Visuospatial and visuomotor performance in normal and disabled children


Material Information

Visuospatial and visuomotor performance in normal and disabled children
Physical Description:
Oglesby, Jennifer Marie, 1966-
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 28314248
oclc - 50298859
System ID:

Full Text







This dissertation is dedicated to my parents, Carolyn

M. Pickard and Ellis J. Oglesby, Jr., and to my brother,

Ellis J. Oglesby, III. Their love and consistent support

has been an integral part of my entire life. In particular,

this dedication is to my parents, for teaching me to have

confidence in myself and to pursue my dreams, and to my

brother, for his heart of gold and for reminding me by his

example to always remain true to my own values.


I wish to acknowledge many people for their support of

me and this project. I thank my chairperson, Eileen B.

Fennell, Ph.D., for her valuable comments and support and

for being a model of excellence as a scientist and

clinician. I also appreciate the helpful input and

assistance of my committee members, Russell M. Bauer, Ph.D.,

Sheila M. Eyberg, Ph.D., Michael Geisser, Ph.D., F. Joseph

Kemker, Ph.D., and Kytja Voeller, M.D. I particularly thank

Kytja Voeller, M.D., for allowing me access to her sample of

learning disabled children, and Sheila M. Eyberg, Ph.D., for

teaching me research skills as my master's thesis

chairperson. In addition, the assistance of Reid Skeel,

M.S., Melodye Gaskin, M.S., Margaret Booth-Jones, M.S., and

Timothy Tumlin, M.S., was crucial to completion of this

project. Last but definitely not least, I also thank my

husband, Gordon S. Freckleton, for his love and support, and

my cairn terrier, Misty, for bringing a little extra joy

into every single day since she entered my life.




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

LIST OF TABLES ......................................... vii

LIST OF FIGURES........................................ viii

ABSTRACT................................................... ix


1 REVIEW OF THE LITERATURE ......................... 1

Introduction..................................... 1
Visuomotor Skills in Human Development............ 1
Neuroanatomy of Visual Perception................. 3
The Retina, Optic Nerve, Optic Chiasm, and
Optic Tract .................................. 3
Systems of Visual Processing Other than the
Geniculostriate System ....................... 4
The Geniculostriate System ..................... 6
Neuroanatomy of Voluntary Motor Activity.......... 14
The Primary Motor Area ......................... 14
The Premotor Area .............................. 16
The Supplementary Motor Area .................... 17
The Corticospinal Tracts ....................... 17
The Rubrospinal Tract........................... 19
The Reticulospinal Tracts ...................... 19
The Striatum.................................... 20
Development of Visual Perception ................. 21
Development of Motor Skills ...................... 25
Visuoperceptual Disorders in Adulthood............ 29
Motor Impairments in Adults and Children......... 34
Cerebral Lateralization and Visuomotor Skills.... 37
Learning Disabilities: An Overview.............. 40
Dyslexia .................................. ..... 43
Dysgraphia ............................... ...... 46
Dyscalculia.................................... 47
Nonverbal Learning Disabilities ................ 48
Assessment of Visuomotor Skills .................. 49
The Hooper Visual Organization Test (HVOT)..... 50
The Rey-Osterrieth Complex Figure Test (ROCFT). 65
Summary of the HVOT and ROCFT Research with
Children..................................... 74


Problems with HVOT and ROCFT Research with
Children ..................................... 75
Rationale for this Project ....................... 76
Hypotheses for this Project ...................... 78

2 METHOD ........................................... 80

Subjects ......................................... 80
Measures ......................................... 82
Hooper Visual Organization Test (HVOT).......... 84
Rey-Osterrieth Complex Figure Test (ROCFT)..... 85
Wechsler Intelligence Scale for Children-
Revised Estimated IQ (EIQ) ................... 86
Multilingual Aphasia Exam Visual Naming (VN)... 88
Boston Naming Test (BNT) ....................... 89
Picture Completion (PC) Subtest of the WISC-R.. 90
Object Assembly (OA) Subtest of the WISC-R...... 90
Test of Visual Motor-Integration (VMI) .......... 91

3 PROCEDURE ........................................ 93

Procedure for Subjects with Neurological
Impairment ..................................... 93
Procedure for Subjects with Learning
Disabilities ................................... 94
Procedure for Normal Control Subjects ............ 94

4 RESULTS .......................................... 98

Inter-Rater Reliabilities ........................ 98
Demographic Information .......................... 102
Further Descriptive Data for the NI Sample ..... 105
Further Descriptive Data for the LD Sample ..... 112
Further Descriptive Data for the NC Sample ..... 114
Screening of Data ................................ 116
Bonferonni Adjustments ........................... 119
Relationships Between Dependent Variables and
Age......................................... ....... 119
Relationships Between Dependent Variables and
Sex......................................... ....... 123
Correlational Analyses ........................... 125
Determination of Covariates ...................... 130
Group Comparisons ................................ 133
Discriminant Analysis ............................ 134
Post Hoc Analyses ................................ 135
Summary of Results ............................... 142

5 DISCUSSION ....................................... 145

Evidence Related to Scoring Systems.............. 145
Mings' (1987) Scoring System for the ROCFT
Total Score .................................. 145
Waber and Holmes' (1985) Scoring System for
the ROCFT Organizational Score ................ 149


Inter-Rater Reliability for Beery's (1989) VMI. 151
Relationship of the HVOT and the ROCFT to Age.... 152
HVOT Total Score .................. ............ 153
HVOT Isolate Response Score..................... 154
ROCFT Total Score.............................. 154
ROCFT Organizational Score...................... 155
Implications of these Developmental Changes.... 156
Relationship of the HVOT and the ROCFT to Sex.... 157
Construct Validity for the HVOT and the ROCFT.... 159
HVOT Total Score ................................ 159
HVOT Isolate Response Score .................... 162
ROCFT Total Score............................... 163
ROCFT Organizational Score...................... 165
Differences Between Samples on the HVOT and the
ROCFT .......................................... 166
HVOT Total Score ................... .... ...... 167
HVOT Isolate Response Score .................... 168
ROCFT Total Score.............................. 169
ROCFT Organizational Score...................... 169
Implications of Differences Among Samples....... 170
Limitations of the Present Study.................. 172
Directions for Future Research ................... 175
Summary and Conclusions .......................... 178


REFERENCES ............................................. 188

BIOGRAPHICAL SKETCH.................................... 200




1 Inter-Rater Reliabilities ......................... 99

2 Frequency of Different Percentages of Agreement
Between the Raters........................... 100

3 Demographic Information ........................... 103

4 Diagnoses of the NI Sample ........................ 106

5 Demographic Data by Acuteness of Illness........... 108

6 Demographic Data by Tests Administered for NI
Sample ...................................... 110

7 Diagnoses of the LD Sample ........................ 113

8 Demographic Data for Geographic Subsets of NC
Sample................................... .......... 115

9 HVOT Scores by Condition and Age.................. 121

10 ROCFT Scores by Condition and Age ................. 122

11 HVOT and ROCFT Scores by Condition and Sex......... 124

12 Correlations Between the HVOT and Other
Variables.................................. 127

13 Correlations Between the ROCFT and Other
Variables.................................. 128

14 Correlations Between the ROCFT and Other
Variables for the Combined Sample........... 129

15 HVOT and ROCFT Scores by Condition by IQ
Level...................................... ........ 139




1 The Optic Nerve, Chiasm, and Tract ................ 5

2 The Geniculostriate Visual System ................. 8

3 Motor Areas and the Corticospinal Tracts........... 15

4 Timeline of Visual Development .................... 22

5 Timeline of Motor Development ..................... 26


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



Jennifer M. Oglesby

December 1994

Chairman: Eileen B. Fennell, Ph.D.
Major Department: Clinical and Health Psychology

Visuoperceptual ability and visuomotor skill in early

and middle childhood provide the foundations for more com-

plex cognitive functions as children grow older. The evalu-

ation of visuoperceptual and visuomotor performance is

therefore particularly important in the neuropsychological

assessment of children. However, necessity has forced many

pediatric neuropsychologists to employ adult assessment

instruments in their work with children despite somewhat

limited normative and psychometric information. Recent

research exploring the use of such measures in children has

supported the use of the Hooper Visual Organization Test

(HVOT) and the Rey-Osterrieth Complex Figure Test (ROCFT).

This project investigated developmental trends, construct

validity, and differences between normal and disabled

children's performance on the HVOT and the ROCFT in



The HVOT and the ROCFT, along with several other tests

of general, visuoperceptual, or visuomotor ability, were

administered to 25 normal control children between the ages

of 6 and 11. In addition, archival data were obtained from

the files of 43 neurologically impaired children and 32

learning disabled children aged 6-11 years.

Consistent with prior research, older children's

performances on the HVOT and the ROCFT were significantly

better than younger children's performances. Promising

preliminary support for the construct validity of the HVOT

and the ROCFT emerged in the form of several high

correlations between the HVOT and the ROCFT and other

measures of visuospatial, constructional, and visuomotor

ability. Finally, the ROCFT demonstrated some usefulness in

discriminating between neurologically impaired children and

normal or learning disabled children.

Thus, the HVOT and the ROCFT have the potential to

provide valuable information in the assessment of visuospa-

tial and visuomotor skills in children. Both the HVOT and

the ROCFT are sensitive to developmental changes in middle

childhood (6 through 11 years). Current evidence provides

promising preliminary support for their construct validity

as tests of visuospatial, constructional, and visuomotor

abilities. Although neither the HVOT nor the ROCFT revealed

differences between the normal and learning disabled

samples, the ROCFT was significantly worse in neurologically

impaired children and therefore has particular potential as


a tool in the assessment of neurologically impaired





Child neuropsychology, like adult neuropsychology,

relies on assessment instruments to measure the existence

and extent of neurological dysfunction. However, because

child neuropsychology is a relatively new field, the

usefulness of a number of instruments used with children

has not been extensively evaluated empirically. Indeed, a

number of adult neuropsychological tests have begun to be

employed for evaluation of children despite relatively

little information about the instruments' value in

assessments of children.

Visuomotor Skills in Human Development

Williams (1983) has pointed out that perceptual and

perceptual-motor tasks are particularly important in

childhood. Although perceptual and perceptual-motor skills

for older children and adults are not particularly

correlated with intelligence, studies with young children

(e.g., five year olds) display significant associations with

intelligence (Williams, 1983). Indeed, the ability to

perceive accurately and to interact with the environment

effectively has been identified as a virtual prerequisite

for cognitive development and academic achievement (Melamed



& Melamed, 1985; Williams, 1983). Therefore, evaluating

deficits in perception and in perceptual-motor tasks is an

important component of neuropsychological assessment of

children (Melamed & Melamed, 1985). Because humans rely so

heavily on vision to interpret environmental stimuli, visual

perception is of particular interest and importance.

Piaget's (1927/1977; 1936/1977; 1937/1977; Piaget &

Inhelder, 1969) theory provides a theoretical framework for

understanding the importance of visuomotor skills in human

development. Piaget proposed that the first year and a half

of life is a sensorimotor period, in which the child

develops greater awareness of his or her environment and

becomes increasingly aware of how he/she can affect that

environment. Development of visual perception and of motor

activity is key to this stage. The dawning awareness of the

"laws of nature" and of how he/she can influence the

environment is a prerequisite for the child to enter the

preoperational or symbolic stage, which lasts from 18 months

to 7 years of age, and in which the child begins to be able

to think and act symbolically. The child must have an

awareness of concrete information about the environment,

derived through visuomotor activity, and his or her role in

it before he/she can begin to represent that environment

symbolically. The symbolic activity of this stage is

necessary for the child to attain the next stage, formal

operations (ages 7 to 11), in which he/she begins to be able

to reason logically about concrete situations. Finally,


this logical reasoning is expanded to abstract situations,

and the child has attained formal operations. It is clear

from Piaget's theoretical model that visuomotor activity,

particularly of the sensorimotor and preoperational periods,

is crucial to the development of cognitive functioning

(Piaget, 1927/1977; 1936/1977; 1937/1977; Piaget & Inhelder,


With this theoretical model of the importance of

visuomotor skills in mind, children's development of

visuomotor skills will be explored. First, however, it is

important to understand what neuroanatomical areas are

involved in visuomotor tasks.

Neuroanatomy of Visual Perception

The Retina, Optic Nerve, Optic Chiasm, and Optic Tract

Visual perception begins in the retina, where the

primary receptors are rods (for vision in low intensity

light) and cones (for color vision and vision in high

intensity light). The highest proportion of cones is found

in the area of central vision, particularly the fovea

centralis. Bipolar cells, the primary afferent neurons of

the visual system, receive input from rods and cones as well

as from horizontal and amacrine cells. The bipolar cells

terminate on the ganglion cells which form the optic nerve

(Carpenter, 1991; Kandel, Schwartz, & Jessell, 1991;

Nieuwenhuys, Voogd, & van Huijzen, 1988; Spreen, Tupper,

Risser, Tuokko, & Edgell, 1984).


The optic nerve, the pathway for virtually all of the

fibers originating in the retina, moves posteriorly and

somewhat medially from the eyes before it decussates in the

optic chiasm: the nasal half of the retinal image of each

eye crosses to the other hemisphere (Figure 1). Thus, after

passing through the optic chiasm, visual information from

the right field of vision is processed in the left

hemisphere of the brain, and vice versa. After the

decussation of the optic nerve at the optic chiasm, the

ganglion cells form the optic tract. The optic tract runs

posteriorly and laterally (around the hypothalamus and

thalamus) before terminating in the lateral geniculate body

of the thalamus (Carpenter, 1991; Kandel et al., 1991;

Nieuwenhuys et al., 1988; Spreen et al., 1984).

Systems of Visual Processing Other than the Geniculostriate

Before turning to the geniculostriate system, the

system responsible for most visual processing, visual

systems responsible for functions such as visual attention,

visual orientation, and control of nonvoluntary eye

movements will be reviewed. One such system, which provides

input to assist in regulation of circadian rhythm, consists

of retinohypothalamic fibers which project directly from the

retina to the suprachiasmatic nucleus of the hypothalamus.

