General Cognitive Diagnosis Model for Response Time

This study proposes a general cognitive diagnosis model for response times (CDM-RT). The idea of the model is to consider person parameters defined on fine-grained attributes to improve the model fit and skill diagnosis related to RTs. The model framework is specified under two distinct motivation conditions and two treatments of attribute profiles. By comparing the attribute profile estimated from CDM for RA and CDM for RT, we can investigate students’ problem-solving behavior and detect possible abnormal patterns. Both simulation study and real data analysis are conducted via a fully Bayesian approach with Markov Chain Monte Carlo (MCMC) metho