Aging and the speed/accuracy relation in visual search: evidence for an accumulator model.

Two models of performance in visual classification tasks, the fast-guess model and the accumulator model, offer contrasting accounts of the relation between speed and accuracy. We attempted to distinguish these models in the context of age-related changes in visual search performance. Twenty-four young adults (mean age 21.50 years) and 24 older adults (mean age 66.17 years) performed a visual search task requiring the detection of an upright L among rotated L's. The results supported the accumulator model, in that mean reaction time (RT) was higher for error responses than for correct responses, and there was a positive relation between RT and error rate. Both of these effects were more pronounced for older adults than for young adults, even when visual acuity was covaried statistically. We conclude that age-related changes in visual search performance involve a decline in the efficiency of sampling the amount of evidence necessary to exceed a decision criterion.