EFFECTS OF SIMPLE AND REGRESSED DISCREPANCY MODELS AND CUTOFFS ON SEVERE DISCREPANCY DETERMINATION

Concern over the growing number of students identified as learning dis abled has led school districts to examine the criteria for determinati on and the means by which they are operationalized. Two frequently use d methods for determining a severe discrepancy between ability and ach ievement, a key criterion in LD determination, were applied to a sampl e of 344 students to determine how a change in method might influence the rates and characteristics of students meeting this criterion. The results indicate an increase in numbers when a regression method is us ed over a simple difference score method. When the change proposed inc luded moving to a more severe cutoff score, the pattern reversed and a 20% decrease was observed. Although IQ correlated with the discrepanc ies when the simple difference score method was used, no correlation w as observed when regression was employed, indicating that regression m ay be a more equitable method for calculating severe discrepancies. Ne ither method resulted in disproportionate racial representation among those meeting the severe discrepancy criterion.