Motion disrupts dynamic visual search for an orientation change

Abstract

Visual search in dynamic environments, for example lifeguarding or CCTV monitoring, has several fundamentally different properties to standard visual search tasks. The visual environment is constantly moving, a range of items could become targets and the task is to search for a certain event. We developed a novel task in which participants were required to search static and moving displays for an orientation change thus capturing components of visual search, multiple object tracking and change detection paradigms. In Experiment 1, we found that the addition of moving distractors slowed participants' response time to detect an orientation changes in a moving target, showing that the motion of distractors disrupts the rapid detection of orientation changes in a moving target. In Experiment 2 we found that, in displays of both moving and static objects, response time was slower if a moving object underwent a change than if a static object did, thus demonstrating that motion of the target itself also disrupts the detection of an orientation change. Our results could have implications for training in real-world occupations where the task is to search a dynamic environment for a critical event. Moreover, we add to the literature highlighting the need to develop lab-based tasks with high experimental control from any real-world tasks researchers may wish to investigate rather than extrapolating from static visual search tasks to more dynamic environments.

Keywords: Dynamic visual search; Feature change; Monitoring; Motion silencing; Response time.

Fig. 1

Timeline of the task. Each trial beings with a fixation cross. The items then appear on screen for 500 ms. In the static condition (golden screen), the items do not move. In the moving condition (turquoise screen), the items move around the screen. After a random interval between 2,000 and 4,000 ms, one item will undergo an orientation change. The panel in the top right shows the starting orientation of all items (a) and the rotated orientation of the target item (b)

Fig. 2

Mean RT for each set size and display type. Error bars show standard error

Fig. 3

Mean RT for each set size and display type. Error bars show standard error