Another system, the extrageniculate pathway, consists of

fibers, called the brachium of the superior colliculus,

which proceed directly through the lateral geniculate body

eye optic chiasm
left visual field

| op i. lateral primary
optic nerve o it geniculate T visual
tract body cortex

right visual field radiation


Figure 1. The Optic Nerve, Chiasm, and Tract C1


to the superior colliculus and pretectal area, which send

efferent fibers to the pulvinar. The pretectal area also

transmits to the bilateral accessory oculomotor nuclei,

which send efferent fibers to the ciliary ganglia which

constrict the pupil in response to light. A complex system

involving multiple cortical and subcortical areas controls

orienting responses, tracking, and other eye movements via

input to the abducens and oculomotor nerves. Areas with

direct or indirect input into this system include the

superior colliculus, pretectal region, vestibular nuclei,

the ventral nucleus of the thalamus, the visual cortex

(primary and association areas), motor cortex, frontal eye

fields, auditory cortex, the pars reticulata of the

substantia nigra, and the paramedian pontine reticular

formation. Finally, the accessory optic system consists of

reciprocal fibers between the optic tract and the dorsal,

lateral, and medial nuclei of the accessory optic system,

which is located in the mesencephalon. Fibers from these

nuclei transmit to the cerebellum via the inferior olive,

and are involved in visual-vestibular interaction

(Carpenter, 1991; Kandel et al., 1991; Nieuwenhuys et al.,


The Geniculostriate System

The lateral qeniculate body

Most processing of visual stimuli occurs through the

geniculostriate system, which originates in the lateral


geniculate body. The lateral geniculate body of the

thalamus has six layers: three (layers 2, 3, and 5) for the

ipsilateral eye and three (layers 1, 4, and 6) for the

contralateral eye. Layers 1 and 2 are magnocellular, while

layers 3 through 6 are parvocellular (Carpenter, 1991;

Kandel et al., 1991; Nieuwenhuys et al., 1988).

Each retinal area involved in perception of the

contralateral visual field corresponds to a particular

location in the lateral geniculate body. Information from

the lower visual field is located medially in the lateral

geniculate nucleus, and information from the upper visual

field is located laterally. The macula, or area for central

vision, is represented caudally in the lateral geniculate

nucleus, while peripheral visual information is represented

laterally and rostrally (Carpenter, 1991; Kandel et al.,

1991; Nieuwenhuys et al., 1988).

Primary visual cortex

From the lateral geniculate nucleus, neurons project

via the retrolenticular portion of the internal capsule to

form the optic radiation (Figure 2). Most of the optic

radiation terminates in primary visual cortex (Brodmann's

area 17), located in the occipital lobe along the calcarine

fissure, and remaining fibers terminate in secondary and

tertiary visual cortex (Brodmann's areas 17 and 18). Most

fibers of the optic radiation terminate in layer IV.

Macular information is processed at the occipital pole, the

most caudal part of primary visual cortex, while peripheral

cortex temporal
(object recognition &
lateral primary color processing)
geniculate I visual
body v^L cortex

tertiary area
opti visual (motion detection a
radiation cortex spatial processing)

Figure 2. The Geniculostriate Visual System


information is processed in rostral areas of primary visual

cortex. The lower visual field is represented superior to

the calcarine fissure, with the upper visual field

represented inferior to the calcarine fissure (Carpenter,

1991; Kandel et al., 1991; Nieuwenhuys et al., 1988; Spreen

et al., 1984).

Cells in the primary visual cortex are more selective

in their responsiveness than retinal ganglion cells and

lateral geniculate body cells. Certain cells respond only

to thin rectangles of light, others to dark bars, others to

edges. In addition to type of stimuli, primary visual

cortex cells respond selectively to the orientation of

stimuli (Carpenter, 1991; Hubel & Wiesel, 1962; Kandel et

al., 1991).

Orientation columns

Cells with particular orientation biases form

orientation columns in the primary visual cortex.

Orientation columns consist of simple cells in layer IV and

complex cells in layers above and below layer IV. Simple

cells receive monocular input, and respond selectively to

stimuli of a particular type and orientation. Complex cells

receive binocular input from multiple simple cells, and

respond selectively to stimuli that move in a particular

direction. Interspersed in layers II and III of the

orientation columns are "blobs," which are not selective to

orientation, and which are involved in color vision


(Carpenter, 1991; Hubel & Wiesel, 1962; Kandel et al., 1991;

Nieuwenhuys et al., 1988).

Ocular dominance columns

In addition to ocular orientation columns, primary

visual cortex is organized in ocular dominance columns.

Each ocular dominance column responds primarily to input

from one eye (cells receiving monocular input react solely

to stimuli from that eye, while cells receiving binocular

input react preferentially to stimuli from that eye). The

ocular dominance columns alternate between left and right

dominance, so that adjacent columns have the opposite

dominance. These ocular dominance columns are involved in

combining the retinal images projected to each eye and are

necessary to depth perception. The only two regions of

primary visual cortex without ocular dominance columns are

those areas representing the "blind spot" in vision and

those areas which represent vision which is so peripheral

that it can only be perceived by the nasal portion of one

retina (Carpenter, 1991; Kandel et al., 1991; Nieuwenhuys et

al., 1988).

The ocular orientation columns and ocular dominance

columns form hypercolumns, which are a set of columns which

respond to stimuli in a particular small region of the

visual field. Such hypercolumns consist of two ocular

dominance columns (one per eye), and two full sets of

orientation columns (one set per ocular dominance column)

(Kandel et al., 1991).


Secondary and tertiary visual cortex

The secondary visual cortex (Brodmann's area 18)

surrounds the primary visual cortex and, along with tertiary

visual cortex (Brodmann's area 19), receives input from area

17 as well as direct input from the lateral geniculate body

and pulvinar. In addition, multiple commissural fibers

connect corresponding areas of visual cortex, resulting in

bilateral rather than contralateral representation of

stimuli. As in the primary visual cortex, cells in

secondary visual cortex respond to particular stimuli in a

specific orientation. However, cells in secondary and

tertiary visual cortex are mostly complex and hypercomplex

(Carpenter, 1991).

The magnocellular system

In addition to the classifications according to

columns, layers, and neuroanatomical location, the visual

processes can also be classified according to function and

the type of cellular input they receive from the lateral

geniculate body. Kandel et al. (1991) identify three

systems: the magnocellular system, parvocellular-interblob

system, and the parvocellular-blob system.

The magnocellular system, involved primarily in motion

detection but also with spatial information and depth

perception, projects from the magnocellular layers of the

lateral geniculate body to layers IV and VI in the primary

visual cortex. From area 17, the magnocellular system

proceeds to the thick stripe of secondary visual cortex as


well as to tertiary visual cortex and to the middle temporal

area, which is concerned with depth and motion perception.

The magnocellular system terminates in the ventral

intraparietal region, the medial superior temporal region,

and the posterior parietal region (involved in spatial

organization) (Haxby et al., 1991; Kandel et al., 1991;

Tranel, 1992). Consistent with the magnocellular system's

termination in parietal regions, McIntosh, Grady,

Ungerleider, Haxby, Rapoport, and Horwitz's (1994) positron

emission tomography results indicated that spatial tasks

localized to right occipitoparietal regions.

The parvocellular-interblob system

The parvocellular-interblob system (interblobs separate

blobs), involved in form perception and somewhat involved in

color perception, proceeds from the parvocellular layers in

the lateral geniculate body to the interblobs in layers II

and III of primary visual cortex. From primary visual

cortex, these cells project to the pale interstripe of

secondary visual cortex and then to tertiary visual cortex.

The parvocellular-interblob system terminates primarily in

the inferotemporal region (important for recognition of

form) and also projects to the posterior parietal lobe

(Kandel et al., 1991). The termination in the

inferotemporal region for this system is consistent with

recent results using positron emission tomography which

indicate that object identification localizes to

occipitotemporal regions (McIntosh et al., 1994).


The parvocellular-blob system

The parvocellular-blob system, involved in color

vision, progresses from the parvocellular layer of the

lateral geniculate body to the blobs in layers II and III of

primary visual cortex. The blobs in primary visual cortex

project to the thin stripe of secondary visual cortex, which

in turn transmits to tertiary visual cortex. The

parvocellular-blob system terminates in the inferotemporal


Other efferent projections of the visual cortex

In addition to the above-described projections to

parietal and temporal cortex, visual cortex efferents

project to the lateral geniculate body, the pulvinar, the

thalamic reticularis nucleus, the pretectal area, and the

superior colliculus (Haxby et al., 1991; Kandel et al.,

1991; Nieuwenhuys et al., 1988; Tranel, 1992; Williams,


The importance of the processing in parietal and

temporal regions, which receive associations from the visual

cortex, cannot be underestimated. In monkeys, disconnection

of visual cortex from such areas of the brain results in

apparent blindness, although the visual cortex displays

normal activation patterns (Nakamura & Mishkin, 1986;

Nakamura, Schein, & Desimone, 1986).


Neuroanatomy of Voluntary Motor Activity

Perception is only the first step in perceptual-motor

tasks. The neuroanatomical aspects of the motor component

of perceptual-motor skills will next be presented. The

major cortical areas for initiating and controlling

voluntary motor activity include primary motor cortex

(Brodmann's area 4), premotor cortex (lateral Brodmann's

area 6), and supplementary motor cortex (medial Brodmann's

area 6) (Figure 3).

The Primary Motor Area

The primary motor cortex, located in the posterior

precentral gyrus, is called agranular cortex because of

increased pyramidal cells in layer III and in layer V (where

the giant cells of Betz are located), thereby obscuring the

internal granular layer. The primary motor cortex

represents contralateral body parts sequentially,

beginning with the pharynx in the most inferior portion of

the precentral gyrus, and proceeding dorsally through the

tongue, face, fingers, arm, and knee, with the ankle and

toes represented at the medial inferior edge of the

hemisphere (Carpenter, 1991; Kandel et al., 1991;

Nieuwenhuys et al., 1988; Snell, 1992).

The primary motor cortex receives afferent fibers from

the ventral lateral and ventral posterior lateral nuclei of

the thalamus (these nuclei receive and thus probably relay

projections from the cerebellum, globus pallidus, and

premotor corticospinal
cortex tract
halamu (crosse d)

primary \anterior
motor corticospinal
cortexI tract
somato- (uncrossed)
ar anterolateral
supplementary decussation corticospinal
motor at level of tract
area medulla (uncrossed)

Figure 3. Motor Areas and the Corticospinal Tracts


substantia nigra), the primary somatosensory area,

somatosensory association areas, premotor cortex, and the

supplementary motor area. This set of input is consistent

with the primary motor area's role of initiating and

controlling voluntary movement. Primary motor cortex's

projections, primarily from layer V, form part of the

corticospinal tract. Other efferent projections transmit to

somatosensory cortex, the supplementary motor area, the

ventral lateral and ventral posterior lateral nuclei of the

thalamus, and the striatum. Ablations of the primary motor

cortex in monkeys result in short-term contralateral

paralysis, hypotonia, and areflexia, and long-term

impairment of skilled movement (Carpenter, 1991; Kandel et

al., 1991; Nieuwenhuys et al., 1988; Spreen et al., 1984).

The Premotor Area

Located anterior to the primary motor area, the

premotor area is agranular, but giant cells of Betz are

absent. The premotor area receives afferent projections

primarily from the posterior parietal cortex, but also from

the supplementary motor area and the ventral lateral nucleus

of the thalamus (which receives input from the globus

pallidus, substantia nigra, and the cerebellum). Like the

primary motor cortex, the premotor area forms part of the

corticospinal tract; in addition, the premotor area projects

to brain stem nuclei that are involved in medial descending

systems (particularly the reticulospinal system). The


premotor area seems to be involved in muscle control and in

spatial preparations for movement (e.g., orienting to a

target) (Carpenter, 1991; Kandel et al., 1991; Nieuwenhuys

et al., 1988).

The Supplementary Motor Area

The supplementary motor area, located on the medial

superior frontal gyrus, receives afferent fibers from

primary motor cortex, premotor cortex, somatosensory cortex,

the contralateral supplementary motor area, and the ventral

lateral nucleus of the thalamus (which receives input from

the cerebellum, globus pallidus, and substantia nigra). The

supplementary motor area projects to the primary motor area,

premotor area, somatosensory cortex, contralateral

supplemental motor area, caudate nucleus, putamen, the

corticospinal tract, and the ventral anterior and medial

dorsal nuclei of the thalamus. Lesions in the supplementary

motor area result in akinesia and poverty of spontaneous

speech. The supplementary motor area is involved in

carrying out programmed movement, planning movement, and

coordinating on-going complex movement (Carpenter, 1991;

Kandel et al., 1991).

The Corticospinal Tracts

Voluntary movements of the body and limbs are

transmitted from the brain via the spinal cord, in spinal

cord tracts called the corticospinal tract, the rubrospinal


tract, and the reticulospinal tracts. Other spinal tracts

(tectospinal and vestibulospinal) are not directly involved

in voluntary movement: the tectospinal tract is involved in

reactive postural movements, and the vestibulospinal tract

pertains to spinal reflexes and control of extensor motor

tone. The corticospinal tract is the only spinal tract to

originate directly from cerebral cortex and is the primary

pathway for voluntary, skilled movement. The corticospinal

tract fibers originate in layer V of the somatosensory

cortical areas (40%), primary motor area (31%), and the

premotor area (29%). Fibers from these areas converge with

thalamocortical fibers in the corona radiata, continue

through the internal capsule, and form the crus cerebri in

the midbrain and the cerebral peduncle in the brainstem.

These fibers form the pyramids in the medulla and also

transmit fibers to the pretectum, superior colliculus, and

medial reticular formation. At the convergence of the

medulla and spinal cord, the corticospinal tract partially

decussates, forming the lateral corticospinal tract

(crossed), the anterior corticospinal tract (uncrossed), and

the anterolateral corticospinal tract (uncrossed)

(Carpenter, 1991; Kandel et al., 1991; Nieuwenhuys et al.,


The lateral corticospinal tract is the largest of these

tracts, comprised of 75-90% of the fibers from the pre-

decussation corticospinal tract. After crossing over at the

level of the medulla, the lateral corticospinal tract is


located in the lateral funiculus, between the posterior

spinocerebellar tract and the fasciculus proprius. The

anterior corticospinal tract is medial, next to the anterior

median fissure, and most of its fibers cross in the anterior

white commissure at upper cervical levels. The

anterolateral corticospinal tract is found in the lateral

funiculus, ventral to the lateral corticospinal tract. Its

fibers, still uncrossed, terminate in the base of the

posterior horn and the interior gray (Carpenter, 1991;

Nieuwenhuys et al., 1988).

