Cognitive psychology meets psychometric theory: on the relation between process models for decision making and latent...

Area of Science:

• Cognitive Psychology
• Psychometrics
• Latent Variable Modeling

Background:

• Diffusion models for response processes can imply item response theory (IRT) models.
• Existing IRT models may not be suitable for all types of psychological measurement, particularly ability tests.

Purpose of the Study:

• To analyze latent variable models from a cognitive psychology viewpoint.
• To investigate the appropriateness of IRT for measuring abilities versus bipolar traits.
• To propose a new IRT model for accuracy and speed based on a reconsidered concept of ability.

Main Methods:

• Analysis of diffusion models and their relationship to 2-parameter logistic IRT models.
• Cognitive modeling of response processes.
• Development of a novel item response model incorporating accuracy and speed.

Main Results:

• Diffusion models for response processes imply IRT models for individual differences.
• Surprisingly, IRT models are appropriate for bipolar traits but not ability tests when diffusion models underlie response processes.
• A new IRT model is developed for accuracy and speed, assuming ability has a natural zero point.

Conclusions:

• The findings challenge traditional assumptions about ability measurement in IRT.
• The new model offers novel perspectives on guessing, response speed, and person fit.
• This work has significant implications for psychological measurement and psychometric theory.