The following is an excerpt of the intervention methodology. For more information, please see the full text of the article on the publisher's website or through your institution's library.

METHOD

Experimental Task A management-training activity called the Railroad Game provided the cooperative task around which the experiment was conducted. The activity consisted of operating a railroad business that received orders to ship many kinds of products between 10 different cities in SOO cars of six different types. [...] The three participants filled orders; kept records on profits, losses, and maintenance charges [...]. [...] There were three positions in the railroad business. The communications officer received and handled the work orders, kept records of car maintenance costs based on the number and type of cars used, and reported each financial transaction to the station master. The shipping officer checked the number of days in transit for each shipment by inspecting the game board and kept records of all profits and losses that the team incurred. The equipment officer kept records of the location of each of the SOO railroad cars in the game, suggested the type of car to be used based on the type of product to be shipped, and reported the number of appropriate cars available at the city of origin. The research subject occupied the position of communication officer, while one white and one black experimental confederate filled the other two positions. [...]

Independent Variables The two independent variables in the experiment were the competence with which group members performed their assigned tasks (three levels) and group success-failure. [...] Subjects were randomly assigned to treatment conditions. [...] The subject interacted with a confederate who was trained to portray either (a) a relatively less competent stimulus person whose task performance handicapped his team, (b) an equally competent stimulus person, or (c) a relatively more competent stimulus person. [...] Thus, when confederate A was relatively more competent, confederate B was necessarily relatively less competent and vice versa. [...] When a confederate enacted the role of the relatively less competent stimulus person, he made numerous errors. [...] When a confederate enacted the role of the equally competent or the relatively more competent stimulus person, he made (and self-corrected) a few minor and insignificant errors at the beginning of the game. [...] Possible effects of variation in competence among the subjects on respect and liking for their group mates (the dependent variable) were minimized in two ways. First, random assignment of subjects to conditions should have precluded any systematic effect of subject competence on between-treatment comparisons. Second, subjects' impressions of their own task competence were kept as similar—and as favorable—as possible by not correcting their errors. [...] Subjects were told at the beginning of the game that they would receive a $5 bonus if their group's profits surpassed the average profits of past groups [...]. Subjects, who were United States Air Force enlisted men, were told in the success condition that they would be competing against groups of air force officers, while the comparison group for the failure condition comprised Army draftees. [...] At the end of each of the five "workdays," the experimenter reported the team's progress. For the success condition, the report was as follows: Day 1, even; Day 2, $500 ahead; Day 3, $500 behind; Day 4, $1,000 ahead; and Day 5, $3,000 ahead. The progress report for the failure condition was as follows for Days 1-5, respectively: even; $500 behind; $500 ahead; $1,000 behind; and $3,000 behind. [...] The stimulus person for whom the subject provided attraction ratings was either black or white. When the stimulus person (one confederate) was white, the second confederate was always black and vice versa. [...]

Subjective Experience Variables A four-item measure of perceived competence and a three-item measure of perceived success were used to assess the subjects' perception of the corresponding independent variables. [...] Two other objective experience variables were included: a one-item measure of the subject's perception of his own task-related competence and an eight-item measure of general satisfaction with the group experience [...]. In addition to these checks on the subjects' experience, we carried out an additional check on the adequacy of the manipulation of the level of competence. A trained observer listened to tapes of the experimental sessions. When the quality of the confederates', treatment-related behavior deviated significantly from the script, subjects were to be discarded. [...]

Criterion Variables After the management task had been completed, participants were separated, and attraction ratings for each of the confederates were obtained from the subjects. The postexperimental questionnaire was prefaced with oral and written assurances of anonymity and confidentiality. [...] The attraction measure consisted of 20 items and, with minor exceptions, was identical to that used by Blanchard et al. (1975). Five of the items concerned ratings of respect and liking; for example, subjects were asked to respond on a 7-point scale whether they would or would not like to be friends with the stimulus person. Nine of the items were adjective ratings from the evaluative dimension of the semantic differential. The evaluative ratings also used 7-point continua and included such dimensions as friendly-unfriendly, honest-dishonest, and valuable-worthless. The last 6 items were intended to specify real-life choices for the subjects. [...]

Subjects The subjects were 60 white airmen who had recently finished basic training and were stationed at Lowry Air Force Base for further training. [...]