The Rubrospinal Tract

The rubrospinal tract originates in the magnocellular

region of the red nucleus in the midbrain tegmentum, which

receives afferent projections, largely via the cerebral

peduncle, from the cerebral cortex, including motor cortex,

and the cerebellum. The rubrospinal tract is involved in

controlling tone in flexor and extensor muscle groups. The

rubrospinal tract decussates in the ventral tegmentum, and

then proceeds in the lateral funiculus, with some fibers

intermingled with and others anterior to the lateral

corticospinal tract (Carpenter, 1991; Nieuwenhuys et al.,


The Reticulospinal Tracts

The reticulospinal tracts include the pontine

reticulospinal tract and the medullary reticulospinal tract,


and pertain to facilitating or inhibiting voluntary

activity, affecting muscle tone, and influencing

respiration, circulation, and sensation. The reticular

formation receives multiple afferent projections, especially

from motor cortex, premotor cortex, the superior colliculus,

and the parietal, occipital, and temporal lobes. The

pontine reticulospinal tract, comprised of largely uncrossed

fibers, originates in the nuclei reticularis, pontis,

caudalis, and oralis in the medial pontine tegmentum, and

descends in the medial anterior funiculus. The medullary

reticulospinal tract, which contains both crossed and

uncrossed fibers, originates in the medial medullary

reticular formation, especially in the nucleus reticularis

gigantocellularis (lateral to the paramedian region and

dorsal to the inferior olivary complex). The medullary

reticulospinal tract descends in the lateral funiculus and

has a wider distribution than the pontine reticulospinal

tract (Carpenter, 1991; Nieuwenhuys et al., 1988).

The Striatum

Nieuwenhuys et al. (1988) note that the striatum has

traditionally been considered "extrapyramidal" and therefore

largely irrelevant to the corticospinal systems described

above. Nieuwenhuys et al. (1988) argue that the striatum,

with its system of afferent and efferent systems, plays a

key role in voluntary motor activity. The mechanisms for

such a striatal influence include the striatal-pallidal-


thalamic connection to premotor cortex, the striatal-nigral

connection to the superior colliculus and reticular

formation, and the striatal-pallidal-nigral-thalamic

connection to the cortex via the nucleus pedunculopontines.

Development of Visual Perception

It is likely that the neuroanatomical structures which

adults employ in visual perception and in visuomotor tasks

serve similar functions in childhood. However, children's

ability to perform tasks requiring visual perception and

visuomotor skills increases over time, with corresponding

changes in the neuroanatomical structures (Kolb & Fantie,

1889; Spreen et al., 1984; Williams, 1983).

Newborn infants' visual behavior consists primarily of

detecting and orienting towards peripheral details or light

(Figure 4) (Graham, 1986; Kolb & Fantie, 1989; Lamb &

Campos, 1982; Spreen et al., 1984). Visual resolution is

quite poor at birth (Lamb & Campos, 1982; Willis &

Widerstrom, 1986). At about six weeks, binocular vision is

apparent and has become quite functional by around four

months of age (Spreen et al., 1984). Also at around four

months of age, the infant has begun to attend to nonmoving

stimuli at the center of the visual field (Gesell &

Amatruda, 1974). Color vision and conjugate eye movements

develop in the first six months of life (Gesell, 1977;

Graham, 1986; Kolb & Fantie, 1989; Spreen et al., 1984). As

the first year proceeds, the infant becomes more and more

detecting/orienting to peripheral stimuli; resolution poor
inocular vision apparent
binocular vision functional; attending to nonmoving stimuli in center of visual field
color vision and conjugate eye movements online
increased skill at recognizing visual stimuli
visuomotor coordination increasing
figure-ground differentiation improving
form constancy present
visual discrimination
excellent increased
of visuomotor
abilities, better
able to mentally
objects in space

1 1 2 3 4 5 6 7
6 months (YEARS)
4 months

6 weeks
Figure 4. Timeline of Visual Development


skillful at recognizing visual stimuli (Lamb & Campos, 1982;

Spreen et al., 1984).

Less dramatic changes in visual ability continue into

later childhood (Graham, 1986; Williams, 1983). Visuomotor

coordination (e.g., hand-eye coordination) grows rapidly

from three to five years of age (Williams, 1983). Figure-

ground differentiation is present at three to four years of

age and improves rapidly from four to six years of age

(Williams, 1983). Form constancy, or the idea that objects'

size and shape are permanent, regardless of the size or

shape of its retinal image, is present by three years of age

and increases greatly at six to seven years of age

(Williams, 1983). By five years of age, children have

become quite skilled at visual discrimination tasks and rely

almost solely on visual discrimination abilities to perform

all visuomotor tasks (Williams, 1983). At six years of age,

children are becoming more adept at both integrating

visuomotor abilities and at mentally manipulating objects in

space (Williams, 1983). All of these visual abilities

appear to have reached adult levels by nine or ten years of

age (Williams, 1983). Thus, many changes in visual

perception occur during middle childhood.

As noted earlier, changes in abilities are typically

related to neuroanatomical development. However, before

exploring what is currently known about the growth of the

neuroanatomical structures of vision, it is important to

understand the general processes of neuroanatomical


maturation. In many instances, neuroanatomical development

is highly related to the amount and type of interaction with

the environment (Carpenter, 1991; Kolb & Fantie, 1989; Lamb

& Campos, 1982; Spreen et al., 1984; Williams, 1983; Willis

& Widerstrom, 1986). The general processes of

neuroanatomical development include specialization of

function, cellular proliferation, cellular migration,

myelinization, an increase in the number of glial cells, the

growth of dendrites and axons, the continuing formation of

synaptic connections, and neurochemical development (Kolb &

Fantie, 1989; Menkes, Till, & Gabriel, 1990). For visual

processes, rapid changes occur in the first year (Kolb &

Fantie, 1989; Spreen et al., 1984; Willis & Widerstrom,

1986). In the first six months, the lateral geniculate

body, cortical neurons, retinal receptors, and

photoreceptors develop, and increased myelinization occurs

in the optic radiation (Kolb & Fantie, 1989; Willis &

Widerstrom, 1986). Between six and twelve months of age,

changes occur in the size of cells in the lateral geniculate

nucleus; in the development of the visual cortex; in the

number of glial cells, interneurons, and dendritic spines;

and in the extent of myelinization in the optic tract, optic

radiation, and corpus callosum (Kolb & Fantie, 1989; Spreen

et al., 1984).

Visual deprivation in animal studies demonstrates the

degree to which visual development depends upon visual

stimulation. In monkeys, although visual deprivation in the


early postnatal period has no impact upon the orientation

columns, any interference with binocular vision results in

decreased binocular cells in primary visual cortex

(Carpenter, 1991). In addition, monocular deprivation

during the first two months of life in kittens changes the

lateral geniculate body and the sizes of the ocular

dominance columns: lateral geniculate body cells for the

deprived eye are smaller than the nondeprived eye, and the

ocular dominance columns for the normal eye become larger

than the ocular dominance columns for the deprived eye

(Hubel & Wiesel, 1970). Such changes cannot be reversed by

later visual experience (Hubel & Wiesel, 1970; Wiesel &

Hubel, 1965).

Development of Motor Skills

Although visuomotor skills were discussed briefly, it

is important to also attend to the development of general

motor skills. At birth, reflexive and random motor

movements are present (Figure 5) (Kolb & Fantie, 1989;

Spreen et al., 1984; Willis & Widerstrom, 1986). These

reflexive movements include orienting towards tactile

stimulation of the face, blinking to light, sucking, and

extending limbs towards or withdrawing body parts from

certain stimulation (Kolb & Fantie, 1989; Willis &

Widerstrom, 1986). The infant slowly develops intentional

motor skills, starting with lifting the chin (one to two

months old) and reaching and grasping (by twelve weeks old)

reflexive and random movements becoming more adept at walking
lifting the chin feeding selves
reeaching end graeping dressing selves
bilateral movements bladder and sphincter control
holding head erect can copy details of drawing
most newborn reflexie have dis appeared orgeniaing drawinge
aore objects among hends and mouth

eerlr fine motor skills
standing and walking

1 year
9-10 months
6 months 6 years
3 months 3-4 years
birth 2 years

Figure 5. Timeline of Motor Development


(Gesell & Amatruda, 1974; Kolb & Fantie, 1989; Spreen et

al., 1984; Willis & Widerstrom, 1986). At three months, the

infant's movements tend to be bilateral, and at four months

of age, the infant becomes able to hold his/her head erect

and turn towards peripheral objects or sounds (Gesell, 1977;

Kolb & Fantie, 1989; Willis & Widerstrom, 1986). By six

months of age, many newborn reflexive movements have

disappeared (Willis & Widerstrom, 1986). Willis and

Widerstrom (1986) note that such reflexes, although they

encourage the infant's development of muscle tone, of

kinesthetic awareness, and of primitive skills for

interacting with the environment, would inhibit the

acquisition of complex motor programs such as walking.

In the second half of the first year, the infant is

able to grasp objects with two hands and manipulate objects

from one hand to the other and to the mouth (around six

months), learns to sit (seven to eight months old), crawl

(nine to ten months old), perform some fine motor skills

(around ten months old), and stand and walk (around one year

old) (Gesell, 1977; Gesell & Amatruda, 1974; Kolb & Fantie,

1989; Spreen et al., 1984). Reflexes are developed during

this period which tend to protect the child from injury and

to facilitate coordinated motor actions (Willis &

Widerstrom, 1986). During the second through fourth year,

children become more adept walkers, learn to feed and dress

themselves, and develop bladder and sphincter control

(Gesell, 1977; Gesell & Amatruda, 1974; Kolb & Fantie, 1989;


Spreen et al., 1984). In addition, during early childhood

increased coordination of various limb movements develops

(Gesell, 1977; Gesell & Amatruda, 1974; Kolb & Fantie, 1989;

Williams, 1983).

In terms of fine motor skills such as drawing, the

child is able to attend to features of a stimulus when

copying it at three and four years of age, but drawings show

minimal organization until about five years of age (Gesell,

1977; Williams, 1983). From five to seven years of age, the

ability to copy drawings rapidly increases and continues to

develop slowly until around ten years of age (Williams,

1983). From five to ten years of age, both gross and fine

motor programs become increasingly coordinated (Gesell,


The improved motor skills in infancy and early

childhood are probably related to neuroanatomical changes in

the first years of life, such as increased myelinization in

the spinal cord and maturation of cortical motor areas and

corticospinal connections (Spreen et al., 1984). In

particular, myelinization changes are observed in the

pyramidal tract in the first 12 months and in associational

areas and the corpus callosum in the first 24 months (Kolb &

Fantie, 1989).

Thus far, visual and motor functions in normal adults

and children have been explored. In addition to the

knowledge gained from examining normal functioning, a


variety of neurological disorders provide information about

mature visual perception and about visuomotor skills.

Visuoperceptual Disorders in Adulthood

A variety of disorders of visual perception have been

described (Bauer, 1993; Benton & Tranel, 1993; Ellis &

Young, 1991; McCarthy & Warrington, 1990). Benton and

Tranel (1993) classify these disorders into visuoperceptual,

visuospatial, and visuoconstructive skills. Visuoperceptual

skills include visual analysis (such as discrimination,

recognition, and figure-ground differentiation), visual

synthesis, and color recognition. Visuospatial skills

involve localization of objects in space, judgment of

direction and distance, and topographical orientation.

Visuoconstructive skills entail reproducing information via

assembly and drawing (Benton & Tranel, 1993).

Warrington and McCarthy's (1990) conceptualization of

visuoperceptual problems includes three primary processes:

visual sensory processing (e.g., ability to recognize

whether shapes are the same or different), perceptual

processing (e.g., ability to process unusual perspectives of

an object in order to recognize it), and visual semantic

processing (e.g., ability to recognize an object's


Because it lacks the ambiguity associated with trying

to make a distinction between "perception" and "semantic

processing," Ellis and Young's (1991) model of object


identification is more compelling (Bauer, 1993). Ellis and

Young (1991) conceptualize object recognition as consisting

of (1) perceiving the initial representation of the object

(i.e., information about edges, brightness contrasts, color,

etc.), (2a) forming a viewer-centered view of object (i.e.,

perceiving a distinct object as it appears in the visual

field) and (2b) forming an object-centered view of the

object (i.e., mentally manipulating the viewer-centered

object in space so that it can be more readily recognized),

(3) translating these representations of the object into

object recognition units, (4) accessing the semantic system

to ascribe meaning to the item, and (5) naming or otherwise

indicating recognition of the object. Although this schema

is an oversimplification in practice (e.g., patients

identified as having deficits accessing the semantic system

often have subtle basic perceptual difficulties as well), it

provides a useful framework for considering visuoperceptual


Observed diseases whose primary defect is in visual

perception and object recognition despite adequate visual

acuity include but are not restricted to achromatopsia,

simultagnosia, apperceptive agnosia, associative agnosia,

deficits in shape discrimination, color anomia, and optic

aphasia (Bauer, 1993; Benton & Tranel, 1993; Ellis & Young,

1991; Warrington & McCarthy, 1990). These disorders will be

discussed according to Ellis and Young's (1991) schema,

despite the oversimplification necessary for such classifi-


cation. Lesions of primary visual cortex result in "blind

spots" in vision and in impaired shape discrimination,

thereby interfering with the initial representation of the

object (Benton & Tranel, 1993; Ellis & Young, 1991;

Warrington & McCarthy, 1990). Achromatopsia, or inability

to see color in all or part of the visual field, also

hinders the initial representation of the object (Bauer,

1993; Benton & Tranel, 1993; Ellis & Young, 1991; Warrington

& McCarthy, 1990). Achromatopsia typically results from

lesions in the lingual and fusiform gyri or the white matter

at the posterior lateral ventricle (Bauer, 1993; Benton &

Tranel, 1993; Warrington & McCarthy, 1990).

Simultagnosia, or inability to perceive and integrate

multiple stimuli, probably reflects difficulties in forming

viewer- and object-centered views of multiple objects

(although simultagnosia is among the most difficult

disorders to classify into Ellis and Young's (1991) schema)

(Bauer, 1993; Benton & Tranel, 1993; Warrington & McCarthy,

1990). Simultagnosia is typically associated with bilateral

parieto-occipital lesions or occipitotemporal lesions

(Bauer, 1993; Benton & Tranel, 1993; Warrington & McCarthy,

1990). Apperceptive agnosia, or inability to recognize,

copy, or match an object, is probably also a deficit in

forming viewer- and object-centered views of objects (Bauer,

1993; Benton & Tranel, 1993; Ellis & Young, 1991; Warrington

& McCarthy, 1990). Apperceptive agnosia typically results

from parieto- or temporo-occipital lesions (Bauer, 1993;


Benton & Tranel, 1993; Warrington & McCarthy, 1990).

Deficits in translating from the viewer-centered to the

object-centered view of an object are associated with right

inferior parietal lesions (Ellis & Young, 1991; Warrington &

McCarthy, 1990).

