6
A FACTOR ANALYSIS OF ITPA AND WISC SCORES OF LEARNING-DISABLED PUP1 LS DONALD A. LETON University of Hawaii The complementary nature of the Wechsler Intelligence Scale for Children (WISC) and the Illinois Test of Psycholinguistic Abilities (ITPA) is assumed pro- fessionally by their joint use to assess learning-disabled pupils. Meyers (1969) has reviewed the factor studies of the ITPA and concluded that there are 7 intellectual abilities represented in the 9 subtests, as follows: (I) Verbal Comprehension, (11) Vocal-Motor Expression, (111) Meaningful Figural Comprehension, (IV) Immediate Visual Memory, (V) Vocal Expression, (VI) Immediate Auditory Memory, and (VII) Vocal Decoding of Receptive Language. Cohen (1959) has factor analyzed the WISC subtests at three age levels, 7%, 10% and 13% years, and concluded that there are 5 intellectual abilities represented in the 12 subtests, as follows: (I) Verbal Comprehension, defined as retention of educative knowledge, (11) Perceptual Organ- ization, (111) Freedom from Distractability, (IV) Verbal Comprehension defined as applied judgment following verbal manipulation, and (V) Coding. WISC and ITPA subtest profiles have been published for various groups of retarded and handicapped children. There are no known factor analytic studies, however, of the intelligence traits of learning-disabled children. The extent to which the WISC and the ITPA subtests assess common and unique factors has not been studied as yet. Learning disabled (LD)’ is a heterogeneous classification of pupils for special education. Although these pupils meet a prerequisite criterion of ‘hormal” intelli- gence they show a variety of symptoms and conditions that are attributed to mini- mal brain dysfunction (Chalfant, 1969). These may include dyslexia, dysphasia, dys- graphia, organically driven behavior, distractibility, memory deficit, and the imper- ception of visual, aural, or tactual stimuli. The learning disability often is known to be, and sometimes assumed to be, an expression of a cerebral disorder. Intactness of the afferent, efferent, and coordinate-proprioceptive systems generally is pre- sumed. The criteria for the diagnosis of learning disability, which are applied in the various clinical specialties, are important determinants of the educational charac- teristics of the classroom groups. METHOD Subjects The pupils in the criterion group for this study were examined at the Learning Disabilities Clinic of the Hawaii State Department of Health. The clinic staff includes a pediatrician, a child neurologist, a psychologist, a speech and language specialist, and SL social worker. The Clinic diagnosis of “learning disabled’’ generally indicated a neurogenic learning or activity-control problem. It precluded mental retardation and psychogenic emotional disturbance as alternative causes for the learning difficulty. The pupils were certified as learning disabled and placed in special classes or in special education programming in the Honolulu School District. All of the pupils in the LD classes are retarded academically. Their scores on the Jastak Wide Range Achievement tests are generally 1.5 to 5.0 grade levels below their actual grade placements. The younger pupils show less discrepancy between 1Abbreviation LD hereafter used to refer to learning disabled.

A factor analysis of ITPA and WISC scores of learning-disabled pupils

Embed Size (px)

Citation preview

Page 1: A factor analysis of ITPA and WISC scores of learning-disabled pupils

A FACTOR ANALYSIS OF ITPA AND WISC SCORES OF LEARNING-DISABLED PUP1 LS

DONALD A. LETON

University of Hawaii

The complementary nature of the Wechsler Intelligence Scale for Children (WISC) and the Illinois Test of Psycholinguistic Abilities (ITPA) is assumed pro- fessionally by their joint use to assess learning-disabled pupils. Meyers (1969) has reviewed the factor studies of the ITPA and concluded that there are 7 intellectual abilities represented in the 9 subtests, as follows: (I) Verbal Comprehension, (11) Vocal-Motor Expression, (111) Meaningful Figural Comprehension, (IV) Immediate Visual Memory, (V) Vocal Expression, (VI) Immediate Auditory Memory, and (VII) Vocal Decoding of Receptive Language. Cohen (1959) has factor analyzed the WISC subtests at three age levels, 7%, 10% and 13% years, and concluded that there are 5 intellectual abilities represented in the 12 subtests, as follows: (I) Verbal Comprehension, defined as retention of educative knowledge, (11) Perceptual Organ- ization, (111) Freedom from Distractability, (IV) Verbal Comprehension defined as applied judgment following verbal manipulation, and (V) Coding.

WISC and ITPA subtest profiles have been published for various groups of retarded and handicapped children. There are no known factor analytic studies, however, of the intelligence traits of learning-disabled children. The extent to which the WISC and the ITPA subtests assess common and unique factors has not been studied as yet.

