ABSTRACT

Cognitive diagnosis models, or diagnostic classification models, give conditional probability distributions for features of performance as functions of multiple aspects of persons’ capabilities and of features of task situations. Although they originated under an information-processing psychological perspective, a sociocognitive perspective adds insight into their situated meanings and practical uses. Ideas are illustrated with Tatsuoka’s mixed-number subtraction example, Yamamoto’s hybrid of item response theory and cognitive diagnosis modeling, and the “conditional sense of fairness” from Chapter 9.