Difficulties accessing the semantic system are probably

involved in associative agnosia, when the patient can copy

or match objects but cannot identify their function (Bauer,

1993; Ellis & Young, 1991; Warrington & Young, 1990).

Associative agnosia typically results from bilateral

occipitotemporal lesions, mesial occipital lesions, or right

occipital lesions (Bauer, 1993; Ellis & Young, 1991;

Warrington & McCarthy, 1990).

Color anomia (inability to name colors despite ability

to match) and optic aphasia (inability to name objects

presented visually despite understanding the object's

function and being able to name items based on auditory or

tactile information) represent deficits in naming (Bauer,

1993; Ellis & Young, 1991; Warrington & McCarthy, 1990).

Color anomia and optic aphasia are associated with left

posterior lesions (Bauer, 1993; Ellis & Young, 1990).

Warrington and McCarthy (1990) conceptualize

visuospatial disorders as including deficits in integrating

complex information for estimation of location (such

information includes the retinal image from each eye, and

information about head and eye position and movement). In

addition, visuomotor coordination involves coordination of


spatial information with somesthetic information about body

position. Attention to different spatial areas and ability

to mentally represent multiple items in space also play a

role in visuospatial phenomena.

Defects in visuospatial abilities include inability to

localize an item in space, discriminate an item's spatial

properties (e.g., bisect a line), distinguish position

(e.g., to recognize when two items are arrayed similarly on

their respective pages), recognize orientation (e.g., to

recognize whether two lines are parallel), search fully a

region of space, and detect or count items within an array

or series (Benton & Tranel, 1993; Warrington & McCarthy,

1990). In addition, visual neglect is an often dramatic

instance of ignoring a major portion of the spatial

environment (Benton & Tranel, 1993; Heilman, Watson, &

Valenstein, 1993; Warrington & McCarthy, 1990).

Visuospatial difficulties are typically related to

occipitoparietal lesions, frequently bilateral, although

temporal regions are also sometimes involved (Benton &

Tranel, 1993; Warrington & McCarthy, 1990). In particular,

optic ataxia, or impaired reaching for an object in space,

localizes to the posterior parietal or occipitoparietal area

(Benton & Tranel, 1993; Warrington & McCarthy, 1990).

Deficits in depth perception are associated with lesions in

superior visual association cortex (Benton & Tranel, 1993).

Visual neglect, which typically results from lesions in

temporoparietal regions, has also been associated with


lesions in the dorsolateral frontal lobe, cingulate gyrus,

striatum, thalamus, mesencephalic reticular formation, and

the posterior limb of the internal capsule (Heilman, Watson,

& Valenstein, 1993).

The final type of adulthood visuoperceptual problem

identified by Benton and Tranel (1993), visuoconstructive

disorders, relates to difficulties in both visual

perception, as described above, and praxis, as described


Motor Impairments in Adults and Children

Apraxia in adults and developmental dyspraxia in

children are disorders of skilled movements which have

implications for normal motor functioning (Cermak, 1985;

Heilman & Rothi, 1993). Developmental dyspraxia represents

a failure to develop coordinated motor programs; apraxia

refers to the loss of previously acquired motor programs

(Cermak, 1985; Heilman & Rothi, 1993). Developmental

dyspraxia tends to reflect a difficulty in sensory

integration and motor planning, whereas apraxia typically is

a deficit in executing sequenced skilled movements (Cermak,

1985; Rothi & Heilman, 1993).

Examination of developmental dyspraxia has indicated

that normal motor skill acquisition consists of three

phases: first, a cognitive understanding of the necessary

motor actions develops; second, movement becomes more

coordinated as practicing provides feedback; and third, a


motor program develops, allowing performance of the motor

sequence with minimal conscious attention (Goodgold-Edwards

& Cermak, 1989).

Dyspraxic errors correspond to these three steps, and

consist of errors in planning, in execution, and in specific

movements (Goodgold-Edwards & Cermak, 1989). Planning

errors can be related to cognitive deficits, spatial

disorientation, visuoperceptual or visuospatial problems,

tactile-perceptual problems, or a poorly developed body

schema (Cermak, 1985; Goodgold-Edwards & Cermak, 1989;

Murray, Cermak, & O'Brien, 1990). Execution errors occur

when visual, proprioceptive, or kinesthetic feedback is

deficient, or when pre-movement preparation is inadequate

(Goodgold-Edwards & Cermak, 1989). Specific movement errors

result from an inability to accurately produce intended

movements, or to perform particular components of a sequence

(Goodgold-Edwards & Cermak, 1989).

Cermak (1985) hypothesized several possible areas of

brain dysfunction which may result in developmental

dyspraxia. Thalamic lesions, particularly in the area of

the thalamus related to somatosensory and auditory

functioning, may contribute to deficits in somatosensory

functions (Cermak, 1985). Cerebellar dysfunction also

results in deficits in coordinating motor sequences (Cermak,

1985). Implicit in Cermak's (1985; Goodgold-Edwards &

Cermak, 1989) conceptualization of dyspraxia is the

potential for dyspraxia to result from deficits in spatial


processing (e.g., posterior parietal regions), somatosensory

processing (e.g., postcentral areas), motor planning and

initiation (e.g., the supplementary motor and premotor

areas), and motor execution (the premotor and motor areas).

In the area of adult voluntary movement, Warrington and

McCarthy (1990) explored simple voluntary movements,

concluding that bimanual coordination localized to the

supplementary motor area, imitating hand position localized

to left parietal regions, and imitation of unfamiliar motor

sequences localized to left parietal and left frontal


Heilman and Rothi (1993) conceptualize the various

forms of apraxia in adults as resulting from different

lesions in the systems necessary for praxis. Sensory areas

frequently necessary for praxis include visual cortex (for

imitation) and Wernicke's area (for comprehension of

instructions). The second necessary step to praxis is

access to praxicons (programs of skilled movement), which

are postulated to exist in the angular gyrus or

supramarginal gyrus, usually of the left hemisphere. Next,

the premotor and supplementary motor areas initiate and

control the movement both directly and via the motor cortex.

The final major aspect of this schema is the corpus

callosum. Lesions in the corpus callosum can result in

disconnections between right and left hemisphere areas

necessary for movement (e.g., a right handed patient might

not be able to carry out commands with his left hand because


of a disconnection of the right motor areas from the left

praxicons). In addition, Heilman and Rothi (1993) note that

disruptions of the striatal-pallidal-thalamic-cortical

circuits can disrupt various steps of praxis.

For apraxia in adults, Heilman and Rothi (1993) have

noted that difficulties with coordinated movement can relate

to damage to premotor or motor cortex, to parietal regions

(spatial disorientation, poorly developed body schema,

possible location of motor engrams), to callosal regions

(disconnection of motor engrams from motor cortex), or to

the striatal-pallidal-thalamic-cortical circuits (which

provide input to cortical areas).

Cerebral Lateralization and Visuomotor Skills

As evident from the previous discussion, right

hemisphere lesions in adults often result in greater

impairment in visuoperceptual and visuospatial skills than

left hemisphere lesions (Bauer, 1993; Benton & Tranel, 1993;

Heilman & Rothi, 1993; Tranel, 1992). Lateralization of

function has been recognized as an important aspect of

cortical organization for many years (Kinsbourne, 1989; Kolb

& Whishaw, 1990). In brief, for right-handers the left

hemisphere is generally dominant for language functions

(e.g., comprehending speech, speaking, reading, writing, and

verbal memory), for complex voluntary movement (e.g.,

writing or the praxicons mentioned above), and for

arithmetic (Heilman & Rothi, 1993; Kinsbourne, 1989; Kolb &


Whishaw, 1990; Tranel, 1992). The right hemisphere tends to

be dominant in interpreting visual information other than

written information, in visuospatial functions, in face

recognition, in comprehending sounds other than speech, in

emotional processing, and in nonverbal memory (Bauer, 1993;

Benton & Tranel, 1993; Heilman, Bowers, & Valenstein, 1993;

Kinsbourne, 1989; Kolb & Whishaw, 1990; Ornstein, Johnstone,

Herron, & Swencionis, 1980; Tranel, 1992; Young, 1983).

Positron emission tomography indicates that the right

hemisphere is more active in object identification and

visuospatial tasks, with presumed callosal connections to

and from the left hemisphere (McIntosh et al., 1994).

Although the two hemispheres work in tandem, right

hemisphere functions tend to relate more directly to

visuoperceptual and visuospatial skills and, to the degree

that visuoperceptual skills are a prerequisite for

visuomotor performance, to perceptual-motor skills. Despite

the right hemisphere's dominance in this area, the left

hemisphere also plays an important role in visuomotor tasks

because (1) language functions are likely to relate to

visuoperceptual and visuomotor skills indirectly, (2) the

motor dominance of the left hemisphere (for right-handers)

means that the left hemisphere motor areas are crucial to

movement, and (3) the praxicons for complex, sequenced

movement are believed to be stored in the left hemisphere

(Heilman & Rothi, 1993; Miller & Rohr, 1980; Ornstein et

al., 1980).


Although the typical cerebral lateralization in adults

has been widely studied, it is less clear precisely how and

when such lateralization occurs. Much evidence suggests

that cerebral lateralization is innate and present from

birth (Kinsbourne, 1989; Young, 1983). For example,

Wernicke's area (crucial to language processing) in the

newborn infant is larger on the left than on the right side

(Kinsbourne & Hiscock, 1981). In addition, left-sided brain

injury prior to 1 year of age results in lower verbal and

performance abilities, whereas similar injury to the right

hemisphere results only in lower performance abilities,

suggesting that the left hemisphere is specialized prior to

one year of age (Kinsbourne, 1989; Woods, 1980).

Furthermore, as in adults, infants demonstrate greater left-

hemisphere excitation to speech and greater right-hemisphere

excitation to non-language sounds (Kinsbourne, 1989;

Kinsbourne & Hiscock, 1981). Evidence also suggests that a

motor preference to use the right hand is present for most

infants in the first months after birth (Bryden & Saxby,

1986; Kinsbourne, 1989; Kinsbourne & Hiscock, 1981).

However, other evidence suggests that cerebral

lateralization is not "set in stone" until later in

childhood. For example, the right-handed preference is not

as common in infants as it is in adults, suggesting that

further development occurs to solidify left hemisphere

dominance (Bryden & Saxby, 1986; Kinsbourne, 1989). Indeed,

a number of studies have not demonstrated a consistent


right-handed preference until around age 8 (Bryden & Saxby,

1986; Kinsbourne, 1989). Similarly, handedness and other

measures of lateralization, such as eye preference, foot

preference, and auditory preference become more highly

correlated with one another in late childhood, suggesting

that these preferences had not crystallized at earlier ages

(Bryden & Saxby, 1986; Kinsbourne, 1989; Kinsbourne &

Hiscock, 1981). Thus, it appears that a certain amount of

cerebral lateralization is present at birth, but that

lateralization continues to develop throughout at least the

first ten years of life.

Learning Disabilities: An Overview

Learning disabilities comprise a final set of

neuropsychological disorders with particular relevance to

visual and visuomotor skills (and with particular relevance

to this project, since learning disabled children comprise

one of the three samples). A diagnosis of learning

disability means that, in the absence of discernible

cognitive, sensory, neurological, emotional, or

environmental causes, a child experiences significant

difficulty learning a particular skill (Rourke, 1989; Spreen

et al., 1984; Taylor, 1988a).

Despite the fact that children and adults with learning

disabilities do not display consistent patterns of

neuroanatomical abnormalities that are clearly attributable

to a learning disability, the type of difficulties they


experience suggests that a currently undetectable

neurological abnormality does exist (Rourke, 1975; Spreen et

al., 1984; Taylor, 1988a, 1988b). Indeed, a small but

growing body of literature is detecting morphological

abnormalities in the brains of some developmental dyslexics

(Dool, Stelmack, & Rourke, 1993; Hynd & Semrud-Clikeman,

1989a, 1989b; Hynd, Demrud-Clikeman, Lorys, Novey, &

Eliopulos, 1990; Leonard et al., 1993). Specific

abnormalities that have been discerned in some but not all

cases include atypical patterns of hemispheric symmetry,

abnormalities in brain electrical activity, focal lesions in

left frontal and perisylvian regions, unusually small

insular regions, extra gyri in the parietal operculum, and

multiple Heschl's gyri (Dool, Stelmack, & Rourke, 1993; Hynd

& Semrud-Clikeman, 1989a, 1989b; Hynd et al., 1990; Leonard

et al., 1993). However, as of yet, no clear pattern has

emerged via which learning disabilities can clearly be

associated directly to brain abnormalities (e.g., control

subjects without learning disabilities often have

morphological abnormalities as well) (Hynd et al., 1990;

Leonard et al., 1993).

Although Morris (1989), Rourke (1989), and Taylor

(1988a) have made excellent cases for improving the

categorization of learning disabilities, no classification

system has emerged that maximizes descriptiveness,

reliability, and validity. The traditional subtypes of

learning disabilities include dyslexia (difficulties


reading), dysgraphia (difficulties writing), and dyscalculia

(difficulties with arithmetic) (Spreen et al., 1984).

Although in theory these subtypes seem straightforward, in

practice the variability of difficulties experienced by

learning disabled children defies such a discrete

classification system (Morris, 1989; Taylor, 1988a).

In addition to the heterogeneity caused by the many

different deficits potentially causing learning

disabilities, Taylor (1988a) has identified numerous areas

in which learning disabled children will differ, producing

even greater heterogeneity. Such areas include

environmental factors (the family and the school's reaction

to, treatment of, and expectations of the child), biological

factors, cognitive abilities, temperament, self-concept,

attitude towards school and learning, comorbid disorders

(attention-deficit disorder has a particularly high

comorbidity rate with learning disabilities), and local

standards for defining learning disability.

Gender differences also account for some of the

heterogeneity (Vogel, 1990). Many more males, in estimated

ratios ranging from 2:1 to 5:1, are diagnosed as learning

disabled; however, females tend to evidence greater deficits

when diagnosed (Taylor, 1988a; Vogel, 1990). In particular,

female learning disabled children typically display lower

Full-Scale, Verbal, and Performance IQs, and have greater

deficits in academic achievement. The only areas in which

female learning disabled children typically display higher


scores than males are in the Coding subtest of the WISC-R

and in written expressive language skills (Vogel, 1990).

Vogel (1990) hypothesized several possible reasons for

these differences, including sample bias (e.g., male

children are more likely to be identified as LD because they

are likely to display more disruptive behavior than

females). Indeed, Shaywitz, Shaywitz, Fletcher, and Escobar

(1990) examined differences between children identified as

learning disabled by the school system and children

identified through their research protocol, concluding that

the proportion of learning disabled males to learning

disabled females is approximately 1.2 males per 1 female.

