Department of Anatomy, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands.
Anat Sci Educ. 2013 Jan-Feb;6(1):29-41. doi: 10.1002/ase.1290. Epub 2012 Jun 6.
Anatomists often use images in assessments and examinations. This study aims to investigate the influence of different types of images on item difficulty and item discrimination in written assessments. A total of 210 of 460 students volunteered for an extra assessment in a gross anatomy course. This assessment contained 39 test items grouped in seven themes. The answer format alternated per theme and was either a labeled image or an answer list, resulting in two versions containing both images and answer lists. Subjects were randomly assigned to one version. Answer formats were compared through item scores. Both examinations had similar overall difficulty and reliability. Two cross-sectional images resulted in greater item difficulty and item discrimination, compared to an answer list. A schematic image of fetal circulation led to decreased item difficulty and item discrimination. Three images showed variable effects. These results show that effects on assessment scores are dependent on the type of image used. Results from the two cross-sectional images suggest an extra ability is being tested. Data from a scheme of fetal circulation suggest a cueing effect. Variable effects from other images indicate that a context-dependent interaction takes place with the content of questions. The conclusion is that item difficulty and item discrimination can be affected when images are used instead of answer lists; thus, the use of images as a response format has potential implications for the validity of test items.
解剖学家在评估和考试中经常使用图像。本研究旨在调查不同类型的图像对书面评估中项目难度和项目区分度的影响。共有 460 名学生中的 210 名自愿参加了一门大体解剖学课程的额外评估。该评估包含 39 个测试项目，分为七个主题。答案格式按主题交替，要么是标记图像，要么是答案列表，从而产生两个包含图像和答案列表的版本。受试者被随机分配到一个版本。通过项目得分比较答案格式。两次考试的整体难度和可靠性相似。与答案列表相比，两个横截面图像导致项目难度和项目区分度增加。胎儿循环的示意图像导致项目难度和项目区分度降低。三张图像显示了不同的效果。这些结果表明，对评估分数的影响取决于所使用的图像类型。来自两个横截面图像的结果表明正在测试额外的能力。来自胎儿循环图的数据表明存在提示效应。其他图像的可变效果表明，与问题内容发生了上下文相关的相互作用。结论是，当使用图像代替答案列表时，项目难度和项目区分度可能会受到影响；因此，将图像用作回答格式可能会对测试项目的有效性产生影响。
