C. Bangeranye is assistant professor, Department of Science Education, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Hempstead, New York; ORCID: https://orcid.org/0000-0002-2266-3206. Y.S. Lim is visiting assistant professor, School of Education, University of Cincinnati, Cincinnati, Ohio; ORCID: https://orcid.org/0000-0003-0225-1527.
Acad Med. 2020 Jan;95(1):145-150. doi: 10.1097/ACM.0000000000002954.
Course evaluations by students are a standard tool that U.S. universities use to monitor the quality of their product. Here, the authors examine an alternative method of monitoring instructional quality that differs from traditional approaches in that it does not rely on students' ratings. The authors sought to glean relevant diagnostic information about course effectiveness from in-class exams used to assess students' learning progress (i.e., cognitively diagnostic assessments that explicitly target instructional content).
The authors used data from an end-of-course, cumulative exam given in 2015 and in 2016 to 200 first-year medical students. They mapped the exam questions to 4 attributes and analyzed the students' overall mastery of the content tested and the percentage of students mastering each attribute.
Analyses of the cognitively diagnostic assessment data revealed the percentage of the cohort who achieved/failed to achieve mastery of each of the attributes, discreet mastery profiles that distinguish among students with similar scores, and the percentage of the cohort within each of the 16 attribute mastery profiles. Analysis allowed the authors to evaluate how well the course content was delivered.
Cognitively diagnostic assessments enable in-class tests to appraise which skills specified in the curriculum have/have not been mastered by the students and how many students have mastered/failed to master which particular skills. Hence, if the learning goals have been well defined at the beginning of a course, then cognitively diagnostic assessments can show to what degree the instructional objectives have actually been accomplished.
学生课程评估是美国大学用来监测其教学质量的标准工具。在这里，作者研究了一种监测教学质量的替代方法，该方法与传统方法不同，它不依赖于学生的评价。作者试图从课堂考试中获取有关课程效果的相关诊断信息，这些考试用于评估学生的学习进度（即明确针对教学内容的认知诊断评估）。
作者使用了 2015 年和 2016 年至 200 名一年级医学生的期末综合考试数据。他们将考试问题映射到 4 个属性上，并分析了学生对测试内容的整体掌握程度以及掌握每个属性的学生比例。
对认知诊断评估数据的分析揭示了每个属性中达到/未达到掌握程度的学生比例、区分具有相似分数的学生的离散掌握情况，以及每个属性掌握情况中每个学生的比例。分析使作者能够评估课程内容的交付情况。
认知诊断评估使课堂考试能够评估课程中规定的技能哪些已经被学生掌握/未掌握，以及有多少学生已经掌握/未能掌握哪些特定技能。因此，如果在课程开始时就明确了学习目标，那么认知诊断评估可以显示教学目标实际上达到了什么程度。
