Abstract
Numerous factors impact attentional allocation, with behavior being influenced by both individual intent and our visual environment. As such, attention research often focuses on how attention can be made to operate more efficiently. In the present study, we investigate the influence of joint search behavior on reaction time, accuracy, and oculomotor kinematics. Within the present context, joint search occurs when one individual controls the visual input of a second individual via a gaze contingent window (e.g. participant 1 controls the window via their eye movements and participant 2—in an adjoining room—sees only the stimuli that participant 1 is directly fixating). In Experiment 1, pairs of participants complete three blocks of a discrimination task (target present/absent) which required them to either 1) Search and perform the discrimination task individually, 2) Search the display while their partner performs the discrimination task, or 3) Perform the discrimination task while their partner searches. Critically, visual search is most efficient (e.g., highest accuracy with minimal search speed reduction) when the person performing the discrimination task is doing so for the second time while the person controlling the visual output is searching for the first time. Performance is least efficient when the person performing the discrimination task is doing so for the first time while the person controlling the visual output is searching for the second time. In two subsequent experiments, task difficulty is enhanced to determine the generality of this effect. This paradigm creates a counterintuitive speed/accuracy tradeoff which combines the heightened ability that comes from task experience (discrimination task) with the slower performance times associated with a novel task (initial search) to create a potentially more efficient method of visual search. As such, the present paradigm has important practical applications to events outside the laboratory, such as x-ray screening at airport terminals.
Meeting abstract presented at VSS 2013