Abstract
The psychologically acceptable upper limits of computer system response time (SRT) are experientially proposed by R. B. Miller (1968). The difference of the kinds of psychological stress under the upper limits are studied in this paper. At several levels of SRT, calculation and odd-even discriminating task were imposed on subjects with a personal computer. Performance time, psychological stress and eye movements were analyzed. As the results, psychological stress are almost acceptable situation under the proposed upper limits by Miller. However the differences of the kinds of psychological stress, such that rather long SRT brings relaxed feelings but rather short SRT brings self-paced feelings, are observed. Too short SRT brings more psychological stress, and when the task requires eye movements, performance time decreases. These results mean there exist lower limits of SRT in the ergonomic sense. In conclusion, Miller's proposed upper limits of SRT are generally accepted in this study. However kinds of stress slightly differ even under the upper limits and the lower limits of SRT appear to exist in the ergonomic sense.
