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.

Specific Experimental Procedures

A number of experiments have now been conducted within this paradigm in various university settings. Results from the first two studies illustrate the experimental conditions we are examining and the data on generalization effects we have obtained thus far. The first experiment was a dissertation undertaken by Marian Rogers (1982), and the second was a dissertation by Keith Edwards (1984). Both employed the three-stage group problem-solving paradigm described above, with the experimental manipulations part of the task instructions in the second phase.

Study 1. Study 1 examined the effects of a structural variable (cooperative vs. competitive interteam reward contingency) crossed with an instructional manipulation designed to focus attention on task versus interpersonal aspects of a team performance. Each experimental session enlisted eight female subjects whose judgments on a series of dot-estimation problems provided the ostensible basis for division into two (equal-sized) categories, designated respectively as underestimators and overestimators. [...] Following the category induction of Phase I, category members were randomly reassigned to two heterogeneous teams, each consisting of two underestimators and two overestimators. [...] In the first part, each team member worked separately to generate a list of four or five of her own personality characteristics she believed would be important qualifications for space travel. After producing these individual lists, the team convened to discuss them and to arrive at a selection of five traits representing the team’s consensus as to the most important characteristics needed for space travel. [...] In the interpersonal focus condition, team members were told that successful task performance required that they form accurate impressions of what their fellow team members were really like. In the task focus condition, they were told it would be important to make an accurate assessment of the quality of each member’s list as a contribution to the team product. [...] In the cooperative condition, subjects were told that the problem solutions of the two teams would be evaluated jointly to determine their joint eligibility for a small monetary reward. In the competitive condition, they were told that the team solutions would be compared and the team with the better product would be eligible for a reward. Following these instructions, both teams worked on the assigned problem and later submitted a final list to the experimenter. At this point participants were asked to make a series of ratings of their own team product and of the individual team members [...]. They then made a series of ratings of the other team and its members paralleling the ones they had made on members of their own team. As our final generalization measure, all participants in the session were shown a videotape of a four-person team supposedly taped from an earlier session and rated the team members on the video [...].

Study 2. Study 2 paralleled the general procedures of Study 1, but the experimental conditions varied category salience and task orientation differently. In each session, six participants were divided into two equal categories based on the dot over- and underestimation procedure. [...] Teams were composed of one person from one category and two persons from the other category, to generate a majority-minority differentiation within each team. [...] When assignment was category based, overestimators and underestimators were assigned to the respective teams explicitly as representatives of their membership in one of the two categories; when individuated, groups were ostensibly formed to maximize distinctiveness among subjects in their opinions on a set of general attitude measures obtained earlier in the session. [...] Following the formation of cooperative teams under the conditions above, the two teams then worked independently of each other on the problem-solving task developed for Study 1. Upon completion of the task, members of each team completed the performance and evaluation measures on members of their own and the other team that were used in Study 1. [...] For the allocation response measure, subjects were asked to distribute up to 100 “chips” to all team members on the basis of their perceived contribution. (Allocations to each team member were independent, so that each could receive from 0 to 100 chips.) One measure of bias was the mean number of chips allocated to outgroup category members on the other team as compared with the mean number allocated to ingroup members. Evaluative trait ratings were similarly compared in terms of mean ratings on combined 7-point scales of “friendliness” and “competence.”