They found that increased behavior problems in males

resulted in greater teacher identification of learning

disabilities in males. Furthermore, they note that the

traditional view that many more males than females are

learning disabled results in underdiagnosis of learning

disabled females by physicians.


It is likely that dyslexia, considered to be the most

common learning disability, is extremely variable in terms

of the specific deficit which causes the reading difficulty

(Taylor, 1988a; Spreen et al., 1984; Vellutino, Scanlon, &

Tanzman, 1994). Reading consists of a wide range of

specific functions, and deficits affecting any of these

functions would result in reading problems (Friedman, Ween,


& Albert, 1993; Taylor, 1988a). Traditionally, the first

step in reading is visual analysis, resulting in

identification of the visual pattern of the letters or the

word. Once visual analysis occurs, reading proceeds in two

different ways (Friedman et al., 1993; Taylor, 1988a). The

word can be matched to the visual image of the word in

memory (lexical or orthographic reading), or can be "sounded

out" and then identified in the same way that speech is

interpreted (phonological reading) (Friedman et al., 1993;

Taylor, 1988a; Vellutino et al., 1994).

Although this schema for reading is conceptually

appealing, reading is much more complex (Vellutino et al.

1994). General language skills, especially auditory

comprehension, have significant impact on reading ability,

and semantics as well as syntactic knowledge greatly

facilitate reading (Vellutino et al., 1994).

Bradley (1983) suggests that lexical reading occurs

first developmentally, and that phonological reading

develops later in childhood (beginning around age 7). In

support of this hypothesis, Bradley (1983) demonstrates that

children experience a period when they can use phonological

cues for spelling, but not for reading. In addition to

lexical and phonological reading, Morris (1991) indicated

that a third and final developmental stage occurs when the

child becomes able to analyze sets of letters which form a

meaningful subset of a word (e.g., prefixes, roots,



Difficulties can occur in any one of these steps

(Friedman et al., 1993; Taylor, 1988a). For example,

problems in visual discrimination can affect the ability to

recognize letters and to distinguish between letters (Spreen

et al., 1984). Difficulties in sequencing or in spatial

perception can also result in visual discrimination errors,

such as mistaking "two" for "tow" (Spreen et al., 1984;

Taylor, 1988a). Indeed, Bradley's (1983) research indicated

that spatial difficulties are present in children who are 18

months behind in reading. However, Vellutino et al. (1994)

conclude that the incidence of generalized language

difficulties is far greater in dyslexia than the incidence

of visual or visuospatial defects.

As further examples, problems with orthographic memory

may make it difficult to recognize letters and words.

Bradley (1983) demonstrated that children who had

difficulties on a task involving visual memory for letter

sequences, when compared to children who attained perfect

scores on this task, demonstrated more errors on reading

phonetically irregular words. Lexical reading can also be

impeded if the lexicon is damaged, or is disconnected from

the anatomical structures employed to analyze the written

word (Friedman et al., 1993).

Deficits in auditory processing or phonological

processing may cause difficulties in translating information

from written to oral form, and vice versa (Spreen et al.,

1984; Taylor, 1988a; Vellutino et al., 1994). In support of


this possibility, Bradley (1983) found that children who

were 18 months behind their expected reading level could

perform purely visual and purely auditory tasks, but

experienced difficulty translating from visual to auditory

information and vice versa. Similarly, problems in auditory

organization can make phonological processing of language

more difficult (Vellutino et al., 1994). In fact, Bradley

(1983) found that children who were 18 months behind their

expected reading level experienced difficulty on an auditory

organization task.

Examination of adults who have lost reading ability

following injury also contributes important information to

the attempt to understand dyslexia (Friedman et al., 1993).

In adults, damage to the occipital lobes, the corpus

callosum, or the angular gyrus results in pure alexia,

defined as severely impaired comprehension of written

language which is frequently accompanied by difficulty

copying and by written acalculia (Friedman et al., 1993).

Alexias have also been reported to result from lesions in

precentral cortex, postcentral cortex, the supramarginal

gyrus, and the inferior, medial, and superior temporal gyri

(Friedman et al., 1993).


Dysgraphia may relate either to general difficulties

with fine motor skills (such difficulties were discussed

with apraxia and developmental dyspraxia, so will not be


reexamined here), or to specific difficulties with writing

language (Roeltgen, 1993; Spreen et al., 1984). Writing,

like reading, has two primary routes, lexical and

phonological (Roeltgen, 1993). Based on adults with lexical

agraphia (inability to write irregular words), the region

most important to lexical writing is the junction of the

posterior angular gyrus and the parieto-occipital lobule,

extending into the white matter (Roeltgen, 1993). Lexical

agraphia has also resulted from lesions in the right

parietal lobe, the left posterior temporal lobe, the left

frontal lobe, and the left caudate nucleus (Roeltgen, 1993).

Phonological agraphia, or inability to write unfamiliar

words or nonwords, results primarily from damage to the

supramarginal gyrus or the insula medial to the

supramarginal gyrus (Roeltgen, 1993). Less commonly,

phonological agraphia has been linked to damage to the

caudate nucleus, internal capsule, and thalamus (Roeltgen,



Dyscalculia can take a number of forms, including

difficulties with verbalizing arithmetic functions, with

reading and/or writing arithmetic functions, with

comprehending the nature of the arithmetic functions, with

placing the components of a written arithmetic task in

appropriate spatial relations to one another, with

sequencing the components of an arithmetic function,


shifting between arithmetic response sets, judgement and

planning, or with using the many memory functions necessary

for arithmetic (Levin, Goldstein, & Spiers, 1993; Spreen et

al., 1984; Rourke, 1993). Rourke (1993) has reported

arithmetic disabilities to be associated features of both

nonverbal learning disabilities (discussed below) and

language disabilities. Reading, writing, spatial functions,

and sequencing functions were discussed above, and therefore

will not be reviewed again.

Nonverbal Learning Disabilities

Rourke (1989, 1994) has been the primary proponent of a

nonverbal learning disability syndrome. He has identified

associated deficits in the following areas: tactile and

visual perception, complex psychomotor skills, attention,

exploratory behavior, tactile and visual memory, concept

formation, problem-solving, oral-motor praxis, prosody,

phonology-semantics, and verbal comprehension. These

difficulties occur in the context of strengths in auditory

perception, simple motor skills, rote memory, auditory and

verbal attention and memory, phonology, verbal reception and

comprehension, and verbal output. As a result of this

pattern of strengths and weaknesses, Rourke (1989, 1994)

indicates that these children display academic strengths in

graphomotor ability (only after age 8), word decoding,

spelling, and verbatim memory, and weaknesses in early

graphomotor ability, reading comprehension, arithmetic, and


science. Rourke (1989, 1994) also considers a key aspect of

nonverbal learning disability to be social-emotional

problems, including poor social skills, emotional

difficulties, and abnormal activity level (hyperactivity at

age 4-5 and hypoactivity in later childhood).

Little (1993), in her review of the literature,

concluded that the empirical evidence for a nonverbal

learning disability subtype was robust. However, she noted

that insufficient evidence existed to conclude that social-

emotional problems are unique to the nonverbal learning

disability subtype. Indeed, Voeller (1991) describes

social-emotional problems as the core feature of social-

emotional learning disability, which in her sample was not

consistently associated with the other features reported by

Rourke (1989, 1994).

Although Rourke (1989, 1994) hypothesizes that white

matter lesions result in the nonverbal learning disability,

compelling arguments have been made by other researchers for

right hemisphere dysfunction to be the key area of

dysfunction in social-emotional difficulties (Voeller,


Assessment of Visuomotor Skills

Developmental information, visual perceptual problems,

difficulties with praxis, and evaluation of learning

disabilities provide an overview of the complexity of visual

and visuomotor processes. As described earlier, these


visual and visuomotor skills are a crucial component of

intellectual development (Piaget, 1927/1977; 1936/1977;

1937/1977; Piaget & Inhelder, 1969; Williams, 1983).

Therefore, it is likely that assessing such abilities in

childhood would provide a valuable tool in the evaluation of

children's neuropsychological functioning. However, few

such assessment instruments have received extensive

empirical examination of their usefulness in

neuropsychological assessment of children. Rather, tests

which have been widely used with adults have been extended

to children. Further data is necessary to demonstrate

empirically that such adult tests can be effectively

generalized to children. The Hooper Visual Organization

Test (HVOT) (Hooper, 1958) and the Rey-Osterrieth Complex

Figure Test (ROCFT) (Osterrieth, 1944; Rey, 1941), which

have been used extensively with adults in the assessment of

visuospatial and visuomotor skills, have emerged as

potentially having some utility in the evaluation of

children (Lezak, 1983; Spreen & Strauss, 1991).

The Hooper Visual Organization Test (HVOT)

The HVOT consists of a series of 30 items in which the

patient is presented with line drawings of objects which

have been divided into two or more parts and randomly

arranged on the page (Hooper, 1958). The patient must

mentally rotate and integrate the pieces of the object in

order to identify the object. The HVOT results in two


different kinds of information: (1) the score, or the total

number correct, and (2) the number of isolate responses

(Hooper Visual Organization Test manual, 1983; Lezak, 1983;

Spreen & Strauss, 1991). In addition, three other sources

of information from the HVOT are perseverations, bizarre

responses, and neologisms (Hooper Visual Organization Test

manual, 1983; Lezak, 1983; Spreen & Strauss, 1991. Isolate

responses are errors which occur when one of the pieces of

the object resembles a different object (e.g., the back

portion of a mouse, including the tail, looks like a pipe)

(Hooper, 1958; Hooper Visual Organization Test manual, 1983;

Lezak, 1983). When the patient names this piece of the

object rather than integrating the various pieces of the

object, this is considered an isolate response (Hooper,

1958; Hooper Visual Organization Test manual, 1983; Lezak,

1983). Although intended to measure visuospatial and object

recognition ability, the HVOT also requires naming; thus,

naming is a confounding factor in interpretation of the HVOT

(Spreen and Strauss, 1991).

HVOT isolate response errors

Walker (1956), in a study of 38 neuropsychiatric

patients (no diagnoses were specified), found that the

number of isolate response errors was significantly

different for neuropsychiatric patients with definite or

suspected cortical involvement than for neuropsychiatric

patients with no cortical involvement. However, he noted

that the low cutoff (1 or fewer isolate response errors to


be considered within the normal range) greatly increased the

likelihood of false positive errors (Walker, 1956).

In a sample of 33 schizophrenics, Walker (1957) found

that those schizophrenics who had 1 or no isolate response

errors on the HVOT were more likely to be discharged within

one year after administration of the HVOT than those

schizophrenics with 2 or more isolate response errors on the

HVOT. These data suggest that the number of isolate

response errors on the HVOT, at least for schizophrenic

patients, is related to severity of impairment. These

studies by Walker (1956, 1957) are the only empirical

evaluation of isolate response errors in adults, and do

offer preliminary evidence that identifying such errors may

be of importance.

The HVOT's reliability

In terms of the HVOT total score, Hooper (1958) found

split-half correlations of .78 and .82 for samples of 166

college students and of 74 psychoneurotic patients (no

diagnoses were provided), respectively. Similarly, the

split-half correlations calculated by Gerson (1974) ranged

from .79 to .80 in his three groups of 16 patients with

various forms of organic impairment, 19 functionally

impaired (schizophrenic) patients, and 33 normal subjects.

The HVOT's relationship to demographic variables

Hooper (1958) reported that the HVOT generally is

unrelated to sex, education, intelligence, and age.

However, he did find that intelligence correlated with the


HVOT at .57 for individuals with an IQ between 30 to 80, and

that older people received lower scores on the HVOT,

presumably the result of normal aging processes (Hooper,

1958). Although Hooper's (1958) claims about the lack of

gender effects have sustained subsequent inquiry, his other

claims have been called into question by more recent


For example, the results of other researchers who

examined the relationship between the HVOT and IQ indicate

that Hooper (1958) was mistaken when he claimed that IQ was

not related to HVOT scores. For example, Wentworth-Rohr,

Mackintosh, and Fialkoff (1974), who employed samples of 200

psychiatric patients with unspecified diagnoses, 85 social

worker nuns, 51 parochial school principals, and 61 college

students, found correlations ranging from .31 to .50 between

the HVOT and IQ. Boyd (1981), in samples of 40 organically

impaired patients and 40 normal controls, and Tamkin and

Jacobsen (1984), in a sample of 211 psychiatric patients

with unspecified diagnoses, also demonstrated that the HVOT

was related to their estimates of IQ. Therefore, IQ scores

should clearly be taken into account when the implications

of HVOT scores are considered.

The data regarding the relationship between the HVOT

and age and education has been mixed. Wentworth-Rohr et al.

(1974) confirmed that, for adults and adolescents, HVOT

scores are unrelated to sex, education, and age. Although

Boyd's (1981) sample displayed a significant correlation


between the HVOT and age for a neurologically impaired

group, he concluded that the severity of illness, rather

than age itself, accounted for this difference. However,

the research of Tamkin and Jacobsen (1984) and Tamkin and

Hyer (1984) indicated that HVOT scores were related to age,

as well as to education, in a sample consisting of inpatient

psychiatric patients in a Veterans Administration Medical

Center. With no information regarding substance abuse

history or chronicity of psychiatric illness (both likely to

be more severe in the middle-aged to elderly population),

Tamkin and Hyer's generalizability to the normal population

or to a neurologically impaired population is questionable.

The association of age and education with HVOT scores is

complicated by Tamkin and Jacobsen's (1984) finding that IQ

correlated with both age and education, making it difficult

to determine the impact of age and education independent of

IQ. Thus, the research on the relationship between age and

education on HVOT scores remains inconclusive.

Nevertheless, in order to address the problem of age

and intelligence affecting HVOT scores, age- and education-

corrected scores based on a regression equation have been

developed for the most recent HVOT manual (The Hooper Visual

Organization Test Manual, 1983). Education was chosen as an

indirect measure of IQ which would be available without

testing (The Hooper Visual Organization Test Manual, 1983).

