A model for screening and classifying potential accounting majors

This study is concerned with developing an explicit methodology with which to screen potential admission into Bowling Green State University's (BGSU hereafter) accounting program. Rather than predicting success in a specific accounting course, the researchers predict success in the accounting program at BGSU. Discriminant analysis was conducted in conjunction with a split sample approach, wherein the model was specified and the parameters estimated on part of the split sample. Then this model was utilized to classify student success or failure on the remaining part of the split sample. The classification results were promising with 85.2% of the students correctly classified.