Litzelman D K, Stratos G A, Marriott D J, Skeff K M
Department of Medicine, Indiana University School of Medicine, USA.
Acad Med. 1998 Jun;73(6):688-95. doi: 10.1097/00001888-199806000-00016.
To examine an instrument for evaluating clinical teaching using factor analysis and to refine the validated instrument to a practical length.
Factor analysis on a split sample of 1,581 student evaluations rating 178 teachers. The instrument was based on the seven-category Stanford Faculty Development Program's (SFDP's) clinical teaching framework and contained 58 Likert-scaled items, with at least seven items per category plus five items measuring "teacher's knowledge." Standard methodology for survey item reduction was used to remove items with low or complex factor loadings and iteratively remove items with low item-scale correlation. Results were replicated on the second sample.
The seven original categories emerged and items originally categorized under "knowledge" statistically combined with "promoting self-directed learning." Over 73% of the variance was explained. Item reduction resulted in 25 items with overall internal consistency over .97 and internal consistency of constructs ranging from .82 to .95.
Factor analysis of student ratings validated the seven-category SFDP framework. An abbreviated instrument to measure the seven categories is described. Results suggest that students may not systematically distinguish between their teachers' knowledge and their teachers' ability to promote self-directed learning, an important finding for both administrators and faculty development programs.
运用因子分析检验一种用于评估临床教学的工具，并将经过验证的工具精简至实用长度。
对1581名学生对178名教师的评价进行拆分样本的因子分析。该工具基于七类斯坦福教师发展项目（SFDP）的临床教学框架，包含58个李克特量表项目，每类至少有七个项目，外加五个测量“教师知识”的项目。采用标准的调查项目缩减方法，去除因子载荷低或复杂的项目，并迭代去除项目与量表相关性低的项目。结果在第二个样本上进行了重复验证。
七个原始类别出现，原本归类在“知识”下的项目在统计上与“促进自主学习”合并。超过73%的方差得到了解释。项目缩减后得到25个项目，总体内部一致性超过0.97，各结构的内部一致性在0.82至0.95之间。
学生评价的因子分析验证了七类SFDP框架。描述了一种用于测量这七类的简化工具。结果表明，学生可能无法系统地区分教师的知识和教师促进自主学习的能力，这一发现对管理人员和教师发展项目都具有重要意义。
