Human optimization with moving optima

Earlier work had examined how humans optimize a two-variable function. This paper extends the study to optimization when the position of the optimum can change with time. Thirty student subjects performed a blind hill climbing task under two levels of display noise and three levels of movement of the optimum. Performance was studied using 13 measures, which were found to group into three factors, ‘deviation size’, ‘hits on target’ and ‘speed’. It was found that the movement of the optimum degraded performance once the general area of the optimum was reached. Display noise also degraded performance. Subjects with formal optimization training performed better on the ‘deviation size’ factor but had no advantage on ‘hits on target’ and ‘speed’ factors.