For adults over 25, these age- and education-corrected

scores would be expected to enhance the HVOT's usefulness;


however, this expectation has not been empirically


Criterion validity data for the HVOT

It is almost universally reported, regardless of either

definition of neurological impairment or sample size, that

mean HVOT scores in adults are significantly lower in

neurologically impaired patients than in non-neurologically

impaired patients (Boyd, 1981; Gerson, 1974; Hooper, 1952;

Tamkin & Kunce, 1985; Sterne, 1973; Wang, 1977). Walker's

(1956) sample of 38 neuropsychiatric patients is the only

study that did not display significant differences between

patients with suspected versus patients with no neurological

impairment. Because Walker's (1956) definition of

neurological impairment was particularly vague and the

preponderance of evidence indicates that the HVOT total

score does discriminate between neurologically impaired and

non-neurologically impaired patients, it is likely that

Walker's (1956) sample size and procedures for classifying

patients were inadequate to discern differences between

neurologically impaired and non-neurologically impaired


The earliest published research using the HVOT

demonstrated that the HVOT distinguished between

neurologically impaired versus non-neurologically impaired

adults (Hooper, 1952; Hooper, 1958). His data, which were

based on a total sample of 200 subjects, indicated that

adults with a variety of brain pathology (of inflammatory,


degenerative, toxic, tumorous, or traumatic etiology)

received significantly different scores on the HVOT than

schizophrenic, psychoneurotic, or normal adults (Hooper,

1952; Hooper, 1958). However, the criteria for these groups

appear to be poorly defined, and the specific data for the

groups are not available.

Gerson (1974) also found that the HVOT scores were

significantly different for his 16 organically impaired

adult patients (with brain disease of degenerative,

traumatic, or toxic origin) versus either his 19

functionally impaired patients (i.e., schizophrenics) or his

33 normal subjects. The schizophrenic and normal groups

were not significantly different. Even when IQ was

controlled through an analysis of covariance, the HVOT

continued to demonstrate a statistically significant

difference among the three groups.

Boyd (1981) examined HVOT data for 40 mildly to

moderately neurologically impaired adult patients (10 with

closed-head injury; 5 with seizure disorders; 4 with

congenital vascular malformations; 3 cases each of normal

pressure hydrocephalus, multiple sclerosis, penetrating head

injuries, and cerebral vascular accident; 2 cases each of

metastatic brain tumor, cortical atrophy, and extrinsic

brain tumor; and 1 case each of Huntington's chorea,

glioblastoma, and aneurysm) and 40 non-psychotic, non-

neurologically impaired patients. The two groups were

equivalent in age, education, and Peabody Picture Vocabulary


Test (PPVT) (Dunn, 1965) scores. The HVOT scores for the

two groups were significantly different, with mean scores of

23.2 (SD = 3.99) for the neurologically impaired patients

and 26.5 (SD = 2.47) for the non-neurologically impaired


Similarly, both Wang (1977) and Tamkin and Kunce (1985)

found that HVOT scores were significantly lower for

neurologically impaired versus normal adult subjects.

Sample sizes were 49 neurologically impaired (15 with left

hemisphere damage, 19 with right hemisphere damage, and 15

with bilateral damage; diagnoses are not provided) and 17

non-neurologically impaired subjects (with diagnoses such as

low back pain, osteoarthritis, peripheral neuropathy, or

spinal cord injury) for Wang's (1977) research, and a total

sample size of 66 patients with unspecified diagnoses for

Tamkin and Kunce's (1985) study.

Finally, Sterne (1973), in a total sample of 75 adult

subjects, demonstrated that the HVOT discriminated between

25 neurologically impaired patients with unspecified

diagnoses, 25 patients of indeterminate neurological status,

and 25 non-neurologically impaired patients. However, he

noted that the HVOT, when given in conjunction with other

tests, did not provide additional information for the

discrimination of organic versus non-organic patients.

Therefore, he indicated that, when sufficient data had been

collected, the HVOT was unnecessary for the determination of

neurological impairment.


The data from all of these studies indicate that the

HVOT is able to discriminate between groups of

neurologically impaired versus non-neurologically impaired

adult patients, although it is typically endorsed primarily

as a screening measure. Despite the HVOT's ability to

distinguish between groups of neurologically impaired versus

non-neurologically impaired patients, it has been quite

difficult to determine a cutoff score that would be useful

in classifying individual patients.

The utility of a cutoff score for the HVOT

Hooper (1958) found false positive rates of 3% in a

sample of 30 junior high school students, and of 6% in a

sample of 166 college students. In Eisenman and Coyle's

(1965) study of 23 female student nurses, the lowest score

obtained was 26.5, well above the cutoff of 20, and within

the 25 to 30 range identified as normal by Hooper (1958).

Further evaluation of the number of false positives and

false negatives in a psychiatric population consisting of 20

organic (including diagnoses of disseminated sclerosis,

meningitis, lesions, tumors, hemorrhages, Parkinson's

disease, and Wernicke's aphasia) and 95 non-organic patients

(15 schizophrenic, 13 manic or depressed, 15 neurotic, 10

behaviorally disordered, 9 alcoholic or epileptic, 6

mentally retarded, 13 antisocial, and 14 court-ordered)

indicated a false positive rate of 25% and a false negative

rate of 30% (Love, 1970). However, when those non-organic

patients with an IQ under 75 were excluded, the false


positive rate dropped to 14.5% (Love, 1970). Love (1970)

concluded that the HVOT is a useful screening device, but

should not be employed alone to discriminate between

neurologically impaired and normal people. In particular,

his data provided further evidence that IQ scores should be

taken into account when examining the significance of HVOT


Other researchers have encountered similar difficulties

in determining a cutoff score that would substantially

diminish false positive and false negative rates. When

evaluating the usefulness of cutoff scores, Gerson (1974)

determined that the false negative rate was 0%, but that the

false positive rate was 51%. Similarly, the false negative

rate was 17% and the false positive rate was 40% for Wang's

(1977) subjects. Boyd (1981) found that, using Hooper's

cutoff of 20, the false positive rate was 2.5%, but the

false negative rate was 85%. All of these data indicate

that the HVOT has a particularly high rate of false

positives, as well as a fair number of false negatives.

Boyd (1981) recommended using a cutoff of 25, which for his

sample resulted in a false positive rate of 20% and a false

negative rate of 32.5%. The high false positive rate

appears to result from the fact that the cutoff is fairly

close to the total possible score, and therefore results in

a ceiling effect so that scores are normal unless the damage

is fairly severe or affects areas crucial to the task. All

of the above data do indicate that the HVOT cutoff scores


are quite problematic, and that the HVOT cutoff score is

frequently unable to provide a conclusive indication of the

presence or absence of neurological impairment.

The HVOT's relationship to location or type of lesion

Despite the fact that the HVOT's most valuable use may

be in detecting specific dysfunction rather than identifying

whether or not a patient has neurological impairment, the

research regarding localization is relatively sparse (Spreen

& Strauss, 1991). In terms of the HVOT's ability to

distinguish between patients based on the site of

neurological impairment, Wang (1977) found that HVOT scores

were statistically equivalent, regardless of site of

impairment (right, left, or bilateral). Nevertheless, he

noted a trend for patients with right hemisphere deficits to

obtain lower scores. The HVOT mean scores were 12.21 for

patients with right hemisphere deficits, 13.93 for patients

with left hemisphere deficits, 10.53 for patients with

bilateral deficits, and 22.35 for normal subjects. Boyd

(1981) also determined that no statistically significant

differences existed within his neurologically impaired group

according to laterality of lesion. However, Tamkin and Hyer

(1984) provided evidence that right hemisphere dysfunction

is associated with decline in HVOT scores in the elderly.

Spreen and Strauss (1991) note that Farver and Farver's

(1982) results were also consistent with a possible

association between right hemisphere dysfunction and low

HVOT scores.


Boyd (1981) found that the HVOT did not relate to the

acuteness of the illness. However, as would be expected,

significant differences occurred on the HVOT according to

type of lesion, with "major tissue-destructive lesions

(intrinsic brain tumor, cerebral vascular accident)" (p. 18)

resulting in lower scores than diseases such as "congenital

vascular malformation, multiple sclerosis, [and] normal

pressure hydrocephalus" (Boyd, 1981, p. 18).

Concluding remarks on the adult HVOT literature

Thus, the data from adult studies with the HVOT support

several conclusions. The HVOT, despite Hooper's (1958)

claim to the contrary, is related to IQ. In addition, any

use of the HVOT cutoff scores is likely to result in a high

number of false negatives and false positives. Finally,

research consistently demonstrates that the HVOT scores are

significantly lower for neurologically impaired versus non-

neurologically impaired adult patients.

Use of the HVOT with children

In the first study involving children (40 7 to 11 year

olds, 20 of whom had undefined neurological impairments and

20 of whom were normal), Jackson and Culbertson (1977) found

that, using a HVOT cutoff of 21 which eliminated false

positives, the false negative rate was 60%. Unfortunately,

they present insufficient data to evaluate what cutoff would

strike a better balance between false positive and false

negative errors. Because later research demonstrated a

substantial increase in scores for children in this age


range (Kirk, 1992; Seidel, 1994), Jackson and Culbertson's

wide age range probably decreased their ability to determine

a cutoff score. Despite these problems in identifying a

useful cutoff score, the HVOT scores were significantly

different for neurologically impaired versus normal


Hilgert and Treloar (1985), employing a sample of 54

elementary school students with a mean IQ of 90.45 who were

referred for psychological evaluation by the school, found

that the mean HVOT score was 20.70. The HVOT scores

correlated significantly (.35) with Performance IQ, but not

with Verbal or Full-Scale IQ. Hilgert and Treloar (1985)

also concluded that the HVOT scores did not vary according

to sex. However, they did determine that age accounted for

a large proportion of the variability on the HVOT. There-

fore, Hilgert and Treloar (1985) concluded that age and

Performance IQ, but not sex, are related to HVOT performance

in children.

Kirk (1992) evaluated HVOT scores in 448 male and

female children aged 5 to 13 with normal confrontation

naming ability. Children were excluded if their scores on

either a confrontation naming task or on the HVOT were more

than 2.5 standard deviations from the mean for their age.

Both age and gender significantly affected performance on

the HVOT. Specifically, male children obtained higher

scores than female children at all age levels. The male

children scored at adult levels (25) by age 12, but the


female children had not achieved adult scores by age 13. In

Kirk's sample, older children performed better than younger

children, with significant differences for both sexes

between the younger ages (5-6) and the older ages (11-13).

In addition, younger children displayed more isolate

response errors than older children.

Most recently, Seidel (1994) examined the HVOT in 211

randomly selected normal subjects aged 5 through 11 years

and 49 children aged 6 through 11 from a clinical sample

including children with spina bifida, brain tumors, head

injuries, seizure disorders, and severe learning disorders.

Inclusion and exclusion criteria for the clinical sample

were unclear. His results replicated Kirk's (1992) normal

sample's scores, including the pattern of improved

performance with increasing age. Like Kirk (1992), he found

that the youngest children (5-, 6-, and 7-year-olds)

performed significantly worse than the oldest children (10-

and 11-year-olds), who approached adult scores. However,

unlike Kirk (1992), Seidel (1994) did not find significant

differences according to sex.

Seidel's (1994) internal consistency for the HVOT among

his normal sample was .723 for a Spearman-Brown corrected

odd-even correlation and was .716 for a Cronbach's alpha.

These were not different to a statistically significant

degree from the internal consistency of .78 and .82 reported

by Hooper (1958) (Seidel, 1994). In support of construct

validity, Seidel (1994) determined that the HVOT in his


clinical sample correlated significantly (.624) with the

Perceptual Organization Deviation Quotient of the WISC-R,

but did not correlate significantly with the Verbal

Comprehension Deviation Quotient (r=.196) or the Freedom

from Distractibility Deviation Quotient (r=.211). In terms

of individual subtests of the WISC-R, the HVOT correlated

significantly with Block Design (.625), Picture Arrangement

(.492), Object Assembly (.488), Picture Completion (.459),

and Vocabulary (.312). Along similar lines, Seidel's (1994)

principal components analysis for his clinical sample

indicated that the HVOT is related to visuospatial and

visuomotor skills, and not to verbal information processing

abilities or to reading and writing skills. Finally, Seidel

(1994) found that, with IQ controlled by matching a subset

of his normal sample with the IQs of his clinical sample,

his clinical sample nonetheless scored significantly lower

than his normal sample on the HVOT.

Concluding remarks on the HVOT literature

In general, the data on the HVOT indicates that it can

be a valuable measure for adults when combined with other

assessment instruments. Although little research has been

done with children, the preliminary normative data and

evidence of reliability and validity is promising. Indeed,

the HVOT may be more useful with children than with adults.

Specifically, the HVOT has a ceiling effect, which results

in normal scores for many neurologically impaired adults.

Because children in the age ranges studied thus far have not


generally reached the ceiling which adults encounter, the

HVOT may be of particular benefit in assessing children.

Furthermore, perceptual skills are not as fully developed in

children as in adults, and are more highly related to

cognitive skills and academic achievement in young children

than in adults (Melamed & Melamed, 1985; Williams, 1983).

Therefore, tests like the HVOT which examine visual skills

may provide information that is more relevant to the

neuropsychological functioning of children than of adults.

Thus, it is important to continue to empirically investigate

the usefulness of the HVOT in children.

The Rev-Osterrieth Complex Figure Test (ROCFT)

The Rey-Osterrieth Complex Figure Test (ROCFT)

(Osterrieth, 1944; Rey, 1941) consists of a complex drawing

which the patient must copy (Lezak, 1983). The drawing is a

rectangular shape, with a large triangle at one end and with

a number of details both within and outside the drawing

(Lezak, 1983; Rey, 1941). The examiner provides the patient

with a new colored pencil each time a section of the drawing

is completed, so that the patient's approach style can be

evaluated (Lezak, 1983; Spreen & Strauss, 1991). The task

requires the patient to correctly perceive the stimulus and

to use fine motor skills to reproduce the drawing. Denckla

(1994) noted that the ROCFT is an excellent measure of

executive function because of its emphasis on motor planning

and organizational ability. The ROCFT results in two


primary scores: one for the number of features either

correctly copied or approximated, and one for the approach

style or organizational quality (Lezak, 1983; Spreen &

Strauss, 1991).

The ROCFT total score

Typically, the total score, or the number of features

correctly reproduced, is derived by examining eighteen

primary details of the figure (Lezak, 1983; Osterrieth,

1944; Spreen & Strauss, 1991). For each of these details,

the patient can receive two points if the detail is both

correct and properly placed, one point if either its

accuracy or its placement (but not both) are correct, and a

half point if it is poorly reproduced and incorrectly placed

(Lezak, 1983; Osterrieth, 1944; Spreen & Strauss, 1991).

Therefore, the maximum score possible is 36.

