Cope M K, Baker H H, Fisk R, Gorby J N, Foster R W
West Virginia School of Osteopathic Medicine, 400 N Lee Street, Lewisburg, WV 24901, USA.
J Am Osteopath Assoc. 2001 Feb;101(2):84-5, 89-90.
To predict student performance on the Comprehensive Osteopathic Medical Licensing Examination (COMLEX-USA) Level 1 examination based on academic performance during the first 2 years, stepwise regression analysis of COMLEX-USA Level 1 performance with preadmission grade point averages, Medical College Admission Test scores, and academic performance was performed on the class of 2000 to develop three formulae that were then used to predict performance on COMLEX-USA Level 1 for the class of 2001. Models ranged in accuracy of predicting the pass/fail status from 95.2% (all available data) to 96.8% (first-year grades and admissions data). A predictive model for student performance on COMLEX-USA Level 1 can be developed and has a high degree of accuracy. The model with the most variables available to choose from predicts the most failures.
为根据前两年的学业成绩预测学生在综合整骨医学执照考试(COMLEX-USA)一级考试中的表现,对2000届学生进行了COMLEX-USA一级考试表现与入学前平均绩点、医学院入学考试成绩及学业成绩的逐步回归分析,以制定三个公式,然后用这些公式预测2001届学生在COMLEX-USA一级考试中的表现。预测通过/未通过状态的模型准确率在95.2%(所有可用数据)至96.8%(一年级成绩和录取数据)之间。可以开发一个预测学生在COMLEX-USA一级考试中表现的模型,且该模型具有很高的准确性。可供选择变量最多的模型预测出的未通过人数最多。