Toward a complete decision model of item and source recognition

Area of Science:

• Cognitive Psychology
• Neuroscience
• Psychology

Background:

• Item recognition and source recognition are distinct memory tasks.
• Receiver operating characteristic (ROC) curves differ for these tasks, suggesting different decision processes.
• A unified model is needed to explain both item and source recognition.

Purpose of the Study:

• To develop a single model explaining both item and source recognition.
• To account for observed differences in ROC curves between the two tasks.
• To test a detection-theoretic model on existing memory data.

Main Methods:

• Postulated a detection-theoretic decision space with item strength and source strength dimensions.
• Developed a model incorporating constant likelihood ratio boundaries, source guessing, and attention allocation.
• Applied the model to three canonical recognition memory data sets.

Main Results:

• The proposed model successfully accounts for data from three independent data sets.
• The model explains ROC curves without invoking recollection-specific processes.
• Observed and predicted ROCs align, validating the model's predictive power.

Conclusions:

• A unified detection-theoretic model can explain both item and source recognition.
• Cognitive processes underlying recognition memory can be described by item and source strength.
• The model provides a parsimonious explanation for recognition memory phenomena.