Numerous prior studies have reported that rats, pigeons, and humans prefer predictable over unpredictable rewards of equal frequency and magnitude. A frustration-theory analysis of this preference suggests that it obtains because the unpredictable partial reinforcement procedure is aversive whereas the predictable discrimination procedure loses its aversiveness. The preference, on such an analysis, arises due to the tendency to avoid the unpredictable of two alternatives. Since frustration varies as a function of magnitude of reward, the avoidance tendency should increase with increases in reward magnitude in the unpredictable alternative. One group of rats in the present study showed a clear preference for seven versus five 45-mg Noyes Pellets. A second group showed the oft reported preference for five pellets predictable versus five pellets unpredictable. A third group of rats showed a preference for a five-pellet predictable reward over a seven-pellet unpredictable reward. The results of this experiment provide evidence for a frustration-theory analysis of the preference for predictable reward.
An experimental test of a “frustration” theory of reminiscence
