Abstract
Some real-world tasks involve observers visually searching in pairs or teams (e.g. searching buildings for weapons). One issue that has not yet been explored is how search behaviour adapts or shifts when a partner in a pair shows signs of missing targets or making false alarms. In the present study, we examined participants' eye movements and behavioural performance (accuracy and reaction time) in a visual search task where two targets could appear. The participants were instructed that they must search for one target, and were informed that a partner had already searched the same displays for a second target. The participants were told that they could also search for the partner's target if they so wished. Following each trial, they were given feedback relating to both their own performance and their partner's performance. The key manipulation that we implemented was relating to the partner's performance since the partner did not, in fact, exist. Partner responses were determined according to a pre-defined signal-detection theory criterion. One group of participants were teamed with a conservative-criterion partner (where the partner had a bias to respond 'absent'), while a second group were teamed with a liberal-criterion partner (where the partner had a bias to respond 'present'). While overall accuracy was equivalent across conservative-criterion and liberal-criterion conditions, reaction times for those paired with a conservative-criterion partner were slower than for those paired with a liberal-criterion partner. Longer reaction times for this group were driven by more fixations and increased verification times (time between first fixating the target and responding 'present'). These results demonstrate that direct experience of conservative search behaviour in a partner influences an observers search speed but not accuracy. We suggest that working with a partner who is likely to miss targets fosters a conservative approach to responding in visual search tasks.
Meeting abstract presented at VSS 2014