Learning disabled (LD)’ is a heterogeneous classification of pupils for special education. Although these pupils meet a prerequisite criterion of ‘hormal” intelli- gence they show a variety of symptoms and conditions that are attributed to mini- mal brain dysfunction (Chalfant, 1969). These may include dyslexia, dysphasia, dys- graphia, organically driven behavior, distractibility, memory deficit, and the imper- ception of visual, aural, or tactual stimuli. The learning disability often is known to be, and sometimes assumed to be, an expression of a cerebral disorder. Intactness of the afferent, efferent, and coordinate-proprioceptive systems generally is pre- sumed. The criteria for the diagnosis of learning disability, which are applied in the various clinical specialties, are important determinants of the educational charac- teristics of the classroom groups.

METHOD Subjects

The pupils in the criterion group for this study were examined a t the Learning Disabilities Clinic of the Hawaii State Department of Health. The clinic staff includes a pediatrician, a child neurologist, a psychologist, a speech and language specialist, and SL social worker. The Clinic diagnosis of “learning disabled’’ generally indicated a neurogenic learning or activity-control problem. It precluded mental retardation and psychogenic emotional disturbance as alternative causes for the learning difficulty. The pupils were certified as learning disabled and placed in special classes or in special education programming in the Honolulu School District.

All of the pupils in the LD classes are retarded academically. Their scores on the Jastak Wide Range Achievement tests are generally 1.5 to 5.0 grade levels below their actual grade placements. The younger pupils show less discrepancy between

1Abbreviation LD hereafter used to refer to learning disabled.

Page 2: A factor analysis of ITPA and WISC scores of learning-disabled pupils

32 DONALD A. LETON

their actual and expected performances. A few children are classified as LD because of their organically-driven behavior. Some may require special class placement, and others may be observed in their regular classes during the course of prescribed medical or nonmedical behavior modification treatments. The mild LD cases, ix. , pupils who are achieving near their expected age and intelligence levels and have mild perceptual-motor distractibility problems, are not placed in special classes. These pupils continue in regular classes and receive supplementary instruction from an LD resource teacher. The mild LD pupils enrolled in supplementary programs are not included in this study.

Proceduws The WISC and ITPA data analyzed in this study were obtained from clinic

reports and from reports of case studies by the psychological examiners in the Honolulu School District. For the 187 LD pupils enrolled in the 16 elementary grade classes in January 1971, 167 complete WISC profiles were available. This represents 89% of the enrollment.2 In some cases, the Stanford-Binet was admin- istered; in some cases, only the WISC Verbal or Performance subtests were ad- ministered; in a few cases, the WISC was administered by psychologists in private practice, and subtest results were not included in the clinic reports.

There were 109 LD pupils for whom the ITPA profiles (1961 edition) were available. The ITPA was included in the psychological assessment for about 70% of the sample, however, in some cases only selected subtests had been administered. Pupils who were administered the 1969 edition of the ITPA were not included.

Of the 109 pupils in the ITPA sample, there were 17 for whom complete WISC profiles were not available. The analysis reported in this study is based on an end sample of 92 pupils for whom complete WISC and ITPA profiles were available. Sixty-seven of the 92 pupils or 73% of the sample were boys. The majority of boys among LD cases typically is observed. The ages for the administration of the Wechsler Intelligence Scale ranged from 5 years 1 month to 13 years 9 months, with a mean age of 7 years 7 months. The standard deviation of the distribution of ages for the administration of the WISC was 20.5 months. The ages for the administra- tion of the ITPA ranged from 5 years 3 months, to 13 years 9 months, with a mean of 7 years 9 months. In some cases the WISC was administered by the school psy- chologist and the ITPA by the clinic psychologist. In these cases there was a 1- to 6-month interval between the WISC and the ITPA. I n the majority of cases the two tests were administered by the same psychologist, a t the same testing session or within a 2- to 3-week interval.

RESULTS The summary statistics for the WISC were as follows: Verbal I& mean 82.8,

SD 8.60; Performance I& mean 92.8, SD 10.93; Full Scale I& 86.4, SD 8.14. The Verbal Intelligence scores ranged from 56 to 105; the Performance scores ranged from 69 to 124; the Full Scale scoresranged from 71 to 112. The ITPA total language score mean was 142.6, SD 30.2.

PForty-two pupils enrolled in four secondary level classes were not included in this survey because of the wider variabilit in age-at-identification and because of the variety of clinics and specialists formerly involved in t i e identification.

