Semantic Memory Retrieval: Analysis by Speed Accuracy Tradeoff Functions

Abstract

Semantic memory retrieval for verifying category—example associations was tested by a speed accuracy tradeoff method: present the category for 2 s, present a correct or incorrect example followed, after a variable lag (0, 0.1, 0.2, 0.3, 0.4, 0.6, 0.8, 1, 2, or 3 s), by a signal to make a “yes-no” response in about 0.2 s. Although the strength of the category—example association is higher for high dominance examples of a category, retrieval dynamics did not vary with dominance level. Recognition for category—example associations appears to be a direct-access (parallel) retrieval process. Priming a category by repeated testing of the same category over three consecutive trials had no effect on either asymptotic strength or retrieval dynamics. Partitioning into short, medium, and long latency responses at each lag produced microtradeoff functions which did not lie on the same macrotradeoff function. Retrieval dynamics were invariant with long-term practice.