A component task analysis of stereoscopic displays

Abstract

Considerable research has centered around the issue of whether stereoscopic (3D) viewing allows improved viewer performance for tasks that involve three-dimensional information. Taken as a whole, such previous research indicates that the potential stereoscopic advantage may be dependent on the nature of the task being examined. This task dependency makes it difficult to predict whether stereoscopic viewing will improve viewer performance for a given untested task. By measuring performance over a variety of component tasks, this research examined the potential task-dependent nature of the stereoscopic advantage. In addition, a method was proposed to employ such component-task data for predicting the stereoscopic advantage within future unknown tasks.

A set of 12 component tasks (in six task groups, each with two representative tasks) was developed to represent the various task demands of processing 3D visual information. Participants performed each of the 12 component tasks in both a monoscopic (2D) and a stereoscopic (3D) viewing condition. Performance was measured in terms of viewing time, percent accuracy, and a generic mental effort rating.

Results indicate that when certain display guidelines are not violated, stereoscopic display improves or at least maintains the overall level of viewer performance for most tasks. Furthermore, the results clearly indicate that the stereoscopic advantage is dependent on the nature of the task. Although further refinement to the set of component tasks is necessary before the precise nature of the task dependency can be determined, the component task method displays considerable promise for being able to predict the stereoscopic advantage for any number of complex 3D tasks.