Bridgham R G, Rothman A I
Med Educ. 1982 Jul;16(4):219-22. doi: 10.1111/j.1365-2923.1982.tb01252.x.
A practice examination has been prepared annually at the University of Toronto for use by those who will take the Qualifying Examination of the Medical College of Canada. Passing the Qualifying Examination is an important step in gaining the Licentiate of the Medical College of Canada (the LMCC) and thus in becoming licensed. The effect of taking the practice examination on scores on the LMCC exam was investigated. The practice examination is a smaller replica of the LMCC exam, with questions spanning major medical specialties presented in both multiple-choice and patient management problem (PMP) format. It was conjectured that the effect of practice would be greater on the PMP score than on the multiple-choice score. Analysis of covariance was used to compare the LMCC examination scores of senior students, who had taken the practice examination, with the scores of those who had not. The comparison was performed for two graduating classes, separately, and involved the use of a score summarizing each student's performance in the third year of the curriculum as a covariate. In each graduating class, the practice effect on PMP scores is statistically significant. In neither class alone is the practice effect on multiple-choice scores statistically significant, although a combination of results in the two classes is statistically significant. The presence of practice effects can be taken to indicate a flaw in the testing process. The elimination of practice effects, through experiences like the practice examination, would make the entire testing process more valid.
多伦多大学每年都会准备一场模拟考试，供那些准备参加加拿大医学院资格考试的人使用。通过资格考试是获得加拿大医学院执照（LMCC）从而取得行医执照的重要一步。研究了参加模拟考试对LMCC考试成绩的影响。模拟考试是LMCC考试的一个较小的复制品，其问题涵盖主要医学专业，采用选择题和患者管理问题（PMP）两种形式呈现。据推测，模拟考试对PMP成绩的影响比对选择题成绩的影响更大。协方差分析用于比较参加过模拟考试的高年级学生与未参加过模拟考试的学生的LMCC考试成绩。分别对两个毕业班进行了比较，并将每个学生在课程第三年的综合成绩作为协变量。在每个毕业班中，模拟考试对PMP成绩的影响具有统计学意义。虽然单独在任何一个班级中，模拟考试对选择题成绩的影响都没有统计学意义，但两个班级的结果综合起来具有统计学意义。模拟考试效果的存在可以被视为测试过程中的一个缺陷。通过类似模拟考试的经历消除模拟考试效果，将使整个测试过程更加有效。
