Dennis P. Saccuzzo
Nancy E. Johnson
San Diego State University
San Diego, CA
Identifying Underrepresented Disadvantaged Gifted and Talented Children: A Multifaceted Approach was a 3-year grant funded from October 1990 through December 1993 by the U.S. Department of Education, Jacob Javits Gifted and Talented Discretionary Grant Program. The purpose of the grant was to evaluate various models for using traditional psychometric tests for selecting diverse students for gifted and talented programs. The testing ground for this endeavor was the San Diego City School District, a system serving over 123,000 children of whom approximately 29% are Latino/Hispanic, 38% Caucasian, 16% African-American, and the remainder composed of five additional ethnic backgrounds.
In support of the objectives of the grant, the district made available a large archival data set of all children who had been evaluated for giftedness between 1984 and 1990, and allowed us to input all data on children referred and evaluated during the grant period. In the end, an extensive data file of over 26,000 potentially gifted children had been created. Of these, over 9,000 had been given the Wechsler Intelligence Scale for Children-Revised (WISC-R) and over 16,000 were given the Raven Standard Progressive Matrices (SPM) Test.
During the 1984-1990 period, the WISC-R had been the primary instrument used to determine giftedness. Students who obtained a Full Scale WISC-R IQ of 130 or greater or a Full Scale WISC-R IQ of 120 with at least two of six risk factors (cultural, language, emotional, economic, health, and environmental) were certified as gifted. Extensive analysis of the data led to two major conclusions. First, there were inequities in the referral process. For example, based on their proportion in the district as a whole and assuming that giftedness is evenly distributed across ethnic backgrounds, Latino/Hispanic children were underrepresented in the referral process by a factor of 4 (i.e., the number tested represented only 25 percent of their actual proportion in the district). Second, an exhaustive analysis that evaluated all major systems and models for weighting WISC-R subtests revealed that the WISC-R could not be used to produce ethnically proportionate representation (i.e., children selected across ethnic backgrounds in proportion to their actual numbers in the district population). These findings and conclusions are documented in a monograph (Saccuzzo, Johnson, & Guertin, 1994) and in articles presently under editorial review.
Given the referral bias uncovered by our analysis of the archival data from the 1984-1990 period, the school district made an effort to achieve proportionate representation in the referral process through teacher training (to help identify potentially gifted traditionally underrepresented students) and through central nominations. At the same time, the district shifted from the WISC-R to the SPM in order to find a culture-reduced measure of intellectual giftedness.
There was a considerable shift toward proportionate representation in the referral process during the 1991-1993 period. Moreover, the use of the SPM in conjunction with an evaluation for risk factors led to the identification of thousands of traditionally underrepresented children who otherwise would not have been selected for the gifted program. While the SPM did lead to increased equity for all ethnic groups in that each ethnic group was selected in greater proportion to their numbers in the population as a whole, it did not produce a completely balanced result for all groups. Again, these results are presented in a monograph (Saccuzzo et al., 1994) and in papers in submission.
In brief, our results comparing the WISC-R and SPM revealed that the two measures had equal predictive validity and showed no differential validity as a function of ethnic background. The SPM proved to be far better than WISC-R in terms of a proportionate representation model of bias, but was not entirely free of such bias. We conclude, based on our findings and on previous reviews of psychometric tests (Kaplan & Saccuzzo), that no traditional test, as presently used, can meet the rigors of proportionate representation.
Given the large data set, we were able to conduct numerous analyses of special interest, as reported in our monograph. In one study, intellectually gifted children from diverse ethnic and cultural backgrounds as well as varying levels of risk were evaluated to determine the effect of risk on gifted children when intelligence level has been controlled. Each of the 7,323 children from six ethnic backgrounds had achieved a standardized intelligence test score (Wechsler Intelligence Scale for Children-Revised or Raven’s Standard Progressive Matrices) at least two standard deviations above the mean. Although each child in the sample had demonstrated high intellectual potential, differences were found between groups defined on level of risk: no risk, low risk (one and only one area of risk), and high risk (more than one area of risk). High-risk gifted children were disadvantaged relative to those at low or no risk in all measures of both aptitude and achievement, as assessed with the Developing Cognitive Abilities Test and the Comprehensive Test of Basic Skills. Furthermore, those at high risk demonstrated lower WISC-R Verbal IQ scores than children at lower levels of risk.
Our data also allowed us to analyze gifted underachievers. A well-defined sample of gifted underachievers was compared to a sample of gifted high-achievers. All children had full scale WISC-R IQ scores of 130 or greater. Analysis of gender, ethnicity, and risk revealed a greater concentration of non-Caucasian males with at least two risk factors in the underachieving group. Our findings suggested that gifted underachievers are not as motivated or interested in acquiring traditional factual information as high-achievers. Creative teaching strategies are recommended to maximize the talents of underachievers.