robb, S. A., Kendall, B. E ., Boyd, S. G. (1330). Myelinatic., patterns on magnetic resonance of children with devel­ opmental delay. Developmental Medicine and Child Neuroloou. 32. 235-303.

Heaton, R. K . (1381). Wisconsin card sorting test manual. Odessa, Florida: Psychological Assessment Resources, Inc.

Huballah, M . Y ., & Hoffman, H. J. (1387). Early repair of myelomeningocele and simultaneous insertion of ventric- uloperitonal shunt: Technique and results. Neurosur- geru. 20(1), 21-23.

Kozulin, A. (1986). Vygotsky in context. In L. Vygotsky, Thought and language (rev. ed.) (pp. xi-lxi). Cam­ bridge, Massachusetts: The MIT Press.

Lambert, N. (1981). Diagnostic and technical manual : AAMD adaptive behaivor scale, school edition. Monterey, California: Publishers Test Service.

Lambert, N. M ., Windmiller, M ., & Cole, L. J. (1375). AAMD adaptive behavior scale. public school version. Washington, DC: American Associaotn on Mental Defi­ ciency . 38

Lambert, N., Windmiller, M . , Tharinger, Cole, L. C1981). AAMD Adaptive behavior scale, school edition. Mon­ terey, California: Publishers Test Service.

Leland, H. (1383). Adaptive behavior scales. In J. L. Matson & J. A. Mulick (Eds.), Handbook of mental retar­ dation (pp. 215-225). New York: Fergamon.

Levandowski, 0. H. (1385). Discrimination of normal. brain-damaged and conduct disordered children using the Luria-nebraska Children's Batteru. Unpublished doc­ toral dissertation, Rosemead School of Psychology, La Mirada, California.

Luria, A. R. (1365). L. S. Vygotsky and the problem of localization of functions. Neuroosuchologia. 3, 387- 332.

Luria, A. R. (1366). Higher cortical functions in man. New York: Basic Books, Inc.

Luria, A. R. (1373). The working brain, an introduction to neuroDSuchologu. New York: Basic Books.

Lyons, 6. R ., Moats, L ., & Flynn, J. M. (1388). From assessment to treatment: Linkage to interventions with children. In M. 6. Tramontana & S. R . Hooper (Eds.). Assessment issues in child neuroosuchologu. (pp. 113- 142). New York: Plenum.

Mazur, J. M., Aylward, G. P., Colliver, J., Stacey, J., & Menelaus, M. (1388). Impaired mental capabilities and hand function in myelomeningocele patients. Zeit- schrift fur Kinderchiruroie 43(Suppl. II), 24-27.

McLone, D., Czyzewksi, D., Raimondi, A. J., and Sommers, R. C. (1382). Central nervous system infections as a limiting factor in the intelligence of children with myelomeningocele. Pediatrics. 70(3). 330.

Mittelmeier, C ., Rossi, J. S., & Berman, A. (1383). Dis­ criminative ability of the Trail Making Test in young children. International Journal of Clinical Neuro- Dsuchologu. 11(4). 163-166.

Morris, R . , Blashfield, R ., & Satz, P. (1386). Develop­ mental classification of reading-disabled children. Journal of Clinical and Experimental Neuroosuchologu. 8(40), 371-332. 59

Nolan D. R., Hammeke, T. fit., & Barkley, R. A. (1983). fit comparison of the patterns of the neuropsychological performance in two groups of learning disabled chil­ dren . Journal of Clinical Child Psucholoou. ISC 1), 22- 27.

Pavlov, I. (1957). Experimental psucholoou and other es- saus. New York: Philosophical Library.

Plaisted, J. R ., Gustavson, J. L ., Wilkening, G . N . & Gold­ en, C . J. (1983). The Luria-Nebraska Neuropsychologi­ cal Battery-Children’s Revision; Theory and current research findings. Journal of Clinical Child Psucholo- gy., 12(1), 13-21.

Reines, 5., & Goldman, J . M . (1980). The development of the brain. Gpringfield: Thomas.

Reitan, R. M. (1971). Trail Making Test results for normal and brain-damaged children. Perceptual and Motor Skills. 33, 575-581.

Robbins, S. L ., Cotran, R. S., & Kumar, V. (1981). Patho- lonic basis of disease (3rd ed.). Philadelphia: W. B. Saunders Company.

Rourke, B. P. (1982). Central processing deficiencies in children: Toward a developmental neuropsychological model. Journal of Clinical Neuropsucholoou. 1(1), 1-18.

