Short-term memory scanning viewed as exemplar-based categorization.

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Human Memory

Background:

• Exemplar-similarity models, like the exemplar-based random walk (EBRW), formally account for multidimensional classification probabilities and response times (RTs).
• Exemplar models have been used to explain old-new item recognition and the relationship between classification and recognition.
• A gap exists in testing these models on RTs and performance within the Sternberg short-term memory-scanning paradigm.

Purpose of the Study:

• To evaluate the exemplar-based random walk (EBRW) model's ability to account for performance in the Sternberg short-term memory-scanning task.
• To bridge the gap in research by applying exemplar models to a classic recognition-RT paradigm.
• To explore potential shared cognitive operating principles between multidimensional classification and short-term recognition memory.

Main Methods:

• Application of the exemplar-based random walk (EBRW) model.
• Analysis of response times (RTs) and other performance measures.
• Testing the model within the Sternberg (1966) short-term memory-scanning paradigm.

Main Results:

• The EBRW model naturally accounts for a wide range of phenomena in short-term memory scanning.
• The model demonstrates efficacy in explaining RTs and performance in the Sternberg task.
• Evidence supports the EBRW model's applicability beyond its original scope of multidimensional classification.

Conclusions:

• The exemplar-based random walk (EBRW) model provides a robust theoretical account for short-term memory scanning.
• Similar cognitive operating principles may govern both multidimensional classification and short-term old-new recognition.
• This research validates the EBRW model's utility across different memory and cognitive domains.