Monitoring the effects of increasing resource demands on the clarity of visual images.

Abstract

This paper examines the changes in the clarity of images over time under conditions varying in resource demands. In the experiments, subjects listened to descriptions of arrays of objects and moved a lever according to their assessment of the clarity of their visual images of the whole array. The position of the lever was sampled every half second by computer. The rated clarity of visual image (Y) increased over time (X) and fitted well to the power function Y = aXb for arrays of one to five objects. The increases in set size produced increasing demands on the resources of the limited capacity visual imagery system leading to an exponential‐like decline in the final rating values across set size. In the first two experiments, increasing presentation rate produced a differentiation in the slopes (a) of the power functions fitted to each set size, because visualization is more difficult at the faster rate. As it is relatively more difficult to visualize larger arrays at the faster rate, a relativity hypothesis suggests that visualizing smaller arrays is relatively easier in the fast condition, producing the differentiation in slopes. The last two experiments further confirmed the relativity hypothesis and correspondingly falsified an alternative hypothesis predicting that images decaying over time produced these findings; for instance, in the last experiment, the slope (a) and exponent (b) of the power function were higher under the slow presentation rate indicating that this condition was easier. This was further confirmed by verbal reports and by the final rating values. The image decay hypothesis predicted the contrary result as the possibility of image decay was reduced in the fast condition. These results also elucidate previous experiments that demonstrated an increase in visualization latencies as a function of set size. In such experiments latencies increase as a function of set size because the clarity of the image is slower to reach criterion as the number of objects to be visualized increases.