作者: Edward H. Ip
关键词:
摘要: The testlet response model is designed for handling items that are clustered, such as those embedded within the same reading passage. Although a powerful tool item clusters in educational and psychological testing, interpretations of its parameters, conditional correlation between pairs, information function not always clear. For example, a-parameter model, which sometimes referred to discrimination parameter interpreted conventional way, often directly compared standard models. In this article, it argued despite deceptively similar forms a-parameters, interpretation comparison can be misleading. A marginalized version introduced derive quantities comparable This mean (marginalization) approach contrasted with temptingly simple but inapp...