ERIC Number: ED477458
Record Type: Non-Journal
Publication Date: 2003-Apr
Reference Count: N/A
Consequential Validity Impact of Choosing Different Aptitude-Achievement Discrepancy Models in Identifying Students with Learning Disabilities.
Glasnapp, Douglas R.; Poggio, John P.
This study used computer simulation to provide information on the percentage of students with learning disabilities expected to be identified under different aptitude-achievement discrepancy eligibility models and criteria and to demonstrate the consequential effects in terms of the extent to which the different models identify students of different ability levels. The two primary models of concern were the regression discrepancy model and the straight discrepancy model. In addition, the true score discrepancy model was included in the comparisons. Data with predetermined parameters were generated to simulate conditions mirroring those found in actual practice. Four base data sets for samples of 10,000 cases each were simulated. Findings demonstrate the consequences that result when one of the models is chosen as the preferred model for inclusion in procedures for identifying and qualifying students for services for learning disabled students. Results of the simulation studies indicate that the inflation of qualification rates can range from 12% to 31% when using the straight discrepancy model versus the regression discrepancy model, depending on the score reliabilities for the two measures and the extent of their correlation. (Contains 10 figures, 14 tables, and 6 references.) (Author/SLD)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: N/A