Despite the fact that Osterrieth's (1944) scoring

criteria are poorly operationalized, Carr and Lincoln (1988)

obtained an inter-rater reliability of .99 for forty ROCFTs

scored by two raters. Similarly, Karapetsas and Kantas

(1991) found 92% agreement between two scorers of 80 ROCFTs

completed by children. However, neither Carr and Lincoln

(1988) nor Karapetsas and Kantas (1991) describe their

training method for these raters, and it is likely that they

were trained together and developed identical scoring

criteria. As some researchers, such as Bennett-Levy (1984),

have noted, scorers with various training backgrounds may

frequently employ quite different scoring criteria for the


ROCFT. Consequently, more specific criteria for scoring

accuracy and placement have been developed (Bennett-Levy,

1984; Mings, 1987). Using such operationalized standards,

Bennett-Levy (1984) found inter-rater reliability of .96 for

25 drawings. Unfortunately, Mings (1987) does not report

inter-rater reliabilities; however, it is likely, given the

specificity of his criteria, that inter-rater reliability

would be relatively high.

Organizational scoring for the ROCFT

How patients organize their drawing of the ROCFT is

also an important indicator of brain function. For example,

Osterrieth (1944) reported that 83 percent of his normal

adult sample organized their drawing using some

organizational structure (reported in Lezak, 1983). Both

Visser (1973) and Binder (1982) found that lack of

organization is frequently indicative of brain damage

(reported in Lezak, 1983). Similarly, research has

indicated that subjects' approach to the drawing is

significantly correlated with the total score for the

drawing (Bennett-Levy, 1984; Waber & Holmes, 1986).

To score approach style, Osterreith (1944) identified

seven approaches to copying the figure: (1) beginning with

the rectangle and adding the details, (2) beginning with a

detail, completing the rectangle, then finishing the

details, (3) beginning with the outline of the figure, then

filling in details, (4) placing details without any

organizing structure, (5) drawing details with no


organization whatsoever, (6) drawing a different object, and

(7) scribbling a figure which is unrecognizable (reported in

Lezak, 1983).

Because Osterrieth's (1944) approach styles are

qualitative, a number of researchers and clinicians have

2| developed scoring criteria which operationalize specific

standards for determining approach style. For example,

Bennett-Levy (1984) developed a strategy score, based on two

subscores: one subscore based on the amount of continuation

(i.e., when a line is drawn continuously, rather than in

separate pieces), and one subscore for the amount of

symmetry (i.e., when symmetrical aspects of the drawing are

drawn sequentially). For a similar purpose, Waber and

Holmes (1985) have devised a system based on specific

details of the drawing to distinguish five levels of

increasing organization. These levels of organization

correlate highly (r = .82) with clinical impressions of the

drawing, and have high inter-rater reliability (.88 to .94)

(Waber & Holmes, 1985, 1986).

The ROCFT's relationship to demographic variables

The effects of a number of demographic variables on

adults' performance on the ROCFT have been examined. In

terms of age, King (1981), in a sample of 114 subjects with

neurological impairment (no diagnoses specified) and 71

normal control subjects, found that subjects over 60 years

of age, regardless of whether they had neurological disease,

displayed poorer performances on the ROCFT total score than


subjects under 60. Similarly, Bennett-Levy (1984)

demonstrated in a sample of 107 normal subjects that age was

significantly correlated with the ROCFT total score.

Ardila, Rosselli, and Rosas (1989), in a sample of 200

normal right-handed subjects, concluded that age affected

ROCFT scores only for poorly educated subjects, suggesting

that an interaction effect may exist for age and educational


Intelligence and educational level have also been

associated with the ROCFT (Spreen & Strauss, 1991). King

(1981) and Bennett-Levy (1984) both found that IQ was

significantly correlated with adults' ROCFT performance, and

Spreen and Strauss (1991) report correlations between the

ROCFT and IQ ranging from .23 to .47. Furthermore, Ardila

et al.'s (1989) results indicated that educational level

significantly affected the ROCFT.

King (1981) found that adult subjects' sex did not

significantly affect scores on the ROCFT. However, Bennett-

Levy (1984) and Ardila et al. (1989) found that, for poorly

educated subjects, male subjects' copies of the ROCFT were

significantly better organized than females' drawings.

Similarly, Casey, Winner, Hurwitz, and DaSilva (1991) found

that, for a sample of 244 college students, subjects' sex

did not significantly affect ROCFT scores. These results

indicate that, as for age, there may be an interaction

effect for subjects' sex and educational level.


Validity data for the ROCFT

In terms of the validity of the ROCFT, King (1981)

found that the copy score for adults' ROCFT correlated

significantly with the Wechsler Memory Scale (Wechsler,

1945) visual recall, suggesting that both assess similar

visuoconstructive skills. In support of criterion validity,

the ROCFT total score was significantly better for a control

group than for patients with brain injuries (King, 1981).

The ROCFT's relationship to location of lesion

Visser (1973) and Binder (1982) found that adult

patients with right hemisphere lesions tended to have more

details missing, whereas patients with left hemisphere

lesions tended to have more fragmented details (i.e.,

details which are not drawn as a unit) (reported in Lezak,

1983). However, King (1981), in his study of 114 patients

with brain dysfunction and 71 control subjects, demonstrated

that neither laterality of lesion nor type of lesion

affected the ROCFT total score. He did not analyze approach

style or types of errors (King, 1981), raising the possi-

bility that type or location of brain dysfunction may affect

a patient's approach to the drawing rather than the total

score obtained.

For example, Messerli, Seron, and Tissot (1979), in

their study of adult patients with frontal lobe lesions,

found a number of frequent errors (reported in Lezak, 1983).

In order of descending frequency, these were perseverations

(either resulting from a disorganized approach, or from


drawing excessive elements of a set), transformations of a

detail into a more familiar figure (e.g., a circle with

three dots is drawn as a face), and omissions (reported in

Lezak, 1983). Pillon (1981) demonstrated that patients with

frontal lesions tended to have difficulty planning their

approach to copying the figure, whereas patients with

parietal-occipital lobe lesions typically had trouble

spatially organizing the figure (reported in Lezak, 1983).

Concluding remarks on the adult ROCFT literature

Thus, the total score, the organizational style, and

the types of errors have provided valuable information in

adult neuropsychological assessment. Specifically, the more

details included and the better the approach style, the less

the indication of brain dysfunction. In adults, as would be

expected in a complex task, IQ, education, and age appear to

affect ROCFT performance, indicating that such factors

should be taken into account (Spreen and Strauss, 1991).

The primary areas of difficulty with ROCFT literature are

scoring issues. Namely, the original scoring procedures

were not operationally defined, prompting a number of

researchers to design their own scoring systems. This is

particularly problematic since no research has been

conducted to discern which of these scoring systems is most

clinically valuable.

Use of the ROCFT with children

In a sample of 454 normal children aged 5 to 14, Waber

and Holmes (1985) found that children's copies of the ROCFT


tend to increase in organization with age. More speci-

fically, the youngest children tended to organize their

drawings around only the vertical axis, older children

typically organized their drawings around both the vertical

and horizontal axes, and the oldest children tended to be

able to organize their drawings around the vertical,

horizontal, and diagonal axes (Waber & Holmes, 1985).

Unfortunately, Waber and Holmes (1985) do not describe at

what approximate ages such shifts in capacity typically

occur, except to note that children 13 years old and older

tend to have the most organized drawings. Despite this lack

of age-related information, the general pattern of changes

with age suggests an increasing ability to organize and

integrate a variety of information.

Age and handedness, but not sex, significantly affected

organization scores on children's ROCFT (Waber & Holmes,

1985). Karapetsas and Kantas (1991), in a sample of 420

right-handed 5 1/2 to 12 1/2 year old children in Crete,

obtained similar results for age effects on ROCFT total

scores. More specifically, 5 1/2 to 6 1/2 year old children

performed significantly worse than all older children on

ROCFT total scores, suggesting that a developmental shift

occurs at around age 6 to 7 (Karapetsas & Kantas, 1991). In

a sample of 514 left-handed children in Greece, Karapetsas

and Vlachos (1992) obtained similar results: 5 1/2 to 6 1/2

year olds scored worse than all age groups over 7 1/2, and 6


1/2 to 9 1/2 year olds had significantly lower scores than

all other age groups.

Although Waber & Holmes (1985) reported that sex does

not affect how children approach the ROCFT, Karapetsas and

Kantas (1991) and Karapetsas and Vlachos (1992) found that

male children received significantly lower total scores than

female children. Specifically, around age 8 1/2, right-

handed female children began obtaining significantly higher

scores on the ROCFT than did male children (Karapetsas &

Kantas, 1991). In left-handed children, females signi-

ficantly outscored males for the age range from 7 1/2 to 9

1/2, but were statistically equivalent in younger and older

age ranges (Karapetsas & Vlachos, 1992). The contrasting

results obtained by Waber and Holmes (1985) in Boston and

Karapetsas and colleagues (1991, 1992) in Greece present two

possibilities: (1) that, although a child's sex does not

affect organization of the ROCFT, female children are able

to accurately copy more details at certain ages; and (2)

that cultural differences between Greece and the North-

eastern United States account for the different findings.

In the only study comparing a normal sample with a

clinical sample, Waber and Bernstein (1994) found that the

organizational score (Waber & Holmes, 1985) was substan-

tially greater for control versus learning disabled

children. Waber and Bernstein (1994) noted that the

learning disabled children's performance was significant for

the lack of improvement with age. Because the normal


children's scores increased steadily with age, reliable

differences between the two groups clearly emerged around 9

years of age (Waber & Bernstein, 1994).

Concluding remarks on the ROCFT's use with children

The relationship between demographic characteristics

and the ROCFT appear to be different in children than in

adults. Specifically, age and sex relate more strongly to

ROCFT performance in children than in adults. Older

children, particularly older female children, perform better

on the ROCFT total score than younger children. This is in

contrast to unreplicated HVOT data, which indicated that

male children may perform better than the female children

(Kirk, 1992). This pattern suggests that male children may

be more skilled at mentally manipulating figures in space,

but female children display greater ability to reproduce

visual information through drawing. In any event, it is

clear that age and sex are important components to keep in

mind when employing the HVOT and the ROCFT in children.

Summary of the HVOT and ROCFT Research with Children

The HVOT and the ROCFT both display the potential for

being quite useful in the neuropsychological assessment of

children. Both the HVOT and the ROCFT assess visuospatial

and visuomotor skills which are crucial to children's intel-

lectual development. The HVOT and the ROCFT both display

significant improvements in scores as children age, sugges-

ting that the HVOT and the ROCFT may be able to detect


developmental changes in visuospatial and visuomotor tasks.

Furthermore, since children have not begun to attain the

adult ceiling level which is sometimes noted, the HVOT and

the ROCFT may be particularly able to detect subtle deficits

in children's visuospatial and visuomotor processes.

Problems with HVOT and ROCFT Literature with Children

Despite the promising results of the preliminary

research previously outlined, much more research is

necessary before forming conclusions about the utility of

the HVOT and the ROCFT in children. For example, only three

projects have evaluated either the HVOT or the ROCFT in

neurologically impaired or learning disabled children

(Seidel, 1994; Jackson & Culbertson, 1977; Waber &

Bernstein, 1994). These studies provide promising but

limited data for the use of the HVOT and the ROCFT with a

clinical population. The remainder of the research with the

HVOT and the ROCFT in children has focused upon normal

children (Hilgert & Treloar, 1985; Karapetsas & Kantas,

1991; Kirk, 1992; Waber & Holmes, 1985; Waber & Holmes,

1986). Such research has provided valuable normative data

(Karapetsas & Kantas, 1991; Kirk, 1992; Waber & Holmes,

1985), validity information (Hilgert & Treloar, 1985), and

scoring criteria (Mings, 1987; Waber & Holmes, 1985; Waber &

Holmes, 1986). These studies provide the necessary

foundations for continuing to explore the utility of the

HVOT and the ROCFT in neuropsychological assessment.


However, a number of questions must be answered before the

HVOT and the ROCFT can be employed with strong confidence in

a clinical population. For example, only Hilgert and

Treloar (1985) have examined the effect of IQ on children's

HVOT performance; no one has examined the relationship

between IQ and the ROCFT in children. In addition, only

Hilgert and Treloar (1985) and Seidel (1994) has compared

the HVOT with other measures of visual-perceptual and

visuomotor skills to examine construct validity; no one has

performed such correlations with the ROCFT. Furthermore,

the data on the HVOT's and the ROCFT's ability to

discriminate between normal and neurologically impaired

children has been fairly limited. This project attempts to

provide data that compensates for these limitations.

Rationale for this Project

This research project addressed (1) the criterion

validity of the HVOT and the ROCFT in terms of the ability

to detect developmental changes in middle childhood, (2) the

construct validity of the HVOT and the ROCFT by correlating

them with other measures of visuospatial and visuomotor

skills, and (3) the criterion validity of the HVOT and the

ROCFT by exploring whether any differences exist among

children with medically identified neurological impairment,

children with learning disabilities, and normal children.

Despite the methodological problems inherent in a

learning disabled sample (Durrant, 1994; Taylor, 1988a), the


addition of a learning disabled sample provided valuable

additional information. Learning disabled children

presumably have some neurological dysfunction that places

them at an intermediate level between the normal sample and

the neurologically impaired sample. Therefore, the learning

disabled sample furnished a more rigorous test of the HVOT

and ROCFT's ability to discriminate between normal and non-

normal samples. In addition, learning disabled children's

performance on the HVOT and the ROCFT was important to

document because neuropsychological testing plays such an

important role in identifying the specific difficulties

learning disabled children experience (Rourke, 1994; Taylor,

1988a, 1988b). Rourke (1994) and Taylor (1988b) have noted

the importance of thorough neuropsychological evaluation

(including visuoperceptual, visuospatial, and visuomotor

assessment) with learning disabled children in order to

assist in diagnostic issues and treatment planning.

The measures employed in this project were selected to

either provide important screening information, to assess

demographic information, or to evaluate visuoperceptual or

visuomotor skills. To screen for problems which might

exclude a child from study, both estimated IQ and naming

ability were assessed. IQ scores served 3 purposes: (1) to

screen out children who did not meet certain IQ criteria;

(2) to examine visuospatial and visuomotor skills while

statistically controlling for general intellectual ability;

and (3) to determine the association between IQ scores and


HVOT and ROCFT scores. Since naming ability is a

prerequisite for the HVOT, it was also necessary to screen

for naming deficits (Spreen & Strauss, 1991).

