A multinomial modeling analysis of the recognition-failure paradigm

Area of Science:

• Cognitive Psychology
• Memory Research
• Computational Neuroscience

Background:

• The recognition-failure paradigm is a key area in memory research.
• The Tulving-Wiseman function is a prominent theoretical model within this paradigm.
• Previous studies have noted exceptions to the Tulving-Wiseman function, prompting further investigation.

Purpose of the Study:

• To critically evaluate the explanatory power of the Tulving-Wiseman function using empirical data.
• To develop and apply a novel computational model for analyzing memory storage and retrieval processes.
• To investigate the conditions under which the Tulving-Wiseman function fails and propose an alternative explanation.

Main Methods:

• Statistical model comparison assessing the fit of the Tulving-Wiseman function to recognition-failure data.
• Development of a multinomial model grounded in retrieval-independence theory.
• Meta-analysis of existing recognition-failure studies using the developed computational model.

Main Results:

• The Tulving-Wiseman function demonstrated a poor fit to the data, irrespective of the statistical metric used.
• Violations of the Tulving-Wiseman function were systematically linked to conditions characterized by weak memory storage and strong retrieval.
• A theoretically derived alternative equation, assuming a positive correlation between storage and retrieval, matched the data nearly as well as the original function.

Conclusions:

• The Tulving-Wiseman function is inadequate for fully explaining recognition failure phenomena.
• Retrieval-independence theory provides a robust framework for modeling memory processes in recognition failure.
• The findings suggest that the interplay between storage strength and retrieval efficiency is crucial for understanding memory performance.