Evaluation of the “Time Tunnel” Display Design Technique

Abstract

Three experiments were conducted to evaluate the time tunnel display design concept. This technique is similar to traditional “strip chart” or “trend” displays in that temporal information (i.e., the value of system variables over time) is provided. It is different in that perspective geometry and the “depth plane” are used to present this information. In Exp. 1 a time tunnel display and a “quickened” display were evaluated using a dynamic simulation. The results indicate that display quickening improved performance significantly, but that the time tunnel display did not. In Exp. 2 a redesigned version of the time tunnel display was evaluated; the overall pattern of results did not change substantially. In Exp. 3 an alternative methodology was used. Participants were shown “snapshots” of system states (states generated in Exp. 2) and completed three tasks: 1) fault detection (confidence rating), 2) fault identification (categorical assessment), and 3) estimation (estimate of quickened variable). Three displays were evaluated: a baseline display, the redesigned version of the time tunnel display and a traditional “trend” display. The results indicate that both displays with temporal information (time tunnel and trend displays) produced better performance than no temporal information (the baseline display). The results also indicate that the time tunnel display produced significantly better performance than the trend display.