Sequential sampling models in memory

In the domain of memory research, there are strong relationships between the accuracy and response times (RT) for choices in tasks such as recognition memory and recall. The strongest unification of these variables to come out of cognitive psychology is in the sequential sampling model framework, in which evidence accumulates in a noisy fashion until a threshold for a given response is reached. In this chapter, we provide an overview of these models within the memory domain. We begin with the classical diffusion model along with the motivations for the "full" diffusion model which includes variability in drift rates and starting points and describe how such models can produce interpretable parameter estimates for testing psychological theory. In addition, we summarize the connections between diffusion models and signal detection theory, extensions to confidence, recall, and continuous responses, and how such models can serve as back-end decision models for theories of retrieval.