Discrete-state and continuous models of recognition memory: testing core properties under minimal assumptions

Area of Science:

• Cognitive Psychology
• Psychological Models

Background:

• The debate on recognition memory hinges on whether it uses continuous or discrete processes.
• Existing models include signal detection theory (SDT) for continuous and 2-high-threshold for discrete processes.
• Previous comparisons predominantly used receiver operating characteristic (ROC) data.

Purpose of the Study:

• To propose and validate a new, simpler model-comparison approach for recognition memory.
• To evaluate the continuous vs. discrete process debate using this novel method.
• To determine if memory processes are better described by continuous or discrete models.

Main Methods:

• A new model-comparison technique based on ranking judgments was developed.
• This method avoids distributional assumptions and costly experimental manipulations.
• Paired-comparison tests were used, replacing complex goodness-of-fit analyses.

Main Results:

• The proposed ranking judgment method proved effective for model comparison.
• Empirical results from two experiments were analyzed.
• Findings were consistent with a continuous memory process, supporting the SDT model.

Conclusions:

• The ranking judgment approach offers a more accessible and less assumption-laden method for comparing cognitive models.
• Evidence supports continuous memory processes over discrete ones in recognition tasks.
• This study advances the understanding of recognition memory mechanisms.