Page 3: A factor analysis of ITPA and WISC scores of learning-disabled pupils

TA

BL

E 1.

SUM

MA

RY

ST

AT

IST

ICS

FOR

WIS

C S

UB

TE

ST

SCO

RE

S FOR

LEA

RN

INQ

-DIS

ABL

ED

PUPI

LS

Info

. C

omp.

A

rith

. Si

mil.

V

ocab

. D

igits

C

ompl

. A

rr’t.

D

esig

n A

ss’b

ly

Cod

ing

__

-.

Pict

. Pi

ct.

Blo

ck

Obj

ect

Mea

n 6.

85

7.79

7.

76

7.09

6.

52

6.47

9.

38

9.C

6 9.

02

8.22

8.

34

SD

2.13

2.

78

3.03

3.

07

2.54

2.

78

2.05

2.

80

2.81

3.

12

2.63

TA

BL

E

2. SU

MM

AR

Y

STA

TIS

TIC

S FO

R I

TPA

SU

BT

EST

SC

OR

ES FO

R L

EA

RN

ING

DIS

AB

LE

D

PU

PIL

S

Aud

itory

A

udito

ry

Visu

al

Aud

itory

V

iual

V

ocal

V

isual

M

otor

V

ocal

M

otor

V

ocal

V

ocal

M

otor

A

udito

ry

Aut

omat

ic

Dec

odin

g E

ncod

ing

Ass

ocia

tion

Sequ

enci

ng

Enc

odin

g Se

quen

cing

A

ssoc

iatio

n D

ecod

ing

Mea

n -1

.76

- .2

3 - .5

3 -1

.10

-1.3

4 -

.44

- .9

3 -

.49

- .4

1 SD

1.

08

1.05

.9

5 1.

19

.63

.81

1.06

.9

5 1.

22

TA

BL

E

3. IN

TE

RC

OR

RE

LA

TIO

N

OF

WIS

C A

ND

IT

PA SW

TEST

S*

~ ~~

Subt

est

23

45

67

8

1.

Info

rmat

ion

38

-04

43

38

17

07

05

2. C

ompr

ehen

sion

-04

35

25

31

29

-03

3.

Ari

thm

etic

-0

1 -1

1

-06

-15

-07

4. Si

mila

ritie

s 16

18

27

08

6. D

igits

26

10

8.

Pict

ure

Arr

ange

men

t 9.

B

lock

Des

ign

5. V

ocab

ular

y 11

20

20

7. Pi

ctur

e C

ompl

etio

n - 0

4

10.

Obj

ect A

ssem

bly

11.

Cod

ing

13.

Vis

ual D

ecod

ing

14.

Mot

or E

ncod

ing

15.

Aud

.-Voc

. Ass

’n.

16.

Vi.-

Mot

or S

eq.

17.

Voc

al E

ncod

ing

19.

Vis

ual-M

otor

Ass

’n.

20.

Aud

. Dec

odin

g

12.

Aud

.-Voc

. Aut

o

18.

Aud

.-Voc

. Seq.

9 10

-19

01

06

01

04 -01

07

-12

-04

14

0009

19

,24

23

-02 30

11

12

13

14

15

16

17

18

19

20

-00

23

17

-08

-08

-08

-03

40

18

17

14

20

02

14

45

17

14

05

18

18

24

-03

06

-01

25

-08

10

23

07

-13

-14

10

21

17

25

-11

14

21

-01

-11

-12

05

16

48

-00

13

33

-02

00

17

-12

00

32

25

18

03

13

29

-07

10

20

01

12

02

-15

23

27

49

08

20

35

16

23

08

- 02

36

26

- 08 30

26

35

19

10

-11

16

- 10 41

18

26

44

08

12

-09

23

18

02

07

27

07

23

18

38

07

30

15

08

14

01

16

14

08

07

22

28

10

12

34

13

21

33

05

08

18

43

05

14

14

18

- 23

11

18

02

13

07

-06 15

-c

3 30

01

24

46

- 01 27

19

05

P

0

w Y 2 P 3 8 ;; d u)

d 0

s L

*Dec

imal

s om

itted

0

w

Page 4: A factor analysis of ITPA and WISC scores of learning-disabled pupils

34 DONALD A. LETON

The means and standard deviation for the distributions of the WISC and the ITPA subtest scores are presented in Tables 1 and 2 respectively.

