Integrating cognitive process and descriptive models of attitudes and preferences

Area of Science:

• Cognitive Science
• Decision Science
• Behavioral Economics

Background:

• Discrete choice experiments (DCEs) are increasingly favored for preference measurement over traditional methods like Likert scales.
• Standard analysis of DCE data uses random utility models, which offer good measurement but limited cognitive process insights.

Purpose of the Study:

• To extend a cognitive model, previously successful for simple decisions, to analyze complex, multi-attribute DCE data.
• To integrate response time data with choice data for a deeper understanding of decision-making processes.

Main Methods:

• Applied an established cognitive model to analyze two distinct DCE datasets (dermatology appointments, mobile phones).
• Validated the extended model against standard choice models using data fit and parameter analysis.
• Tested the model on a perceptual judgment task analogous to best-worst DCEs, analyzing both choice and response time data.

Main Results:

• The extended cognitive model demonstrated comparable fits and parameters to standard choice models for complex DCE data.
• The model successfully accounted for both choices and response times in a perceptual judgment task.
• This suggests the cognitive model provides process-level insights into DCEs.

Conclusions:

• Discrete choice experiments are valuable across multiple research fields.
• Accumulator-based cognitive models from response time research can serve as process-level instantiations for random utility models in DCE analysis.
• This integration enhances the understanding of cognitive mechanisms underlying preference formation and choice behavior.