Evaluating the Congruence among Methods of Reading Disability Identification

Abstract

Diagnostic procedures used to identify reading disabilities must meet acceptable psychometric standards and produce groups of subjects who possess known characteristics of the disorder. The issue addressed in this study is whether there are significant differences in the numbers of children identified as reading disabled according to an achievement cut-off method and a regression method of discrepancy analysis. These approaches then are compared to clinic diagnosis and other criteria. The achievement cut-off approach identified a significantly higher number of subjects as reading disabled than did the other methods. This approach also resulted in significant reading-disabled/non-disabled group differences that were not seen when groups were formed using the regression method. The achievement cut-off approach had a lower level of agreement with clinic diagnosis than did the regression method and met fewer of the evaluative criteria than did the regression method. Use of the achievement cut-off approach resulted in the apparent overidentification of reading disabilities in the current sample, thus this approach cannot be recommended for use without more supportive research.