Abstract
The role of memory in visual search has lately become a controversial issue. Horowitz and Wolfe (H&W, 1998) asked participants to search displays for a letter “T” among letters “L” in two experimental conditions: In the static condition, the displays remained unchanged, whereas in the dynamic condition, all letters were randomly re-located every 111 ms. If search involves a memory-based mechanism that keeps track of the previously examined locations, observers would be expected to have great difficulties searching the dynamic display. Surprisingly, the target-present search rates in the dynamic did not differ from those in the static condition. Because a memory-based mechanism would be of no use in the dynamic condition, H&W concluded that memory is likewise not involved in the static condition.
One alternative explanation for the results of H&W is that participants adopted a sit-and-wait strategy, which consists of attending to a region of the display and waiting for the target to appear there. Although H&W tried to rule out such an alternative explanation, this study argues that their participants have opted for a more sophisticated sit-and-wait strategy, one that allows the attentional focus to encompass several stimulus locations, and where the attentional focus can be shifted to other areas after some time has elapsed.
This hypothesis is supported by experimental data showing that performance in H&W's dynamic condition does not differ from another dynamic condition (aperture condition), in which observers are forced to adopt a sit-and-wait strategy by being presented with a limited region of the display only. These results seriously question H&W's assumption that the dynamic condition provides an “analytical” counterpart to the static condition, which would allow us to infer how search is performed with normal, static displays. This paves the way for other theories emphasizing the role of memory in visual search.