Admission to Accounting Programs: Using a Discriminant Model as a Classification Procedure.
使用判别分析构建会计专业录取政策，基于三个指标（45学分时GPA、大学数学成绩、英语写作成绩）以78%的统计准确率分类学生，并能正确识别90%应被拒绝的学生。
Abstract ABSTRACT: This paper uses discriminant analysis as a foundation for an admissions policy for accounting majors. The discriminant model is capable of correctly classifying students into Accept/Reject groups with a statistical accuracy of 78 percent using only three characteristics--grade-point average at 45 semester hours, grade in college mathematics, and grade in English composition. The model, more importantly, correctly classifies 90 percent of those students who, according to the model's criterion, should be rejected. Alternatively, the model is capable of identifying the same percent of those students who perform exceedingly well in accounting. This technique, together with a review and appeal procedure, provides a method of restricting enrollment while still meeting today's demand for quality accounting students.