No abstract

No abstract

Only a few studies have investigated experimenter bias in intelligence testing. Larrabee and Kleinsasser (Note 1) found significant expectancy effects on verbal Wechsler Intelligence Scale for Children (WISC) subtests with sixth graders. Masling (1959) found a significant difference between the scores attributed to memorized responses given in a cold or a warm manner. In these studies, there is no way to distinguish between bias in administration and bias in scoring. In fact, Masling's experimenters were found to behave differently toward the cold and the warm subjects. In this study only bias in scoring, controlling for the effect of actual administration, was investigated.The subjects were 18 graduate students who were asked to score a WISC record of an Israeli fifth grader. The same record was given to all subjects with one of two cover sheets portraying either an underachieving disadvantaged child or a high-achieving, upper-middle class child. Of the subtests used in the standardized Hebrew version of the WISC, we included Information, Comprehension, Vocabulary, and Coding. We did not include Digit Span and Arithmetic, which are usually directly scored during administration.The means of the high-and low-expectancy groups were compared for each subtest, and also for the sum of scores of the three verbal tests. In all comparisons the mean of the high-expectancy group was higher than that of the lowexpectancy group, and the sign test for the five comparisons between pairs was statistically significant (p = .031). Significant differences were found for Comprehension (3-point difference, t = 3.59, p < .005) and for the total verbal sum of scores (4.22-point difference, t = 1.94, p < .05). (One-tailed tests were used, since the ex-Reprints and an extended report of this study may be obtained without charge from Elisha Y.

Not suitable.