No abstract

No abstract

The partial reinforcement extinction effect (the FREE) in classical aversive conditioning was investigated in 2 experiments. In the first, the nictitating membrane responses of 120 rabbits were conditioned at a 250-msec. interstimulus interval (ISI) under continuous reinforcement, partial reinforcement with the unconditioned stimulus (US) omitted (Group PO), or partial reinforcement with the US delayed to 1,500 msec. (Group PD). These 3 groups were factorially extinguished under US-Omitted, US-Unpaired, or US-Delayed extinction regimens. A significant FREE was obtained, but only for PO training and US-Omitted extinction. The second experiment, employing human subjects in a masked eye blink conditioning task, produced parallel results. A general discrimination view of the classical FREE seems applicable, but one in which neither cognitive factors nor intertrial conditioning of reinforcement aftereffects play a significant role.

The effect of magnitude of reinforcement upon resistance to extinction (Rn) was examined using simple partial-reward (PR) sequences (Experiment 1), or simple continuous-or varied-reward sequences (Experiment 2), using rats in a straight alley. Extinction tests replicated the typical interaction between reward magnitude and reward schedule (large, as opposed to small, reward increases Rn in PR groups but decreases Rn in 100% groups) using these elementary training sequences. However, the principles underlying this interaction were found to depend heavily on the specific sequence involved, over and above reward amount per se. The sequential hypothesis of Capaldi was shown to be consistent with the extinction results, whereas interpretation from the standpoint of frustration or dissonance theory was less satisfactory.

No abstract