Other measures employed were intended to assess various

aspects of visuomotor skills. For example, the Picture

Completion (PC) subtest of the Wechsler Intelligence Scales

for Children or the Wechsler Preschool and Primary Scales of

Intelligence (Wechsler, 1974; Wechsler, 1989; Wechsler,

1991) requires visual skills similar to the HVOT and ROCFT:

recognition of the item (HVOT) and visual analysis of

details of the item (both HVOT and ROCFT). Similarly, the

Object Assembly (OA) subtest of the Wechsler Intelligence

Scales for Children is essentially similar to the HVOT,

except that Object Assembly requires manual rather than

mental manipulation of the parts of the object.

Additionally, Beery's (1989) Test of Visual-Motor

Integration (VMI) is a commonly used test of graphomotor

ability, and is therefore similar to the ROCFT. These tests

were employed to evaluate the construct validity of the HVOT

and the ROCFT.

Hypotheses for this Project

The specific hypotheses included the following:

(1) All of the following dependent variables would be

significantly associated with (a) age and (b) sex: (1) HVOT

total score, (2) the number of HVOT isolate response errors,

(3) the ROCFT total score, and (4) the ROCFT organizational



(2a) The HVOT total score would correlate signifi-

cantly with IQ, Picture Completion, and Object Assembly.

(2b) Because the HVOT isolate response score probably

relates to general abilities or executive functioning rather

than to visuoperceptual or visuospatial skills per se, the

HVOT isolate response score would correlate significantly

with IQ.

(2c) The ROCFT total score would correlate

significantly with IQ, Picture Completion, Object Assembly,

and the Test of Visual-Motor Integration.

(2d) The ROCFT organizational score would correlate

significantly with IQ (as with the HVOT isolate response

score, the ROCFT organizational score probably relates more

to general level of functioning and to executive systems

rather than to visuoperceptual, visuospatial, or visuomotor

abilities themselves).

(3) After controlling for any effects of age, IQ, and

sex, the children with neurological impairment or with

learning disabilities would display significantly more

impaired performance than normal children on the following

measures: (1) HVOT total score, (2) the number of HVOT

isolate response errors, (3) the ROCFT total score, and (4)

the ROCFT organizational score.



Three samples were collected: a sample of

neurologically impaired (NI) subjects, a sample of learning

disabled (LD) subjects, and a sample of normal control (NC)

subjects. In all three samples, children were aged 6

through 11. These ages were selected because, as described

earlier, these are the ages in which children are developing

visuomotor skills. At younger ages, children are

insufficiently developed to participate fully in the tasks,

and, at older ages, they have frequently come close to

attaining adult scores. Therefore, this stage of late

childhood was chosen as being the most instructive in

evaluating the development of visuomotor skills.

Of the 56 NI subjects obtained through the files of

Shands Hospital's Psychology Clinic who were in the

appropriate age range and who had confirmed central nervous

system disease, 10 were not eligible because their IQs were

below the cutoff of 70. An IQ of 70 was selected as the

cutoff score in order to balance the need, on the one hand,

to have an NI sample that was fairly representative of the

general NI population, with the needs, on the other hand,

(1) to avoid inordinately skewing the results by having



severely impaired NI children, and (2) to avoid excessive IQ

differences between samples.

All 5 children in this age range who met medical

criteria for inclusion but received neither the HVOT nor the

ROCFT had IQs below 70. These 5 children's ages (in years)

were 8 (1 child), 9 (2 children), and 10 (2 children).

Three of these children were female, 2 were male. Their

diagnoses were epilepsy with history of viral meningitis,

severe cortical atrophy secondary to anoxia, epilepsy and

left hemiparesis secondary to focal encephalomalacia,

cerebral palsy, and brain stem tumor.

In addition to the 10 children excluded because of IQ,

two children were not eligible because they had received the

HVOT only, and their confrontation naming scores were two

standard deviations or more below the mean for their age or

grade. Finally, 1 child's data was not employed because the

raw data was not available in his file. Of the 43 children

remaining, two children's HVOT scores were deleted because

of low naming; however, their ROCFT and other scores were


Fifty children obtained archivally through the

Psychology Clinic and Pediatric Behavioral Neurology Clinic

at Shands Hospital were the appropriate age and had been

diagnosed as LD by a qualified pediatric neurologist or

neuropsychologist. Of this sample, 16 children were

ineligible because they received neither the HVOT nor the

ROCFT. These children's diagnoses included dyslexia,


general language disability, visuospatial disability,

language processing disability, and visuomotor disability,

with 1 secondary diagnoses of spelling disability. Further

demographic data on these 16 children is explored in the

results section. In addition to these 16 children, 2

children were ineligible because of IQ scores below 80 (80

and 85 are standard IQ cutoffs for LD children). Because

only 6 children in the LD sample had received the HVOT, the

data for the HVOT within the LD sample was insufficient for

statistical analysis. Therefore, the 3 children who had

received only the HVOT were excluded, leaving 29 children in

the final LD sample.

In the NC sample, 40 children were initially recruited.

Six of these children were reported to be gifted on initial

screening, and therefore did not receive testing. Eight

additional children were found to have estimated IQs above

126, and were therefore excluded from data analysis (no

child had an IQ below the lower cutoff of 75). One child

did not receive testing because she had migraine headaches.

Two children's HVOT scores were not analyzed because of low

naming scores. Therefore, the NC final sample consisted of

23 children with all data available for analysis, and 2

children with ROCFT and other data appropriate for analysis.


In addition to the HVOT and the ROCFT, several other

measures were employed to evaluate visuomotor and other


functions. An estimated IQ or Full-Scale IQ was obtained

for each subject. The estimated IQs were based upon the

Vocabulary and Block Design subtests of the Wechsler

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

1974). Full-scale IQs were measured by the WISC-R or by the

Wechsler Intelligence Scale for Children-III (WISC-III)

(Wechsler, 1991) or the Wechsler Preschool and Primary Scale

of Intelligence-Revised (WPPSI-R) (Wechsler, 1989). Because

the ability to name pictures of objects is an important

component of the HVOT, confrontation naming ability was

screened using the Visual Naming subtest of the Multilingual

Aphasia Exam (MAE) (Benton & Hamsher, 1989) or the Boston

Naming Test (Goodglass, Kaplan, & Weintraub, 1978). The

HVOT and the ROCFT require visual perception and visual

analysis; these abilities were also evaluated employing the

Picture Completion subtest of the WISC-R. Because the HVOT

is intended to measure patients' ability to internally

manipulate objects' positions in space, evaluation of the

ability to physically reorganize parts of objects into

meaningful wholes was explored using the Object Assembly

subtest of the WISC-R. Finally, a measure of copying

ability which is less complex than the ROCFT and is already

widely used among children, Beery's (1989) Test of Visual-

Motor Integration, was given to subjects.


Hooper Visual Organization Test (HVOT)

Since the HVOT was reviewed above, only administration

information is provided here. For the HVOT, the directions

given to normal children were, "This is a test of your

ability to recognize pictures of objects when the pictures

have been cut up and rearranged. Look at each picture and

then decide what it might be if it were put together. For

example, look at the first picture. What would it be if it

were put together?" (Hooper Visual Organization Test manual,

1983). If the child said "Fish," the examiner responded

"That's right. It's a fish. Now do the other pictures in

the same way. Toward the end they become rather hard. Try

to give an answer even if you are not sure of it" (Hooper

Visual Organization Test manual, 1983). If the child did

not say "Fish" on the first card, the examiner said, "The

correct answer is 'fish.' You see, here is the head, the

fins, and the tail. If it were put together correctly, it

would be a fish. Now do the other pictures in the same way.

Toward the end they become rather hard. Try to give an

answer even if you are not sure of it" (Hooper Visual

Organization Test manual, 1983).

In addition to the total raw score (determined

according to the manual's scoring criteria), an isolate

response score was computed. The following responses

counted one point toward an isolate response score: "duck"

or "goose" for item 1, "hoe" for item 11, "net" for item 12,

"pencil" for item 14, "anvil" for item 18, "island" for item


21, "pipe" for item 22, "tray" for item 29, and "toilet

plunger" for item 30 (Hooper Visual Organization Test

manual, 1983; Lezak, 1983). These isolate responses were

collected because little is known about children's

tendencies to commit such errors.

Rey-Osterrieth Complex Figure Test (ROCFT)

Since the ROCFT was reviewed earlier, reliability and

validity information are not repeated here. Prior to

presenting the figure, the examiner gave normal children a 8

1/2 by 11 inch piece of white paper (presented horizontally)

and said, "I am going to show you a complicated drawing.

I would like you to draw it for me on this piece of paper.

I will be giving you a new pencil from time to time. When

I do that, please trade pencils with me, and continue

working." The Rey-Osterrieth Complex Figure was then

presented to each child horizontally, with the diamond to

the child's right. The examiner timed the child's copy of

the ROCFT, and provided the child with a new pencil when the

child began working on a new section of the drawing. If a

child did not receive the pencils in the predetermined

order, the order was recorded on a separate sheet of paper.

The total score for the ROCFT was computed using Mings'

(1987) version of Osterrieth's (1944) scoring. Mings (1987)

provides operationalized criteria for judging the accuracy

of a detail and its placement. The organizational level of


the ROCFT was determined using Waber and Holmes' (1985)

scoring criteria.

Wechsler Intellicence Scale for Children-Revised (WISC-R)
Estimated IQ (EIQ)

The Wechsler Intelligence Scale for Children-Revised

(WISC-R) is an excellent measure of Verbal IQ, Performance

IQ, and Full-Scale IQ (Spreen & Strauss, 1991; Vernon, 1984;

Wechsler, 1974). However, because of its length, it was

impractical to give the entire WISC-R to all of the normal

children who were recruited. Therefore an Estimated IQ

(EIQ) was obtained. The WISC-R was selected instead of its

recent revision, the Wechsler Intelligence Scale for

Children-III (WISC-III) because the majority of the clinical

population and learning disabled children had received the


Fortunately, a fairly accurate estimate of the WISC-R

IQ can be obtained by giving two subtests, Vocabulary and

Block Design (Sattler, 1988; Spreen & Strauss, 1991). The

Vocabulary subtest requires children to define verbally a

number of words, and the Block Design subtest involves using

red and white blocks to reproduce a red and white design

presented to the child. These subtests were administered

according to the WISC-R manual (Wechsler, 1974). Despite

the potential for systematic differences between estimated

IQs and IQs based on the entire WISC-R, the full WISC-R was

employed for NI and LD children because of its enhanced

accuracy over estimated IQ.


Vocabulary and Block Design have extremely high

internal consistency (r = .86 and r = .85, respectively) and

test-retest reliability (r = .82 and r = .81, respectively)

(Wechsler, 1974). Furthermore, the combination of Vocabu-

lary and Block Design has a higher validity coefficient

(.906) than any other two WISC-R subtests (Sattler, 1988).

Vocabulary correlates at .66, more than any other subtest,

with the Verbal IQ score (Wechsler, 1974). Similarly, Block

Design has the highest correlation with Performance IQ (r =

.66) of all the subtests (Wechsler, 1974). Finally, the

correlation between Full-Scale IQ and subtest scores is

highest for Vocabulary (.65 <= r <= .88), Block Design (.67

<= r <= .88), and Picture Arrangement (r = .67) (Haynes,

1982; Ryan, 1981; Wechsler, 1974). Given these data,

Vocabulary and Block Design provide the most accurate

Estimated IQ that can be obtained from only two subtests.

A few of the children obtained through the Psychology

Clinic (13 NI children) or the Pediatric Behavioral

Neurology Clinic (3 LD children) received the WISC-III

(Wechsler, 1991). Although the possibility of systematic

differences between WISC-III Full-Scale IQ and WISC-R

Estimated IQ poses problems, Vocabulary and Block Design for

the WISC-III do correlate highly (.77 and .76, respectively)

with the WISC-R (Wechsler, 1991). Because a WISC-III IQ

score would be more valuable than no IQ at all, the WISC-III

Full-Scale IQ scores were employed. However, the WISC-III

Full-Scale IQ tends to be about 5 points less than the


WISC-R Full-Scale IQ (Wechsler, 1991). Although the

differences between the Vocabulary-Block Design estimated

IQs for the WISC-R and Full-Scale IQs for the WISC-III have

not been evaluated, it is likely that the WISC-III Full-

Scale IQs are also lower than the WISC-R estimated IQs. For

accurate comparison among subjects, these scores required

equalization. Because more children received the WISC-R,

the WISC-R was used as the standard. Therefore, 5 points

were added to all WISC-III IQs or estimated IQs.

One NI child and one LD child received the Wechsler

Preschool and Primary Scale of Intelligence Revised

(WPPSI-R) (Wechsler, 1989). The WPPSI-R Full-Scale IQ

correlates at .85 with the WISC-R; however, it yields scores

about 7 points lower than the WISC-R. Seven points were

therefore added to all WPPSI-R IQ scores.

Multilingual Aphasia Exam (MAE) Visual Naming (VN)

The Visual Naming (VN) subtest of the Multilingual

Aphasia Exam (MAE) (Benton & Hamsher, 1989) requires the

subject to name 30 pictures of objects (e.g., circle, piano,

fork) or parts of objects (e.g., pedals, handle). For nor-

mal children, the examiner pointed to the picture (or part

of the picture) and asked the child "What do you call this?"

or "What is the name of this part?" (Benton & Hamsher,

1989). For adults, VN scores correlate highly with educa-

tional level, but do not correlate with age or sex (Benton &

Hamsher, 1989). Furthermore, after controlling for educa-


tional level, VN scores tend to be lower for aphasic than

for normal adult subjects: 18% of normal subjects receive

scores of 48 or less, versus 90% of aphasic subjects obtai-

ning scores of 48 or less (Benton & Hamsher, 1989). In

children, VN scores increase with grade, and approach adult

levels in 6th grade (Benton & Hamsher, 1989). Finally, in

support of criterion validity, 46% of children with severe

communication difficulties (dyslexia, stuttering, and

expressive language disorders) obtained impaired scores on

VN (Benton & Hamsher, 1989).

Boston Naming Test (BNT)

Some of the children collected archivally received the

Boston Naming Test (Goodglass, Kaplan, & Weintraub, 1978)

instead of the MAE VN subtest. The BNT requires the patient

to name up to 60 line drawings of objects, and provides

semantic and phonological cues when the patient is unable to

spontaneously name the item. The BNT discriminates between

dyslexic and normal readers (Cohen, Town, & Buff, 1988; Wolf

& Obregon, 1992). Yeates (1994), in his review of the lite-

rature, noted that in all studies children's scores gene-

rally increased with age. Using composite norms from all

five normative studies of the BNT in children, Yeates (1994)

determined that, until 11 years of age, each 1-year age

range received scores which were significantly larger than

the scores of those children one year younger. Yeates'

(1994) composite normative data were used to determine whe-