Abstract
Some previous work suggests the surprising possibility that when subjects are given valid feature information (e.g., color) about non-target items, it actually hurts rather than helps their performance (e.g. Moher & Egeth, 2012; Tsal & Makovski, 2006). They suggest that this effect is due to a search process where observers first select the to-be-ignored item and then inhibit it, similar to the selection process of the "attentional white bear" effect. In Moher & Egeth (2012), observers were randomly cued with the color of the to-be-ignored item. Since observers were given a different color on each ignoring trial, learning a specific feature cue was not possible. However, is search still inefficient when observers learn that the target will never contain a constant specific feature (e.g. will never be red)? Using a similar paradigm to Moher & Egeth (2012), observers were prompted with either information about a non-target item in the search display (i.e. "ignore trials") or no information about the search display (i.e. "neutral trials"). Afterwards, they were asked to search a display for a capital letter "B" or "F" among other colored letters. Critically on ignore trials, observers were given one color (randomly assigned and counterbalanced across subjects) to learn to ignore for the duration of the experiment. Results revealed that early on in the experiment, there was a cost in reaction time (RT) on ignore trials compared to neutral trials, similar to Moher & Egeth (2012). However, after extended practice, RTs on ignore trials decreased significantly below RTs on neutral trials, suggesting that observers had learned to use more efficiently the information about the irrelevant feature. Overall, when observers learn to use information about non-target items, this process allows for more efficient search, just as when they learn to use information about target items (Sireteanu & Rettenbach, 2000).
Meeting abstract presented at VSS 2014