Trail Carla, Reiter Harold I, Bridge Michelle, Stefanowska Patricia, Schmuck Marylou, Norman Geoff
McMaster University, Hamilton, ON, Canada.
Adv Health Sci Educ Theory Pract. 2008 Aug;13(3):253-61. doi: 10.1007/s10459-006-9037-z. Epub 2006 Oct 25.
A consistent finding from many reviews is that undergraduate Grade Point Average (uGPA) is a key predictor of academic success in medical school. Curiously, while uGPA has established predictive validity, little is known about its reliability. For a variety of reasons, medical schools use different weighting schemas to combine years of study. Additional concerns relate to the equivalence of grades obtained from different fields of study and institutions, with little hard data to guide conclusions. At the Michael G. DeGroote School of Medicine Class of 2007 at McMaster University, every undergraduate grade of 2,138 applicants, along with field of study and post-secondary educational institution, was analyzed. Individual grades were aggregated into an overall uGPA using published algorithms from several medical school, and correlated with a non-weighted sum. Correlations of the different schemas with equal weights ranged from 0.973 to 0.990. The extent of the difference between fields of study was small, accounting for only 1.5% of the variance. However, differences among 16 Ontario universities were larger, and accounted for 9.3% of the variance. The results of this study suggest that all weighting schemas are virtually equivalent, making any formulation reasonable. Differences by field of study are small, but do not show any bias against non-science students. Differences by institution are larger, amounting to a range in average score from 78.7 to 84.6; however it is not clear whether this reflects candidate ability or institutional policy, so attempts to correct for institution may be difficult.
许多综述一致发现，本科平均绩点（uGPA）是医学院学业成功的关键预测指标。奇怪的是，虽然uGPA已确立了预测效度，但其可靠性却鲜为人知。由于种种原因，医学院使用不同的加权模式来综合多年的学习成绩。其他问题涉及不同研究领域和院校所获成绩的等效性，几乎没有确凿数据可指导得出结论。在麦克马斯特大学迈克尔·G·德格罗特医学院2007届班级中，分析了2138名申请者的每一项本科成绩，以及学习领域和高等教育机构。使用几所医学院公布的算法将个人成绩汇总为总体uGPA，并与未加权总和进行关联。等权重的不同模式之间的相关性在0.973至0.990之间。不同学习领域之间的差异程度较小，仅占方差的1.5%。然而，安大略省16所大学之间的差异较大，占方差的9.3%。这项研究的结果表明，所有加权模式实际上都是等效的，任何一种公式都是合理的。不同学习领域的差异较小，但并未显示出对非理科学生的任何偏见。不同院校之间的差异较大，平均分数范围为78.7至84.6；然而，尚不清楚这是反映了考生能力还是院校政策，因此试图校正院校因素可能会很困难。
