On an Application of Statistical Ranking and Selection Tables to the Identification of the Learning Disabled

Abstract

The statistical consequences of giving a student more than one achievement test to determine whether he or she is learning disabled (LD) is discussed. The standard rule is based on one achievement (and one intelligence) test so that if several of the former (say,b) are taken and the lowest score is used, there will be a discrepancy between the proportion of students that are actually labeled LD and the proportion we would originally expect. Ranking and selection tables are used in this paper in a dual manner: (a) to find the expected proportion of LD students as a function of the cut-off point for different values ofb and different correlation structures, and (b) to find the appropriate cut-off point as a function of the given expected proportion of LD students. The appropriate table for problem (a) is included.