Mielke Ina, Breil Simon M, Hissbach Johanna, Ehrhardt Maren, Knorr Mirjana
Department of Biochemistry and Molecular Cell Biology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany.
Department of Psychology, University of Münster, Münster, Germany.
Adv Health Sci Educ Theory Pract. 2024 Sep 30. doi: 10.1007/s10459-024-10379-3.
Situational Judgement Tests (SJTs) are popular to screen for social skills during undergraduate medical admission as they have been shown to predict relevant study outcomes. Two different types of SJTs can be distinguished: Traditional SJTs, which measure general effective behavior, and construct-driven SJTs which are designed to measure specific constructs. To date, there has been no comparison of the predictive validity of these two types of SJTs in medical admission. With the present research, we examine whether the HAM-SJT, a traditional SJT, and the CD-SJT, a construct-driven SJT with an agentic and a communal scale, administered during undergraduate medical admission can predict OSCE (i.e., objective structured clinical examination) performance in a low-stakes (n = 159) and a high-stakes (n = 160) sample of medical students. Results showed a moderate positive relation between the communal scale of the CD-SJT and performance in OSCE stations with trained patients in the high-stakes sample (r =.20, p =.009). This SJT had also an incremental value in predicting the OSCE performance above and beyond GPA (i.e., grade point average), a science test (i.e., HAM-Nat), and gender (ß = 0.18, 95% CI [0.03; 0.33], p =.020). That is, individuals who chose more communal behavioral responses in the SJT were rated more favorably in interactions with trained patients in the OSCE. A comparable correlation coefficient was observed for the HAM-SJT when controlling for range restriction due to admission (r = 0.14 vs. r = 0.20). Our research provides a first indication for the predictive validity of construct-driven SJTs in high-stakes undergraduate medical admission.
情境判断测试（SJTs）在本科医学入学时用于筛选社交技能很受欢迎，因为它们已被证明能预测相关的学习成果。可以区分两种不同类型的SJTs：传统SJTs，用于测量一般有效行为；以及结构驱动的SJTs，旨在测量特定结构。迄今为止，尚未对这两种类型的SJTs在医学入学中的预测效度进行比较。通过本研究，我们考察了在本科医学入学期间进行的传统SJT即HAM-SJT，以及具有能动和社群量表的结构驱动SJT即CD-SJT，能否在低风险（n = 159）和高风险（n = 160）的医学生样本中预测客观结构化临床考试（OSCE）成绩。结果显示，在高风险样本中，CD-SJT的社群量表与在有培训患者参与的OSCE考站中的表现之间存在中等程度的正相关（r =.20，p =.009）。该SJT在预测OSCE成绩方面，相对于平均绩点（GPA）、科学测试（即HAM-Nat）和性别，也具有增量价值（ß = 0.18，95%置信区间[0.03；0.33]，p =.020）。也就是说，在SJT中选择更多社群行为反应的个体，在OSCE中与培训患者互动时的评分更有利。在控制因入学导致的范围限制时，观察到HAM-SJT的相关系数与之相当（r = 0.14 vs. r = 0.20）。我们的研究首次表明了结构驱动的SJTs在高风险本科医学入学中的预测效度。
