Knowledge of solution strategies and IRT modeling of items for transitive reasoning

Componential item response theory (CIRT) is presented as a model-oriented a pproach to studying processes and strategies underlying the incorrect/corre ct responses to cognitive test tasks. CIRT is contrasted with a data-orient ed approach in which verbal explanations for incorrect/correct responses ar e collected during the test phase and incorporated in the scoring. Alternat ively, the psychologically meaningful data are modeled by unidimensional it em response theory models. Verbal explanations for each examinee and task w ere collected from transitive reasoning tasks in addition to the incorrect/ correct responses. Two datasets were compiled, one reflecting the common in correct/correct scoring and one showing whether a deductive strategy had be en used to produce a correct response. The Mokken model of monotone homogen eity, the partial-credit model, and the generalized one-parameter logistic model were used to analyze both polytomous datasets. Results showed that co mbining knowledge of solution strategies with IRT modeling produced a usefu l unidimensional scale for transitive reasoning.