Generalization of operant conditioning of verbal output in three-man discussion groups

Abstract

The present study investigated the generalization of verbal operant conditioning using a group discussion paradigm. The hypotheses were: Hypothesis A: Verbal output of Target !s would increase with the presentation of positive reinforcers for talking and punishment for silence. Hypothesis B: Later testing of Target Ss in new groups would show generalization of increased verbal output. Hypothesis C: Later testing of Target Ss in new groups would show the reinforced Target Ss being rated as performing a greater degree of leadership function than non-reinforced Target Ss. Hypothesis D: The degree of generalization of verbal output would be a function of time, with Target Ss retested one day later showing more generalization than Target Ss retested one week later. Ss were female volunteers from an introductory psychology course. Twenty-four groups of three Ss were tested in a soundproofed room, with a hooded panel containing a red and green light situated on the table in front of each S. Ss were assigned to either a Reinforced or Non-Reinforced treatment, and each group met on three successive days for 30 minutes each day. On Day 1 no reinforcement was presented, in order to establish the operant level of verbal output for each Ss who spoke the least amount of time within their respective groups were selected as the Target Ss (N = 24). On Days 2 &3, Ss under the Reinforced treatment were instructed that they would receive a green light when they facilitated the group discussion and a red light when they hindered the group process. Actually Target Ss were positively reinforced (green light) for talking and punished (red light) for silence, while Non- Target Ss were punished for talking. Ss under the Non-Reinforced treatment received neither positive reinforcement or punishment. On Day 4, which occurred either one or seven days after Day 3, Target Ss were assigned to new groups with Ss who had not participated in the first discussion group. Target Ss received neither positive reinforcement or punishment during Day 4. At the end of Day 4, Ss ranked each other on a sociometric questionnaire concerned with leadership. Hypothesis A was supported by the results, with Target Ss showing an increase in verbal output under the Reinforced treatment (F2, 44 = 20.93, p<.01), and no change under the Non-Reinforced treatment (p>.05). Hypothesis B was not supported, with Reinforced Target Ss showing no significant difference (p).05) in verbal output from Non-Reinforced Target Ss on Day 4. However, Target Ss showed a significant increase of verbal output from Day 1 (First Group) to Day 4 (Second Group), for both the Reinforced and Non-Reinforced treatments (F1,20 = 22.77, p ( .01). Hypothesis C was not supported. Reinforced Target Ss did not show a significant difference (p) .05) between Day 1 (First Group) and Day 4 (Second Group) in leadership function, whereas Non-Reinforced Target !s showed a significant increase (F1,20 = 8.71, p<.01) in leadership function. Hypothesis D was not supported. Target Ss who were retested one day later showed no significant difference (p>.05) from Target Ss retested one week later.