Abstract
Consider one of the many visual search tasks that often confronts each of us, such as searching for your own car among a sea of other cars in the parking lot. If you were given a preview of the colors or the shapes of the cars before launching into an actual search, could you more efficiently locate your car than if you had no feature-based preview? It seems clear that a preview would enable you to isolate the group of cars that has the same feature as your car, then restrict your search to just this subset of cars, and without the preview you wouldn't have this head-start on your search. Nevertheless, a recent study (Olds & Fockler, 2004) found little response-time advantage for a feature preview relative to no preview. We thought that these puzzling results may have arisen from bottom-up factors interfering with participants' ability to use the information from the preview to impose top-down control over their searches. In a series of experiments we found that the effectiveness of feature preview varies with the strength of bottom-up factors.