Premium
An experimental simulation of an internation truel
Author(s) -
Willis Richard H.,
Long Norma Jean
Publication year - 1967
Publication title -
behavioral science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.371
H-Index - 45
eISSN - 1099-1743
pISSN - 0005-7940
DOI - 10.1002/bs.3830120104
Subject(s) - subject (documents) , psychology , statistical significance , statistical analysis , social psychology , statistics , computer security , mathematics , computer science , library science
A simple experimental simulation was conducted of a three‐sided internation conflict situation (i.e., a “truel”), in which each nation was represented by one subject. On each of up to 50 trials, each subject made a trichotomous decision not to attack, to attack the first of the others, or to attack the second of the others. To be attacked was to be eliminated, but a victim had the opportunity to make one more decision after being attacked. A 2 × 2 × 2 orthogonal design was employed in which the independent variables were (a) knowledge of source of attack, (b) knowledge of maximum number of trials, and (c) sex of subjects. Three groups were run in each of the eight conditions, making 24 groups and 72 subjects. The following results were among those worthy of note: 1. Attacks occurred in 75 percent of the groups. There was a trend, approaching statistical significance, for no‐attack groups to occur more frequently in the Source Known conditions. Conversely, almost twice as many individual attacks occurred in the Source Unknown condition. 2. An initial attack was invariably followed by at least one, and often two or three, additional attacks on the next trial or two. Closely related to this “chain reaction” effect is the fact that there were no groups with two survivors. Either there was a single victor, or‐much more commonly‐all group members experienced a common fate. 3. Fully 90 percent of all initial attacks occurred during the first five trials, and almost half occurred on the second trial alone. Out of 21 initial attacks, only one came after the seventh trial. 4. There was a pronounced and highly significant interaction between experimental conditions and strategies adopted by individual subjects as these jointly determined the chances of survival. Although a strategy of aggression always failed, a nonaggressive strategy had varying chances of success depending upon the experimental condition. In the maximally structured condition (Source Known, Trials Known), nonaggression succeeded half the time , while in the minimally structured condition (Source Unknown, Trials Unknown), it invariably failed.Results are discussed in terms of hypotheses suggested concerning the outcomes of inter‐person, intergroup, or internation truels.