Diagnosis and Predictability of Learning Disabilities

Abstract

Two studies were completed to determine predictability of group membership in learning disability and normally-achieving categories with a psycho-educational test battery measuring intelligence, achievement, visualmotor perception and personality constructs. Study 1 included upper elementary students in grades four to six (controls = 211, learning disabled (LD) = 25). Study 2 included grade three children (controls = 73, LD = 70). Several additional psycho-educational measures were added in Study 2 following the analysis of results in Study 1. A linear regression analysis revealed that, in Study 1, only 35 per cent of the total variance was accounted for in predicting group membership (LD vs controls). In Study 2, however, variance estimates increased to 77 per cent when new data were entered into the equation. The findings support the clause of discrepancy in identifying LD students and agree with previous research findings that tests used for a differential diagnosis of learning disabilities provide little evidence for prediction of group membership. Further study is recommended to examine the heterogeneous nature of learning disabilities.