Computer-based curriculum-based measurement (CBM) is a relatively common practice, but surprisingly few studies have examined the reliability of computer-based CBM. This study sought to examine the reliability of CBM administered via paper/pencil versus the computer. Twenty-one of 25 students in two third-grade classes (N = 21) participated in two generalizability studies. The primary facet of interest, format of assessment, was examined for two measures, a fact operation (fact family measure) and a more complex operation (multidigit addition). Researchers administered four alternate forms of each measure under both computer and paper/pencil conditions. Results indicated ideal reliability of measurement within all conditions and results were consistent across assessment formats for the fact operation. However, assessment format explained a large proportion of variance for the more complex skill (multidigit addition). Results indicate a need to consider the reliability of transitioning assessment from paper/pencil to computer-based response on a measure-by-measure basis.

基于计算机的基于课程的测量 (CBM) 是一种相对普遍的做法，但令人惊讶的是，很少有研究检查基于计算机的 CBM 的可靠性。本研究旨在检验通过纸/铅笔与计算机管理 CBM 的可靠性。两个三年级班的 25 名学生中有 21 名（N= 21) 参加了两项普遍性研究。主要关注的方面，评估格式，检查了两个测量，一个事实运算（事实族测量）和一个更复杂的运算（多位数加法）。研究人员在计算机和纸/笔条件下对每种测量进行了四种替代形式。结果表明，在所有条件下测量的理想可靠性，并且结果在事实操作的评估格式中是一致的。然而，评估格式解释了更复杂技能（多位数加法）的大部分差异。结果表明，需要逐项考虑将评估从纸/笔转换为基于计算机的响应的可靠性。