An eye movement analysis of contextual cueing effects

Abstract

Contextual cueing is the implicit association of objects (or ‘cues’) in a visual scene due to repeated exposure, either spatially or semantically. Such associations can aid people in detecting a specific target more quickly and easily when it appears in a familiar context. The goal of this study was to investigate whether and how people utilize “distractors” as contextual cues during a visual target search. Ten undergraduate participants performed a luggage-screening task during which they were connected to an eye tracker. First, participants were trained using 25 luggage images, each of which contained a target (i.e., knife) and a specific distractor (i.e., iPod). During the post-training session, participants screened 100 bags with target base rate set at 50%. The bags contained, either, the distractor and the target (25 bags), the target only (25 bags), the distractor only (25 bags), or neither the distractor nor the target (25 bags). Participants’ fixation counts, saccade counts, saccade amplitudes, and scan paths were assessed. It was found that in the presence of the distractor, participants demonstrated fewer fixations and saccades as well as shorter saccade amplitudes than in the absence of the distractor. This was particularly salient in the absence of the target. The results suggest a higher level of search efficiency when the distractor cue was present, and less organized and more aggressive search pattern in the absence of the distractor cue. We contend that participants formed an implicit association between the distractor and the target and used this association to improve the efficiency of their visual search.