Response bias correction in the process dissociation procedure: approaches, assumptions, and evaluation

Area of Science:

• Cognitive Psychology
• Psychometric Methods
• Multinomial Processing Tree Models

Background:

• The process dissociation procedure (PDP) is a key method for estimating cognitive processes.
• Previous expositions of PDP within multinomial modeling have obscured critical process overlaps.
• Understanding these overlaps is essential for accurate interpretation of cognitive data.

Purpose of the Study:

• To propose a general classification of response bias corrections for the PDP.
• To clarify the assumptions underlying existing response bias correction models.
• To illustrate a framework for classifying corrective models based on process interactions.

Main Methods:

• Developed a general classification scheme for response bias corrections to the PDP.
• Derived three specific corrective models within this new classification framework.
• Applied and evaluated these derived models using existing empirical data.

Main Results:

• The proposed classification scheme effectively clarifies assumptions of previous correction methods.
• The derived corrective models offer a more transparent approach to addressing response bias in PDP.
• Evaluation using Buchner et al. (1995) data demonstrates the utility of the framework.

Conclusions:

• Clarifying process interactions in PDP is crucial for accurate modeling and interpretation.
• The proposed classification provides a structured approach to understanding and developing response bias corrections.
• This framework enhances the application of multinomial modeling in cognitive research.