CRITERION-RELATED BIAS WITH THE GUIDE TO THE ASSESSMENT OF TEST-SESSION BEHAVIOR FOR THE WISC-III AND WIAT - POSSIBLE RACE ETHNICITY, GENDER, AND SES EFFECTS

Using the Guide to the Assessment of Test-Session Behavior for the WIS C-III and WIAT (GATSB), Anglo examiners recorded test observations for 969 children between the ages of 6 and 16 years. The children came fr om the standardization and validity-study samples of GATSB ratings com pleted with the Wechsler Intelligence Scale-Third Edition (WISC-III) ( Wechsler, 1991). The sample differed by race (Anglo, black, Latino), s ocioeconomic status (SES) (high, middle, low), and gender. GATSB ratin gs and WISC-III Verbal, Performance, and Full Scale IQs were compared. Correlations between the GATSB and WISC-III were generally moderate ( average r = -.27). Children who exhibited higher levels of avoidance, inattentiveness, and uncooperative behaviors while being tested tended to exhibit lower WISC-III scores. This pattern held true for Anglos, blacks, and Latinos; for girls and boys; and for those from high-, mid dle-, and low-SES homes. Evidence was generally absent that Anglo exam iners display bias in black-Anglo, gender, or SES comparisons. However , consistent differences were noted between Latino and Anglo children. Examiners tended to rate Latinos as displaying better test behaviors than Anglos when children's IQs were below average, but comparable whe n IQs were average and above. Thus, the GATSB generally displayed simi lar intrasession validities for children who differ by race/ethnicity, gender, and SES.