Sequential Expectancy in Visual Search

An experiment was designed to determine whether there was a sequential expectancy effect in visual search by which subjects carried over an expectation of the duration of the search from one trial to the next and produced a shorter search time on the n th trial if the (n-1) th trial had had a search time of similar magnitude. Search time was controlled by the surreptitious insertion of lags between the onset of the background (i.e., start of search) and the actual appearance of the target. Post-target search time (total search time minus lag) was then used as the dependent variable. Three lags (0, 7.5, and 15 s) were used in a random order. Two effects emerged. Post-target search time was found to be reduced if the previous trial had used the same lag as the given trial; and post-target search time was found to increase with lag. Both effects were explained by the construction of a sequential expectancy model.