Note on Concurrent Validity of Structure of Intellect Gifted Screener with Wechsler Intelligence Scale for Children-Revised

Abstract

A primary problem in identification of gifted children is the time required for indiv~dual intelligence testing to establish eligibility. An effective screening instrument could eliminate such expensive testing time. The Structure of Intellect Gifted Screener (2, 3) might fill this need if scores on subtests or the full battery correlate highly with the primary placement tool, the Wechsler Intelligence Scale for Children-Revised (5). The.Screener has 12 subtests from the Structure of Intellect Learning Abilities Test. The test can be administered to groups of children, although the authors note that some portions of the rest may require individual administration for non-readers. Administration time ranges from about 40 min. at the beginning level to about 50 min. for advanced level students. No special training is required for administration. No information is given in the manual concerning the reliability or validity of the instrument, although consmct validity is indirectly addressed through reference to the fact that the test is based on Guilford's Structure of Intellect Factors (1). The present study was done to obtain evidence regarding the concurrent validity of the Gifted Screener in relation to the WISC-R battery. The Gifted Screener was administered in grade-lwel group sessions to 339 pupils, Grades 1 through G in the fall of 1978. The subjects were 163 boys and 176 girls. Following administration, each student was individually given the WISC-R full battery by a qualified psychological examiner. Stepwise multiple regression analysis (4) provided results of parrial or simple regression of each subtest for the Gifted Screener against each subtest of the WISC-R and against the WISC-R composite indicator. When all Gifted Screener scores were employed as predictor variables, all coefficients were significant. The multiple correlation of ,337 (P < .05) between the WISC-R Full Scale IQ and the weighted combination of all Gifted Screener subtests explained only about 11% of the total variance. About 89% of the variance in the WISC-R scores was lefr unexplained by the Gifted Screener. No predictor variable contributed to a significant multiple correlation coefficient when the Full Scale IQ or any combination of WISC-R subtests were used as the criterion variable. Identification of gifted students could be handled better were there a screening instrument which has a high correlation wirh the full WISC-R. The present study produced correlations between the Gifted Screener and the WISC-R too small to consider this test satisfactory. As the intent in development was not screening gifted children, one would nor expect to find a strong relationship between the two tests. Both the screener and the learning sub-scale idenrify areas of strength and weakness in children's intellectual performance and may add to total assessment of gifted children. However, where the WISC-R is ordinarily used as the final identification and placement instrument, use of the Gifted Screener appears inappropriate.