Human efficiency for visual detection of targets on cathode ray tube displays using a two-level multiple-channel time history format.

Human observers were tested for their ability to detect targets on visual displays. The displays simulated the multiple-channel time history format, using two levels of intensity encoding. A target was presented on 50% of the trials at a constant bearing. Observers indicated their judgment as to whether a target was present by using a four-category rating scale. Receiver-operating characteristics (ROCs) were generated from the rating data and compared to ROCs for an optimal detector. In experiment I, data were collected for two signal-to-noise ratios and for 32, 64, and 128 lines of data. Results indicated that observers were 3-5 dB less sensitive than an optimal detector. Performance improved as the number of lines increased but not to the extent predicted by optimal integration of information across lines of the display. In experiment II, the bearing at which subjects judged whether a target was present, was clearly delineated from the background by encoding information presented at this bearing in red. The purpose of experiment II was to determine the degree to which human inefficiency in detecting targets is due to the inability to focus on a single column of data. Subject performance improved when the red encoding was utilized, but subjects still performed 2 dB below ideal. This suggests that subjects use a suboptimal decision rule for judging the presence of the target.