Empirical validation of the diffusion model for recognition memory and a comparison of parameter-estimation methods

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Psychometrics

Background:

• The diffusion model is a prominent theory for binary decision tasks, explaining evidence accumulation over time.
• It models both accuracy and response times, using parameters for decision quality and bias.
• Its application in recognition memory requires rigorous validation.

Purpose of the Study:

• To experimentally test the validity of Ratcliff's diffusion model in recognition memory.
• To assess the convergent and discriminant validity of key model parameters.
• To evaluate the performance of analysis software: fast-dm, EZ, and DMAT.

Main Methods:

• Conducted three recognition-memory experiments.
• Employed specific experimental manipulations targeting three central diffusion model parameters.
• Analyzed data using fast-dm, EZ, and DMAT software.

Main Results:

• Experimental manipulations primarily influenced their targeted diffusion model parameters, supporting convergent validity.
• Minor cross-parameter effects were observed, indicating some limitations in discriminant validity.
• The three software packages provided consistent analyses of the model parameters.

Conclusions:

• The study provides strong support for the convergent validity of the diffusion model's parameters in recognition memory.
• While largely valid, the model exhibits some limitations in discriminant validity, suggesting areas for refinement.
• The findings validate the use of fast-dm, EZ, and DMAT for diffusion model analysis.