Consistency among School Psychologists in Evaluating Discrepancy Scores: A Preliminary Study

Abstract

The purpose of this study was to investigate whether school psychologists possess the psychometric knowledge necessary to evaluate discrepancy scores. Experienced school psychologists were asked to evaluate two discrepancy examples and to rate the usefulness of various test measures for different purposes. Despite knowing that (a) standard scores were the scores of choice for discrepancy analysis, (b) age and grade equivalents offered little usefulness, and (c) standard error of measurement was important, respondents made inconsistent choices in evaluating the discrepancy examples. These findings are explained by the failure of a substantial number of subjects to engage in standard statistical procedures for testing differences between scores. Participants also tended to misinterpret percentile ranks as useful for discrepancy analysis. The implications of these findings for learning disabilities classification and research are discussed.