The 20 subtests in the WISC-ITPA profiles were intercorrelated. The cor- relation matrix, presented in Table 3, was factor analyzed by means of principal components analysis. Seven factors with eigenvalues greater than 1.0 were ex- tracted. The matrix of seven principal axes was rotated orthogonally with a criterion of simple structure, according to Kaiser’s (1959) varimax procedures.

The primary loadings for the 20 subtest variables, the secondary loadings above .40, variable communalities, and the eigenvalues and percent of variance for the 7 factors are presented in Table 4.

TABLE 4. VARIABLE LOADINGS’, COMMUNALJTIES .4ND FACTOR STRENGTHS IN ROTATED FACTOH METHOD MATRIX

Variables I I1 I11 IV v VI VII h2

Information .83 .79 Comprehension .61 .59 Arithmetic ,532 .68 Similarities .66 .65 Vocabulary .46 .48 .46 Digits .81 ‘72 Picture Completion .47 .4s61 .64 Picture Arrangement .65 .69 Block Design .76 .63 Object Assembly - .60 .72 Coding - .43 - .40 .60 Auditory-Vocal Automatic .75 .60 Visual Decoding .74 .72 Motor Encoding .42 58 .60

.72 - .73 .61 Visual Motor Sequencing Vocal Encoding .61 .43 .62 Auditory-Vocal Sequencing .70 .65 Visual-Motor Association .54 .50

.56 Auditory Decoding .53 .61

-.

Auditory-Vocal Association .65 .41

- Eigenvalues 4 . 1 2 . 1 1 . 6 1 .4 1 . 3 1 . 1 1 .0 Percent of Variance 12.8 9 . 6 8.8 9 . 8 7.9 6.5 8.8

*Secondary loadings above .40 appear in italics. -~

Factor I is identified as Verbal Association and is primarily an ITPA language ability. The manifestation of this ability is based on aural receptive and vocal expressive language. The associational process is identified in the Auditory-Vocal Association, the WISC Vocabulary and Coding, and the ITPA Encoding subtests. The semantic, or language meaning, nature of this factor is identified by the Vo- cabulary, Auditory-Vocal Automatic, Auditory-Vocal Association and Auditory Decoding subtests. In Guilford’s (1967) theory of the structure of the intellect this factor would be identified as the cognition of semantic units, semantic relations, and semantic systems (CMU, CMR, and CMS).

Fact’or I1 is identified as Visual Analysis and Motor Association. The primary loadings of the Block Design and Picture Arrangement subtests define the visual analysis requirements in the stimuli, and the Motor Encoding and Visual-Motor Association subtests define the motor-associational nature of the responses. Mani- festation of this ability is based on the perception of figures and the analysis of figural relationships. In Guilford’s theory this factor would be identified as CFU and CFR, the cognition of figural units and figural relationships.

Factor I11 is identified as the Comprehension of Similarities and Differences.

Page 5: A factor analysis of ITPA and WISC scores of learning-disabled pupils

A FACTOR ANALYSIS OF ITPA AND WISC SCORES O F LEARNING-DISABLED PUPILS 35

The nature of the factor is defined in the Visual Decoding and Similarities subtests. With the negative loading of the Object Assembly subtest it appears to be a bipolar factor. This may indicate a basic requirement of figure recognition for the mani- festation of the ability. In Guilford’s theory this factor would be identified as the cognition and convergent production of symbolic relations and classes, CSR, NSR, and NSC.

Factor IV is identified as Auditory Memory. The WISC Memory for Digits and the ITPA Auditory-Vocal Sequencing variables define the nature of this factor. In his factor analysis of the WISC subtests, Cohen (1959) identified the ability assessed in the Memory for Digits and the Picture Arrangement subtests as Freedom from Distraction. In this analysis the Picture Completion subtest loaded on this factor, which reflects a correlation of visual retention and auditory memory abilities. The secondary loading for the Vocal Encoding subtest and a secondary loading of .30 for the Comprehension subtest (not shown in the table) indicate that this factor may include verbal retention as well as immediate recall. I n Guilford’s theory of the structure of intellect this factor would be identified as memory for symbolic systems, MSS.

Factor V is identified as Visual Sequencing. The nature of this factor is defined by the primary loadings of the Picture Arrangement and Visual-Motor Sequencing subtests. The secondary loadings indicate a base in visual analysis and motor response abilities. In GuiIford’s theory this factor would be identified as the memory for figural systems and relations, MFS and MFR.

Factor VI commonly is identified as Logical Reasoning. In Guilford’s theory this would be identified as the cognition of symbolic relations and of semantic sys- tems, CSR and CSS.

