Punishment continuity and self-punitive behavior

Abstract

Partial punishment schedules (20% or 50% trials punished) are as effective as 100% punishment in generating and maintaining self-punitive behavior following avoidance training. If punishment is not continued. running extinguishes fairly quickly.After a ratlearns an avoidance running response, punishment given in the last section of the alley before the goal box is similar in some respects to the avoidance training shock. Given a few avoidance training trials, the animal is usually well on the way to the goal when shock onset occurs and the shock continues during the balance of the run. In the punishment case we have studied, the animal encounters shock 1 ft from the goal box and runs through the punishment zone to complete the trial. We have made use of the presumed similarity of avoidance training and punishment during extinction in predicting that self-punitive running will persist longer following avoidance training than following escape training (Beecroft & Brown, 1967). The first experiment reported in this paper was concerned with a related issue, the resistance to ordinary extinction without punishment which is produced by punishment training as compared to avoidance training. Insofar as punishment training is similar to avoidance training, comparable amounts of such training should produce equal resistance to extinction. EXPERIMENT 1 MethodThe runway and the avoidance training procedures have been described previously (Beecroft, Bouska, & Fisher, 1967). All Ss, female hooded rats 84-94 days old, were trained to one avoidance with a 3 sec avoidance interval. One group of 20 Ss was given 10 more avoidance training trials. The number of additional shocks received by the avoidance Ss ranged from 0 to 7 with a mean of 3.4. The other group of 20 Ss was given punishment training for 10 trials. The punishment was a55 V shock applied in thelastlftofthealley. After receiving either avoidance or punishment training for 10 trials, shock was discontinued for all Ss. The extinction criterion was 60 sec without entering the goal box. A maximum of 100 extinction trials was permitted, although no S approached this limit in Experiment 1. Four Ss were discarded. One did not reach the goal on the first avoidance training trial, one was lost because of an equipment breakdown, and two Ss extinguished during punishment training. ResultsThe Ss given extended avoidance training showed slightly more resistance to extinction than the punishment-trained Ss. The respective means were 23.5 and 17.5 trials completed; medians were 20 and 12.5. The difference between the means was not statistically significant. The resistance to extinction displayed by these groups is somewhat greater than that of Ss who are ex-

Effects of percentage of goal-punished extinction trials on self-punitive behavior

“…The results of the present study confirm the earlier fmdings of Babb and Hom (1971) and Hom and Babb (1975), who found that rats punished in the goalbox during extinction ran significantly faster than rats not shocked during extinction. In addition, the data support the findings of Beecroft, Fisher, & Bouska (1967), Gwinn (1949), and Martin and Moon (1968), who, using an alley shock punishment procedure , found that percent of punished extinction trials had no substantial effects on the magnitude of the self-punitive effect. It may also be noted that, while Melvin(1964) reported extinction run speeds to be an increasing function of percentage of punished trials, the punished groups did not differ significantly among themselves, and only a nonsignificant trend of increased run speed with increased percent of punishment was observed.…”

“…Previous studies examining the effects of periodic punishment, which applied punishment in different parts of the alley but not in the goalbox, have yielded mixed results. Several studies (Beecroft , Fisher, & Bouska, 1967;Gwinn, 1949;Martin & Moon, 1968) found that rats given different schedules of punished extinction respond similarly to rats given punishment on every extinction trial. In contrast , Bender and Melvin (I967)…”

Self-punitive behavior and periodic punishment

“…Melvin (1964) found self-punitive behavior more likely to occur with higher percentages of punishment after 33% or 67% schedules of "partial reinforcement" escape trials during acquisition. Beecroft, Fisher, & Bouska (1967) report self-punitive behavior to be equally probable within 20, 50, and 100% schedules of punishment after avoidance training. Thus, while different training procedures were employed in the two studies (escape vs avoidance), the effects of percentage of punishment have been different.…”