Roger Ratcliff is an American cognitive psychologist whose diffusion model, first published in 1978, has become the dominant model for analyzing response time and accuracy data from two-choice decision tasks. The model's ability to simultaneously account for the shapes of correct and error RT distributions, the relationship between speed and accuracy, and individual differences has made it indispensable in experimental psychology and cognitive neuroscience.
The Diffusion Model
Ratcliff's key insight was to model the decision process as a continuous random walk (Wiener diffusion process) between two absorbing boundaries. The model's parameters have clear psychological interpretations: drift rate reflects evidence quality, boundary separation reflects caution, starting point reflects bias, and non-decision time reflects encoding and motor processes. Through careful mathematical analysis and extensive empirical testing, Ratcliff demonstrated that these parameters can be uniquely recovered from RT distribution data.
Ratcliff has applied the diffusion model to an extraordinary range of topics including lexical decision, recognition memory, perceptual discrimination, aging, clinical populations, and neuroscience. His work with Gail McKoon on memory retrieval and with Jeffrey Starns on recognition memory has demonstrated the model's power as both a measurement tool and a theoretical framework. The EZ-diffusion method, while not developed by Ratcliff, has made his model accessible to a wider research community.