The role of reward variables in vigilance performance of children

Abstract

Sixty male children of elementary school were randomly divided into four reward conditions of a visual vigilance task. The four groups differed on level of reward and order of reward-non-reward sessions. One group received continuous immediate reward on Session 1 and no reward on Session 2; the other 100% group was rewarded for correct responses on Session 2. Similarly, the other two groups were given 50% reward on one of two sessions; all sessions were 24 hours apart. Reward consisted of Hershey's candy kisses and was administered immediately following a correct response. The children were required to monitor a circular display of twelve half-inch red lamps; the lamps were illuminated in sequence so that the display- appeared as a single light moving in rapid "jumps". A correct response consisted of detection of a failure of illumination of a lamp; these signals were presented in a modified Mackworth series with a mean intersignal interval of 1.25 min. The S responded by depressing a push-button, handheld switch; all responses were recorded on paper tape. Masking noise fed into padded earphones was used to control the effects of ambient noise. The results indicated that administration of immediate reward increased the level of performance, i. e., increased percentages of correct responses, and decreased inter-subject variability of performance. However, in the 50% - 0% group, the improvement occurred with cessation of reward. The comparison or the effects of 100% and 50% reward supported the hypothesis of differential influence of level of reward; 100% reward resulted in a higher number of detections. The findings were interpreted in terms of filter theory and frustration effects, and were related to signal detection theory.