One-Parameter Dynamic Choice Measurement Model for Process Data

• In Event: Response Process Data Analysis

Abstract

In this study, we develop a one-parameter dynamic choice (1PDC) measurement model for process data analysis, based on the idea of the action choice submodel in the continuous-time dynamic choice measurement model (Chen, 2020). In the 1PDC model, the predefined effectiveness of the action options links individual latent ability with each action choice during the problem-solving process. Additionally, each problem state in the task is distinguished by the corresponding easiness parameter that is freely estimated. The simulation study shows that, compared with Chen’s (2020) model, the 1PDC model could provide similar or much more accurate parameter estimation when easiness parameters of problem states were equal or unequal, respectively. The empirical example further demonstrates the rationality of the 1PDC model.

Authors

• Yue Xiao, Beijing Normal University

• Hongyun Liu