Feature synchrony-asynchrony and rate of change in visual search

ABSTRACT

Attention is known to be sensitive to the temporal structure of scenes. We initially tested whether feature synchrony, an attribute with potential special status because of its association with objecthood, is something which draws attention. Search items were surrounded by colours which periodically changed either in synchrony or out-of synchrony with periodic changes in their shape. Search for a target was notably faster when the target location contained a unique synchronous feature change amongst asynchronous changes. However, the reverse situation produced no search advantage. A second experiment showed that this effect of unique synchrony was actually a consequence of the lower rate of perceived flicker in the synchronous compared to the asynchronous items, not the synchrony itself. In our displays it seems that attention is drawn towards a location which has a relatively low rate of change. Overall, the pattern of results suggested the attentional bias we find is for relative temporal stability. Results stand in contrast to other work which has found high and low flicker rates to both draw attention equally [Cass, J., Van der Burg, E., & Alais, D. (2011). Finding flicker: Critical differences in temporal frequency capture attention. Frontiers in Psychology, 2, 320]. Further work needs to determine the exact conditions under which this bias is and is not found when searching in complex dynamically-changing displays.

KEYWORDS:

• Visual search
• attention
• feature synchrony
• flicker
• temporal rate

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Although there was no significant interaction with rate, observation of the graphs shows the asymmetry between T_SD_C, and T_CD_S was larger in the slowest (2.5 Hz) rate condition compared to the others. Our experiment may have lacked sufficient power to reveal a significant three-way interaction with rate. It should be noted that the difference between the slow (_s) and fast (_c) items was always a ratio; the 2.5 Hz condition therefore had the largest periodic difference in flicker between the slow and fast items. Given this, it is unsurprising that this rate condition produced the largest asymmetry.

2. We note that we did not monitor eye movements in this study. The long search times in our results (often over 2 seconds in the largest set size conditions) mean that observers were almost certainly making voluntary eye movements searching for the target. We suspect that the effect biasing attention is one which arises from direction of covert attention, and that any effects on eye movements would be a consequence of this. We suspect that similar findings would be obtained regarding the effect of temporal stability if, for instance, observers had to maintain fixation and the search displays were presented in the periphery.

3. An exception to this is a study by Ivry and Cohen (1992). These observers found that a high-rate moving target tended to pop out in search while a slow moving target did not. Thus the manipulated attribute was motion speed rather than flicker frequency and the search asymmetry, though present, was in the opposite direction to ours.