Examiner scoring of ambiguous WISC-R responses

Fabricated test protocols were used to study how effectively examiners agree in scoring ambiguous WISC-R responses. Clinical and school psychologists (N = 62) and graduate students (N = 48) scored WISC-R Similarities, Comprehension, and Vocabulary responses obtained in a group administration procedure. From over 15,000 responses obtained, only unusual, atypical, or ambiguous responses were selected; 726 responses were scored, with 11 groups of 10 raters each scoring 66 responses. Considerable scoring disagreement occurred. Unanimous agreement (within each group of 10 raters) was found on only 13% of the responses, while 80% of the raters agreed on 44% of the responses. Rates of scoring agreement were not related to level of rater experience, but were related to specific subtests. Higher rates of agreement were found on the Similarities and Comprehension subtests than on the Vocabulary subtest for both experienced and inexperienced raters. The results suggest that even with the improved WISC-R manual, scoring remains a difficult and challenging task.