Rourke, B. P., Bakker, D. J., Fisk, J. L ., & Strange, J. 0. (1983). Child neuropsucholoou an introduction to theoru. research, and clinical practice. New York: Guilford Press.

Rourke, B . P., & Finlayson, M. A. J. (1975). Neuropsycho­ logical significance of variations in patterns of performance on the Trail Making Test for older children with learning disabilities. Journal of Abnormal Psu- cholopu. 81(10), 112-121.

Selz, M ., & Reitan, R. M. (1979). Rules for neuropsycho­ logical diagnosis: Classification of brain function in older children. Journal of Consulting and Clinical Psucholoou. 17(2), 258-281.

Siegel, A. W . , Bisanz, J., & Bisanz, 6. L. (1983). Devel­ opmental analysis: A strategy for the study of psycho­ logical change. In 0. Kuhn & J. A. Meacham (Eds.), Contributions to human development (pp. 15-81). Basel: 5. Karger. 1 0 0

Skoff, B. F. (1988). The utility of neuropsychological assessments of mentally retarded individuals. In D. C. Russo & J. H . Kedesdy CEds.), Behavioral medicine with the developmentallu disabled (pp. 181-170). New York: Plenum Press.

Snow, J. H ., & Hynd, G. W. (1985). Journal of Psuchoedu­ cational Assessment. 3, 101-109.

Snow, J. H ., Hynd, 6. W ., & Hartlage, L. C. (1984). Dif­ ferences between mildly and more severely learning- disabled children on the Luria-Nebraska Neuropsycholog ical Battery-Children’s Revision. Journal of Psucho­ educational Assessment. 2, 23-28.

Spreen, 0., Tupper, D ., Risser, A., Tuokko, H ., & Edgell, 0. (1984). Human developmental neuropsucholoou. New York: Oxford University Press.

Teuber, H. L ., & Rudel, R . G . (1962). Behavior after cerebral lesions in children and adults. Dave1opmenta1 Medicine and Child Neuroloou. 4, 3-21.

Tramontana, M. G ., & Sherrets, S. D. (1985). Brain impair­ ment in child psychiatric disorders: Correspondences between neuropsychological and CT scan results. Jour­ nal of the American Academu of Child Psuchiatru. 24(5), 590-596.

Trites, R. L. (1986). Neuropsychological variables and mental retardation. Psuchiatric Clinics of North America. 9(4), 723-731.

Tuber, 0. 5., Ronca, A . E ., Berntson, G . G ., Boysen, 5. T ., & Leland, H. (1985). Heart rate reactivity, habitua­ tion and associative learning in developmentally dis­ abled preschool children. Phusiolooical Psucholoou. 13, 95-102.

Vygotsky, L . 5. (1960). Development of the higher mental functions. (Izd. Akad, Fed. nauk RSFSR; MOSCOW).

Vygotsky, L. S. (1962). Thought and language. Cambridge; Massachusetts Institute of Technology.

Vygotsky, L. S. (1965). Psychology and localization of functions. Neuroosuchologia. 3, 381-386.

Welsh, M. C. & Pennington, B. F. (1988). Assessing frontal lobe functioning in children: Views from developmental psychology. Developmental Neuroosuchologu. 4(3), 199- 230. 1 0 1

Welsh, h. C ., Pennington, B. F., & Graisser, D. B . (in press). A normative-developmental study of executive function: A window on prefrontal function in children. Developmental Neuropsucholoau.

Werner, H., & Kaplan, B. (1956). The developmental ap­ proach to cognition: Its relevance to the psychologi­ cal interpretation of anthropological and ethnolin- guistic data. American Anthropolooist. 58. BB6-B0O.

Wiss, C . A. (1982). A comparative study of the performance of learning disabled, mentally retarded, and normally achieving children on the Modified Wisconsin Card Sort­ ing Test. Unpublished doctoral dissertation, North­ western University, Evanston. APPENDIX A

LNNB-C CLINICAL AND SUMMARY SCALES

LISTING BY NUMBER AND NAME

Clinical Scales:

Cl Motor Functions Scale

CE Rhythm Scale

C3 Tactile Functions Scale

C4 Visual Functions Scale

C5 Receptive Speech Scale

C6 Expressive Speech Scale

C7 Writing Scale

C6 Reading Scale

CS Arithmetic Scale

CIO Memory Scale

Cll Intellectual Processes Scale

Summary Scales:

51 Pathognomonic Scale

52 Left Sensorimotor Scale

53 Right Sensorimotor Scale

1 0 2