Gaze Linking in Visual Search: A Help or a Hindrance?

Abstract

Gaze linking allows team members in a collaborative visual task to scan separate computer monitors simultaneously while their eye movements are tracked and projected onto each other’s displays. The present study explored the benefits of gaze linking to performance in unguided and guided visual search tasks. Participants completed either an unguided or guided serial search task as both independent and gaze-linked searchers. Although it produced shorter mean response times than independent search, gaze linked search was highly inefficient, and gaze linking did not differentially affect performance in guided and unguided groups. Results suggest that gaze linking is likely to be of little value in improving applied visual search.

Searching with and against each other: Spatiotemporal coordination of visual search behavior in collaborative and competitive settings

“…Other aspects of the interface of the task, such as the method of visualizing shared gaze, could also have an influence on the results. Instead of using a simple dot or ring to indicate current gaze position of the partner (Brennan et al, 2008;Messmer et al, 2017;Neider et al, 2010;Yamani et al, 2017), we highlighted the hexagon that is looked at by the partner. This may have contributed to the highly efficient collaborative search behavior displayed by our participants, because our highlighting method has two benefits over a dot or ring for displaying shared gaze.…”

“…This indicates that in Messmer et al's (2017) experiment, seeing the other searcher's gaze behavior through a gaze marker (a dot) caused individual searchers' search efficiency to decrease. Messmer et al (2017), however, did not report analyses of eye-movement data, making it unclear what led to the decrease in efficiency. Two other studies using a different search task (Neider, Chen, Dickinson, Brennan, & Zelinsky, 2010;Yamani, Neider, Kramer, & McCarley, 2017) also reported either no reduction in reaction time when two searchers were provided with shared gaze (Neider et al, 2010), or even an increase (Yamani et al, 2017), compared with a single individual searching the display.…”

“…While Brennan et al (2008) reported that adding a second searcher who collaborated through shared gaze led to the team finding targets twice as fast, an experiment by Messmer, Leggett, Prince, and McCarley (2017) using a very similar search task and setup found a much more modest speedup as their paired searchers were only about 1.2-1.4 times faster than a single searcher. This indicates that in Messmer et al's (2017) experiment, seeing the other searcher's gaze behavior through a gaze marker (a dot) caused individual searchers' search efficiency to decrease. Messmer et al (2017), however, did not report analyses of eye-movement data, making it unclear what led to the decrease in efficiency.…”

Shared Gaze Fails to Improve Team Visual Monitoring

“…The data of Experiment 2 add to a body of mixed evidence as to whether shared gaze does (Brennan et al, 2008; Niehorster et al, 2019; Zhang et al, 2017) or does not (Messmer et al, 2017; Neider et al, 2010; Yamani et al, 2017) facilitate collaborative search. The literature in total suggests that the benefits of shared-gaze awareness to visual search may be nonrobust, or of limited generalizability.…”

“…Another study (Zhang et al, 2017) also found that shared gaze led to shorter search times, but only for one format of gaze representation out of four that were tested, and only for target-present judgments, not target-absent. Other experiments have either found no evidence that shared gaze speeds visual search (Neider et al, 2010) or have found weak evidence that it can interfere with search (Messmer et al, 2017; Yamani et al, 2017).…”