![](https://csdnimg.cn/release/download_crawler_static/19905287/bg1.jpg)
Psychology
in
the
Schools
Volume
24,
July
1987
A LONGITUDINAL STUDY OF THE WECHSLER INTELLIGENCE SCALE
FOR CHILDREN-REVISED OVER A SIX-YEAR PERIOD’
BOONEY VANCE, NORMAN HANKINS,
AND
WESLEY BROWN
East Tennessee State University
The study investigated the stability of the WISC-R
IQ
scores of 32 exceptional students
(24
males,
8
females) over
a
six-year interval. The sample was composed
of
20
learn-
ing disabled and
12
mentally handicapped students who ranged
in
age from 6-5 to
16-1
1.
Analyses of variance for repeated measures and product-moment correlations
were used to analyze the data. Test-retest findings indicated that the
rs
ranged from
.53
to .87, with a median value of .74, with
four
of
the nine values below .70. Im-
plications for clinicians in terms of the importance
of
reevaluation of exceptional
students are discussed.
Stability
of
ability assessment data is essential if placement decisions are not to be
reconsidered at frequent intervals. The most common method for assessing the stability
of
a
test is
to
administer it to a group of subjects and then to readminister the instru-
ment to the same group at a later date. The correlation between the two sets of scores
becomes the coefficient. The longer the delay between the original administration and
the subsequent testing, the lower the stability coefficient, usually.
Wechsler (1974) investigated the stability of the WISC-R by retesting 303 children
from the standardization sample after a one-month period. The results of Wechsler’s
study indicated that the average increase in IQ between the first and second testing was
7
IQ points for the Full Scale, about 4 points for the Verbal Scale, and about 10 points
for the Performance Scale (Sattler, 1982). The stability coefficients ranged from .95 for
the Full Scale to .90 for the Performance and .93 for the Verbal Scale. Sattler explains
these differences as practice effect. Vance, Blixt, Ellis, and DeBell (1981) investigated
the stability of the WISC-R for a sample of exceptional children over a two-year inter-
val. The results
of
the Vance et al. study indicated that the WISC-R is a very reliable
instrument over time when used with learning disabled and retarded youngsters. For
their sample, the stability coefficients were .80 for the Verbal Scale IQ, .91 for the Per-
formance Scale IQ, and .88 for the Full Scale IQ. The mean absolute change in measured
IQ on the Verbal, Performance, and Full Scales was about two points. The sample score
was two points lower on the Verbal Scale IQ and two points higher on the Performance
Scale IQ, with the Full Scale IQs being about equal.
An interesting study conducted by Smith and Rogers (1978) compared the reliabil-
ity of standardized assessment instruments when used with learning disabled children.
Smith et al. found for a sample of learning disabled children, who were tested after
approximately
a
six-month interval, that the coefficients
of
temporal stability on the
IQ indices ranged from .79 to
.82.
In addition, these investigations compared the cor-
relation between the stability coefficients for their LD sample and the WISC-R standard-
ization sample and found a .597
0,
=
,006)
correlation. However, this finding could have
been vitiated by the different test-retest interval for the two groups. Smith et al. con-
cluded that “the examined psychometric properties of these instruments (WISC-R, PIAT,
Requests for reprints should be sent to Booney Vance, Center for Early Childhood Learning and Develop-
‘This study was sponsored in part by a grant from the Research and Development Committee, East Ten-
ment, East Tennessee State University,
P.O.
Box 15, 520A, Johnson City, TN 37614-0002.
nessee State University.
229