Investigating Interactions between Search Mechanisms in the Control of Visual Attention

Olds, Cowan and Jolicoeur (2000) showed that although the mechanisms underlying visual search have traditionally been assumed to be independent, they in fact interact. Using coloured disk stimuli, they interrupted pop-out search (target plus D1 distractors) by adding more distractors (D2s) of a different colour to the display before pop-out processes were able to find the target. In short, partially completed pop-out processes facilitated subsequent difficult search processes (“search assistance”). The present study investigated hypotheses for this interaction. In Experiments 1 and 2, we used methods aimed at determining where the bulk of attentional resources are allocated during search of a visual display assumed to produce search assistance (by measuring the effect of inhibition of return [IOR] between D1 and D2 locations). In Experiment 1, we first presented observers with a search task that has been shown to produce search assistance (using coloured disks; see Olds et al., 2000). Immediately following target response, observers had to determine as quickly and accurately as possible whether a small probe-dot (that appeared on one of the disks) was present or absent. The results of Experiment 1 provided tentative support for a negative prioritization hypothesis, which proposed that some initial distractors (D1s) are eliminated from consideration during the second portion of the display. The sequence of events in Experiment 2 were identical to that of Experiment 1 except that, following target response, observers had to make a temporal order judgement (TOJ) as to which of two physically simultaneous lines (one on a D1, and one on a D2) appeared first. The results of Experiment 2 did not support either of the hypotheses regarding the nature of search assistance. Experiment 3 examined the effect of spatial cues on difficult search by attempting to eliminate the effect of negative prioritization of D1s while measuring a potential effect of positive prioritization of initial locations (D1s plus target location). The results of Experiment 3 provided evidence in support of a positive prioritization hypothesis, which proposed that the initial items are more likely to be searched in the second portion of the display.