Association of neutral objects with rewards: Effects of different numbers of conditioning trials and of anticipated reward versus actual reward

Abstract

The major purpose of the study was to determine the kinds and amounts of “conditioning” that occur when Ss are anticipating the future receipt of rewards rather than actually obtaining rewards. A minor purpose of the study was to determine the possible effects of different numbers of conditioning trials. This resulted in a 2 × 3 design: half of the Ss actually received rewards (pennies) and half anticipated the receipt of rewards a day later, and the Ss were further subdivided into three groups receiving either 5, 15, or 30 conditioning trials. The Ss were 84 children from the third and fourth grades of Nashville public schools. The “conditioning apparatus” was a spin-wheel game in which the stopping of a pointer on one nonsense syllable resulted in the winning of two pennies and the stopping on another syllable resulted in the loss of a penny. A third syllable was experimentally neutral, being associated neither with the receipt of money nor the loss of money. Conditioning effects on the syllables were assessed with five dependent measures: verbal evaluation, reward expectaney, eye movements, pupillary response, and instrumental viewing responses. Significant effects were found for verbal evaluation and reward expectancy but not for the other three dependent measures. Effects due to anticipation were equally strong as effects due to the actual pairing of neutral objects with rewards, and the effects were as strong for Ss receiving only five conditioning trials as for those receiving larger numbers of trials. The results were discussed with respect to (a) the place of reward anticipation in molar behavior theory, and (b) the possible implications for the underlying processes at work in our typical conditioning procedures.