Achievement

Trainee John Wilder, working with Kowler and Feldman, studied the factors that control the timing of saccadic eye movements. There are two ways to control timing: either look at a location as long as needed, or pre-set the duration based on expectations. Wilder et al. found that expected duration was the preferred strategy, that is, people time saccades based on how difficult they expect a visual discrimination to be, not on how difficult it actually is. To understand why this strategy is used, Wilder is developing an ideal observer model of performance. Comparison of the model and performance shows that humans are inefficient and fail to get all the information from the visual displays. Thus, the visual system does not necessarily generate strong signals to indicate when enough information has been acquired. Given such uncertainty, the safer strategy would be to time saccades on the basis of past history.