Doctoral Dissertations

Date of Award

8-1984

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Psychology

Major Professor

Richard A. Saudargas

Committee Members

J. Albery Wiberley, Michael S. Johnson, Charles L. T.

Abstract

This study investigated the relationship among three intelligence tests—the Wechsler Intelligence Scale for Children—Revised (WISC-R), the System of Multicultural Pluralistic Assessment (SOMPA) and the Kaufman Assessment Battery for Children (K-ABC)—and their relationship with concurrent and future reading and arithmetic achievement as measured by the Peabody Individual Achievement Test (PIAT) and the Wide Range Achievement Test (WRAT). This relationship is significant because the theoretical bases of the SOMPA and K-ABC dictate that as more pure measures of academic potential, they should not reflect students' culture, socioeconomic status or urban/rural setting to the degree that the WISC-R does.

Subjects included forty rural, low-income white students enrolled in a small school system in eastern Tennessee. Subjects were all recipients of free school lunches under income requirements established by the United States Department of Agriculture. Numbers of subjects were split evenly according to sex and age (6- or 11-yearsold). The latter was the maximum age span which would allow overlap of the three intelligence tests.

The full scale scores obtained from the three intelligence tests were significantly correlated < .01, adjusted for repeated tests of significance). This suggested that subjects received similar patterns of scores from the three tests.

The remaining analyses of data were conducted by ANOVA procedures which accomodated the use of repeated measures from the subjects. Critical ratios for F were also adjusted for repeated testing for significance.

There were no significant differences with p < .05 found in the sample's mean scores obtained from the three intelligence tests, suggesting that the SOMPA and K-ABC were not discernably less influenced by the subjects' families' income and rural setting than was the WISC-R. Also no effects due to the factors of sex, age, or to interaction of factors was found.

Many state departments of education require that determinations of the existence of learning disabilities be made using an achievement-intelligence discrepancy model which assumes that "normal" achievement standard scores should equal the intelligence standard score. Using data values of [Obtained Achievement]/[Expected Achievement] where Expected Achievement was replaced by each intelligence test's full scale standard score, analyses were made to determine whether one of the three tests was a better predictor of either current or future achievement. No differences in accuracy were found with respect to concurrent or predictive (7 months later) achievement measures; FIAT or WRAT scores; and reading decoding, comprehension or arithmetic computation subtest scores.

The general conclusion was that the three intelligence tests should lead to similar global findings when used with students from a population similar to that of the subjects. The tests did not yield results appreciably different from each other with respect to accuracy in predicting achievement.

The finding of equal means for the three sets of intelligence test scores was inconsistent with previous findings which reported a 5 point SOMPA—WISC-R discrepancy and, with one exception, a 3 to 4 point K-ABC—WISC-R discrepancy for similar subjects. It was suggested that a sample from even lower income or comprised of nonwhite subjects would be a better test of the SOMPA's performance. The present sample was possibly too close to average with respect to culture and living conditions to represent the groups for which the SOMPA was originally intended. There did not seem to be common characteristics among the K-ABC studies to allow suggestions to be made regarding why the present findings were discrepant.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS