A classification model for continuous responses: Identifying risk perception groups on health-related activities

Area of Science:

• Psychometrics
• Educational Measurement
• Statistical Modeling

Background:

• Cognitive diagnostic models (CDMs) traditionally use dichotomous or polytomous responses.
• The application of CDMs to continuous response data is an emerging area of research.
• Existing CDMs lack a Bayesian framework for analyzing continuous response data.

Purpose of the Study:

• To develop the first Bayesian framework for the continuous deterministic inputs, noisy, and gate (DINA) model.
• To propose novel interpretations for item parameters within the continuous DINA model for enhanced analytical clarity.
• To evaluate the performance of the continuous DINA model using a Bayesian approach through simulations and real-world data.

Main Methods:

• Development of a novel Bayesian framework for the continuous DINA model.
• Introduction of new, more interpretable item parameter definitions for the continuous DINA model.
• Conducting simulation studies to assess the model's performance.
• Application of the continuous DINA model to a real dataset on health-related risk perceptions.

Main Results:

• The proposed Bayesian framework successfully implements the continuous DINA model.
• New item parameter interpretations enhance the model's analytical interpretability.
• Simulation results demonstrate the effectiveness of the Bayesian approach for continuous DINA models.
• The model effectively classified individuals in a real-world risk perception study.

Conclusions:

• The developed Bayesian framework represents a significant advancement for continuous cognitive diagnostic models.
• The continuous DINA model, with its enhanced interpretability, offers a promising tool for analyzing complex response data.
• The model demonstrates practical utility in classifying individuals, particularly in health-related risk perception research.