Fallatah Hind I, Tekian Ara, Park Yoon Soo, Al Shawa Lana
Faculty of Medicine, King Abdulaziz University, Jeddah, Saudi Arabia.
Department of Medical Education, College of Medicine, University of Illinois, Chicago, IL, USA.
BMC Med Educ. 2015 Feb 1;15:10. doi: 10.1186/s12909-015-0295-4.
Exams are essential components of medical students' knowledge and skill assessment during their clinical years of study. The paper provides a retrospective analysis of validity evidence for the internal medicine component of the written and clinical exams administered in 2012 and 2013 at King Abdulaziz University's Faculty of Medicine.
Students' scores for the clinical and written exams were obtained. Four faculty members (two senior members and two junior members) were asked to rate the exam questions, including MCQs and OSCEs, for evidence of content validity using a rating scale of 1-5 for each item. Cronbach's alpha was used to measure the internal consistency reliability. Correlations were used to examine the associations between different forms of assessment and groups of students.
A total of 824 students completed the internal medicine course and took the exam. The numbers of rated questions were 320 and 46 for the MCQ and OSCE, respectively. Significant correlations were found between the MCQ section, the OSCE section, and the continuous assessment marks, which include 20 long-case presentations during the course; participation in daily rounds, clinical sessions and tutorials; the performance of simple procedures, such as IV cannulation and ABG extraction; and the student log book. Although the OSCE exam was reliable for the two groups that had taken the final clinical OSCE, the clinical long- and short-case exams were not reliable across the two groups that had taken the oral clinical exams. The correlation analysis showed a significant linear association between the raters with respect to evidence of content validity for both the MCQ and OSCE, r = .219 P < .001 and r = .678 P < .001, respectively, and r = .241 P < .001 and r = .368 P = .023 for the internal structure validity, respectively. Reliability measured using Cronbach's alpha was greater for assessments administered in 2013.
The pattern of relationships between the MCQ and OSCE scores provides evidence of the validity of these measures for use in the evaluation of knowledge and clinical skills in internal medicine. The OSCE exam is more reliable than the short- and long-case clinical exams and requires less effort on the part of examiners and patients.
考试是医学生临床学习阶段知识和技能评估的重要组成部分。本文对2012年和2013年阿卜杜勒阿齐兹国王大学医学院进行的笔试和临床考试内科部分的效度证据进行了回顾性分析。
获取学生临床和笔试成绩。邀请四名教员（两名资深教员和两名初级教员）使用1至5分的评分量表对考试题目（包括选择题和客观结构化临床考试题目）进行评分，以获取内容效度证据。使用克朗巴哈系数来测量内部一致性信度。采用相关性分析来检验不同评估形式与学生群体之间的关联。
共有824名学生完成内科课程并参加考试。选择题和客观结构化临床考试的被评分题目数量分别为320道和46道。在选择题部分、客观结构化临床考试部分与连续评估成绩之间发现了显著相关性，连续评估成绩包括课程期间的20次长病例报告；参与日常查房、临床课程和辅导课；进行简单操作（如静脉插管和动脉血气抽取）的表现；以及学生日志。尽管客观结构化临床考试对参加最终临床客观结构化临床考试的两组学生来说是可靠的，但临床长病例和短病例考试对参加口试临床考试的两组学生来说并不可靠。相关性分析显示，评分者之间在选择题和客观结构化临床考试的内容效度证据方面存在显著线性关联，r分别为0.219，P < 0.001和r为0.678，P < 0.001，在内部结构效度方面，r分别为0.241，P < 0.001和r为0.368，P = 0.023。2013年进行的评估使用克朗巴哈系数测量的信度更高。
选择题和客观结构化临床考试成绩之间的关系模式为这些测量方法用于评估内科知识和临床技能的效度提供了证据。客观结构化临床考试比临床长病例和短病例考试更可靠，并且对考官和患者来说所需精力更少。
