The Effect of Computer System Response Time Variability on Interactive Graphical Problem Solving

An experiment was carried out to examine the effect of variability in computer response time on problem solving in an interactive-graphic environment. Subjects seated at a lightpen-actuated display were required to adjust five parameters to cause a graph plot to lie within an acceptable region; parameter adjustment was achieved by lightpen hits on a light-potentiometer. For a mean response time of 1 s, and a range of 0.2 to 1.8 s, it was found that contrary to a widely held belief, the variability of response time had no effect on the time taken to obtain a solution. The principal effects noted were period effects, demonstrating a change in the subject's strategy and fluency of lightpen use as he worked through three separate attempts at the problem. Some influence of time of day on strategy was in evidence, and there was a suggestion of a degree of correlation between neuroticism and the time taken by the subject to obtain a solution at the first attempt.