Early quantitative models of recognition memory adopt either ‘threshold’ or ‘signal detection’ assumptions. However, recent evidence from a wide variety of behavioral and neuroscientific studies has led most researchers to reject both classes of models in favor of ‘hybrid’ models—such as the Dual Process Signal Detection (DPSD) model—that combine the strengths of both approaches.
Many current models of recognition assume two distinct underlying memory processes such as recollection and familiarity (or item vs source, or item vs associative retrieval). However, various competing dual process models exist ranging from simple 2-parameter measurement models to more complex neurocomputational models of the medial temporal lobes and surrounding cortex. These models make a variety of testable predictions, and empirical work is beginning to differentiate amongst these models.