Academic journal article Canadian Journal of Experimental Psychology

Patterns of Eye Movements during Parallel and Serial Visual Search Tasks

Academic journal article Canadian Journal of Experimental Psychology

Patterns of Eye Movements during Parallel and Serial Visual Search Tasks

Article excerpt

Abstract Eye movements were monitored while subjects performed parallel and serial search tasks. In Experiment la, subjects searched for an "O" among "X"s (parallel condition) and for a "T" among "L"s (serial condition). In the parallel condition of Experiment lb, ""Symbol not transcribed"" was the target and ""Symbol not transcribed""s were distractors; in the serial condition, these stimuli switched roles. Displays contained 1, 12, or 24 stimuli, with both target-present and target-absent trials. RT and eye-movement measures (number of fixations, saccadic error, and latency to move) indicated that search efficiency was greatest in the parallel conditions, followed by the serial condition of Experiment 1a and, finally, by the serial condition of Experiment 1b. This suggests that eye movements are correlated with the attentional processes underlying visual search. How are we able to visually search through our complex environment to find a particular item that we need or want? For example, how are we able to find a pen lying amid papers and books on a cluttered desk? Furthermore, while conducting this search, how do we know that a particular feature in the visual array (e.g., "redness") belongs to one object (e.g., the pen) and not to another (e.g., a book)? Studies attempting to address such questions have made extensive use of the visual search task. In this paradigm, subjects are asked to search stimulus displays for a target among distractors. Typically, both display size (i.e., number of stimuli) and trial type (target present vs. target absent) vary across trials. A target stimulus in a visual search task may be defined either by a distinct feature (feature search task) or by a particular combination or conjunction of features (conjunction search task). For example, in a feature search task, subjects might be asked to look for a red "O" among blue and green "O"s. Here, the target's colour is a unique feature which distinguishes it from the distractors. In a conjunction search task, the target stimulus might be the same red "O," but this time, the distractors could be blue "O"s and red "X"s. In this case, each distractor shares at least one feature with the target, such that the target can only be defined by a specific conjunction of colour and shape. In feature search tasks, the target stimulus is typically found quickly and easily; it seems to "pop out" from the background of distractor stimuli. As a result, the number of distractors in the visual array has little effect on subjects' search latencies. In a conjunction search task, on the other hand, average response time usually increases as a linear function of display size. This increase is more pronounced on negative than on positive trials. In fact, the ratio of the slope on negative trials to that on positive trials is roughly 2:1 (Treisman & Gelade, 1980; Wolfe, Cave, & Franzel, 1989). The feature integration theory of attention was developed by Treisman and her colleagues (e.g., Treisman, 1988; Treisman & Gelade, 1980; Treisman & Gormican, 1988; Treisman, Sykes, & Gelade, 1977) to account for such findings. This theory proposes that features are processed automatically and in parallel. Thus, in a feature search task, response times show little effect of display size because the feature characterizing the target stimulus is detected preattentively and then "calls" attention to the position of the target stimulus in the visual field (Treisman & Gormican, 1988). In contrast, the conjunction of features is thought to require focal attention; as a consequence, stimuli are processed serially in a conjunction search task. To perform such a task, the "spotlight" of attention must be focused on each stimulus in turn, allowing its features to be conjoined to form a unitary object. This process continues until the target stimulus is identified or until the subject, having searched the entire array, decides that it is absent (Treisman & Gelade, 1980; Treisman & Gormican, 1988). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.