Differences among students with or without severe discrepancy between predicted and actual achievement: Does severe discrepancy matter?

Abstract

This study investigated the differences between students with severe discrepancy and those without according to the Reynolds ( 1984) regression-based discrepancy formula. The sample included 264 students who were referred for an initial special education evaluation of !earning problems. Differences between these two groups were examined across gender, age, cognitive variables, and achievement. Results revealed that no significant gender or age differences were found in the presence of a severe discrepancy. However, statistically significant negative correlations were found between Full Scale IQ, and all of the differences between predicted and actual achievement scores with small to medium effect sizes (-1 to -.3). Finally, results from independent tests showed that students without severe discrepencies performed statistically significantly better on all tests of academic achievement than students with severe discrepancies, producing large effect sizes (d) ranging from .79 to 1.4. These results supported research conducted by Kavale (2002) where he found that severely discrepant students scored consistely lower on measures of achievement than non-discrepant students.

Date of Award	2008
Original language	American English
Awarding Institution	Eastern Illinois University
Supervisor	Gary Canivez (Supervisor)