Factor VII is identified as the Verbal-Educative factor. In Cohen’s analysis of the WISC and WAIS this factor is identified as Verbal Comprehension I. He postulates that it reflects verbally retained knowledge from formal education. I n Guilford’s theory this factor would involve several processes, including retentive memory process as assessed in the Information subtest, evaluation process as assessed in the Comprehension subtest, and the cognition of semantics, as assessed in the Vocabulary subtest.

The purpose of this analysis was to determine the minimum number of inde- pendent factors that account for the major portion of the reliable variance in the domain of WISC and ITPA measurements of LD pupils’ abilities. This purpose dictated the criterion of orthogonal factors. Factors with eigenvalues less than 1.0 were not included in the rotation. A greater number of correlated factors was ob- tained in another analysis that used oblique rotations. The extraction of additional factors, however, tended to complicate the task of meaningful interpretation rather than to simplify it.

DISCUSSION This analysis of the WISC and ITPA is based on an assumption that they both

assess perceptual, conceptual, lingual, associative and retentive abilities. They both require visual and aural reception of stimuli and vocal and motor expressions of response. In spite of their common purposes and procedures the task of naming their common dimensions is not an easy one.

Using terms from classical factor-analytic studies of intelligence and from

Page 6: A factor analysis of ITPA and WISC scores of learning-disabled pupils

36 DONALD A. LETON

Guilford’s theory of the structure of intellect, an attempt was made in this article to name the dimensions of ability manifested in the WISC and the ITPA perform- ances of LD pupils. Because of the semantic problem involved with the connotations of terms there is no assurance that another psychologist would use the same terms to identify the ability factors. A statistician’s escape would be merely to number the factors and list the variables loading on each factor.

Inasmuch as the WISC and ITPA are used jointly in the clinical assessment of the intellectual and language abilities of LD pupils, it seems logical that the factor validities of the two instruments should be established for samples of LD pupils. The canonical correlation of the abilities of superior, average, learning disabled and mentally retarded also should be determined. Although perceptual and memory abilities may be prominent in the matrix of abilities of learning disabled, perhaps these traits would be less prominent in the matrix of abilities of children with superior intelligence.

The correlation of measures of visual and auditory memory was observed in this analysis. This should not imply that independent assessments of these two receptive bases for perception and memory are impossible. The existence of independent per- ceptual modalities has been hypothesized by a number of cognition theorists. Al- though this factor analysis of the WISC and ITPA does not support such hypotheses, the theory is not invalidated thereby. Since the WISC and the ITPA subtests were not developed as factor-pure measurements they may not be suited for such dis- tinctions.

The identification of the seven abilities in this analysis should extend to the interpretive value of both instruments. The correlation and factor matrix provide evidence of their common structure and justify their joint use. Only two factors are exclusive instrument factors. Factor I is primarily an ITPA auditory-vocal factor, and factor VII is a WISC verbal intelligence factor. Factor I1 supports an inter- pretation of motor-associative ability independent of auditory-vocal association as identified in factor I. The independence of memory and sequencing abilities is indi- cated in factors IV and V. This distinction may be important for the diagnosis of achievement difficulties and for the prescription of instructional procedures. Al- though factor VI, logical reasoning, is a relatively weaker factor it is easily identified.

There is a false notion among many psychologists that each of the WISC and ITPA subtests assesses a different dimension of ability. Although a majority of psychologists would disclaim such a notion, their clinical interpretations of WISC and ITPA subtest scores tend to belie their disclaimers. The identification of the basic dimensions of these measures should improve their clinical use and, through further research, can add to their predictive validities.

REFERENCES CHALFANT, J. C., & SCHEFFELIN, M. A. Central processing dysfunctions in children: a review of re-

search. Bethesda, Md. : National Institute for Neurological Diseases and Blindness, Monograph No. 9. 1969.

- I - ~ - - COHEN:-J. The factorial structure of the WISC at ages 7-6, 10-6, and 13-6. Journal of ConsuZtinq

GUILFORD. J. P. The nature of human intelligence. New York: McGraw-Hill. 1967. Psychology, 1959, 23, 258-299.

KAISER, H. F. Computer program for varimax rotation in factor analysis.’ Educational and Psycho-

MCCARTHY, J. J., & KIRK, S. A. The Illinois Test of Psycholinguistic Abilities. Champaign, ID.:

MEYERS, C. E. What the ITPA measures: a synthesis offactor studies of the 1961 Edition. E d u w

logical Measurement, 1959, 19, 413-420.

University of Illinois, Institute for Research on Exce tional Children, 1961.

tional and Psychological Measurement 1969, 81, 867-876.