In the previous papers, two behavioral features, the decrement of behavior variability and the facilitation of running response, were found (1) during the reconditioning period following extinction, (2) under partial reinforcement and (3) to the positive stimulus during succesive discrimination in the semicircular maze. These experiments also showed that these features were only found when reinforcement was presented after or between nonreinforcements. When only reinforcement or nonreinforcement was presented continuously, these features were not found. The present experiment was so designed as to examine this interaction effect of reinforcement and nonreinforcement, which we call induction, by use of the repeated conditioning-extinction procedure and to compare this result with those obtained in the three previous experiments.
After exploration, pretraining and preconditioning, twelve rats (the experimental group, i.e. the repeated extinction-conditioning group) were daily given 40 successive trials, the first half being continuous nonreinforcement and the second half being continuous reinforcement for 5 days in a semicircular maze which had a spacious field and eight goal boxes. An extinction period of 60 trials followed in the next day. Other twelve rats (control group) were treated by the same procedure, but they were daily given only 20 reinforcements for 5 days.
The results were:
1. During the repeated extinction-conditioning period, behavior variability (in three different measures) observed during the daily reinforced period of the experimental group progressively decreased with days until it reached lower than the chance level. However, this phenomenon was not found in the nonreinforced period of the experimental group as well as in the reinforced period of the control group.
2. The running responses measured in terms of running time, the number of sectione traversed and the number of pushing of the doors at the goal-boxes reduced in the same manner as behavior variability did. That is, the values of these three measures reduced more in the reinforced period of the experimental group than in the nonreinforced period of this group and in the reinforced period of the control group.
3. With reference to the similar results found in our three previous experiments, it might be concluded that the above two features were produced by the interaction effects of reinforcement and nonreinforcement.
4. Although the running responses during the nonreinforced period of the experimental group were similar to the phenomena revealing in a continuous extinction on each day, these changes progressively disappeared with repetitions of extinction-conditioning procedure.
5. During the extinction of 60 trials, behavior variability increased above chance and in the running responses a greater resistance to extinction was found than was when the rat was treated under partial reinforcement in our previous experiment. These results (4 & 5) may indicate that the order and the number of reinfocrcement and nonreinforcement were important factors in this kind of phenomena.