ABSTRACT

This article compares cut scores from two variations of the Hofstee and Beuk methods, which determine cut scores by resolving inconsistencies in panelists’ judgments about cut scores and pass rates, with the Angoff method. The first variation uses responses to the Hofstee and Beuk percentage correct and pass rate questions to calculate cut scores. The second variation uses Angoff ratings to determine percentage correct data in combination with responses to pass rate questions. Analysis of data from 15 standard settings suggested that the Hofstee and Beuk methods yielded similar cut scores, and that cut scores were about 2% lower when using Angoff ratings. The two approaches also differed in the weight assigned to cut score judgments in the Beuk method and in the occurrence of undefined cut scores in the Hofstee method. Findings also indicated that the Hofstee and Beuk methods often produced higher cut scores and lower pass rates than the Angoff method. It is suggested that attention needs to be paid to the strategy used to estimate Hofstee and Beuk cut scores.

摘要

本文比较了Hofstee和Beuk方法的两个变体中的得分，后者通过使用Angoff方法解决专家对得分和通过率的判断中的不一致来确定得分。第一个变体使用对Hofstee和Beuk百分比正确率的回答以及通过率问题来计算切分。第二种变体使用Angoff评分与通过率问题的答案相结合来确定正确百分比的数据。对来自15个标准设置的数据的分析表明，霍夫斯提（Hofstee）和贝克（Beuk）方法得出的割伤得分相似，使用Angoff评分时割伤得分大约降低2％。这两种方法在Beuk方法中分配给得分得分判断的权重以及在Hofstee方法中出现不确定的得分得分方面也有所不同。研究结果还表明，与Angoff方法相比，Hofstee和Beuk方法通常产生更高的切分和更低的通过率。建议需要注意用于估计Hofstee和Beuk切割得分的策略。