| Summary: | Abstract: p. 59-62.
Autobiography: p. 63.
Thesis (Ph.D.)--Boston University.
Bibliography: p. 58. === A social response was defined as a discriminated operant which includes the behavior of more than one organism in a group and i at least partly under the control of stimuli produced by the group itself.
The purpose of the study was to examine the relationship between reinforcement and the rate of one kind of social response, namely the serial order of the speakers in a conversation. It was assumed that the laws of learning discovered in the study of the behavior of individual organisms would be sufficient to account for this aspect of group behavior, requiring no further theoretical assumptions. When the response was followed by a reinforcing stimulus it was expected to occur more often. When, subsequently, the response was permitted to occur in the absence of reinforcement its rate was expected to decline.
Method. Subjects were asked to come to a series of unanimous decisions about which of several stimuli to try to communicate telepathically to a receiver in another room. After each consensus they were to stop talking immediately, concentrate on that stimulus, and wait silently to be informed whether or not their message was received. After a three-second delay they were signalled whether or not their message had been received correctly and went on to try again.
Each consensus constituted a trial, of which there were 700 in all during seven days. A record was kept of the identity and order of the last two subjects to speak in each trial. In a thee-person group there are six possible two0person trial endings. One of these endings (AB) was the social response studied.
On the first two days the "correct" signal (positive reinforcing stimulus) was sounded after one quarter of the trials on a random schedule, in order to establish a baseline rate for AB.
Days three to five were the conditioning period, during which AB was reinforced whenever it occurred. [TRUNCATED.
|
|---|
