Subjects

Subjects were 192 female undergraduates [...].

Design

Based on their alleged performance on a dot estimation task, eight previously unacquainted females were divided into two social categories [...].

Initially, the two groups of four subjects who shared category identity worked separately on a task designed to establish ingroup cohesiveness.

After evaluating their group’s and the other group’s products, members were reassigned to heterogeneous teams, each composed of two overestimators and two underestimators, to complete a second task.

Teams were assigned either to an inter team cooperation or competition condition and to one of three social orientation conditions (task focus, personal focus, or no-focus), in a 3 x 2 factorial design.

Procedure

Ingroup formation. After seating the eight participants in a common room, a female experimenter told them that the experiment was concerned with how people with different perceptual and analytical skills work together on problem-solving tasks.

She explained that a perceptual task would be used to divide them into two groups.

Subjects were then asked to estimate the number of dots on each of five slides containing between 25 and 55 black dots. Every slide was shown for approximately 5 s with a 10-s intertrial interval enabling subjects to write down their estimates.

The experimenter then collected the estimation sheets, picked up a calculator, and went to an adjoining lab for approximately 4 min, allegedly to score the tests. During this period, she randomly assigned subjects to the overestimator and underestimator groups. After returning, she gave the following information: [Text A]. She then announced the category membership of each subject [...].

The experimenter then led the two groups to separate rooms and gave them written instructions for their first task [...]. When finished, each group evaluated their solution on four 7-point bipolar adjective scales, rating the intelligence, logic, relevance, and strength of the groups’ rankings and justifications. Next, each group was shown the other group’s solution and asked to evaluate it using a form identical to that used to evaluate their solution. Intergroup rivalry or hostility was then enhanced by separately telling each group that they had rated their own solution somewhere between the 90th and 100th percentile of the rating system but the other group had rated it somewhere between the 60th and 70th percentile. Subjects then reunited in a common room and responded to five questions that assessed the effectiveness of the inductions of ingroup cohesiveness and intergroup hostility.

Heterogeneous team formation. To begin the next phase, subjects were told that two workteams, each comprised of two underestimators and two overestimators, would work separately on a new task [...]. Teams were assigned and then separated by escorting one team to an adjoining lab. Once composed, the teams received instructions for a second [...] task.

Manipulation of reward structure. In the cooperative condition, subjects were told that if the final products of both teams reached a certain standard, both would be eligible for a small monetary reward. In the competitive condition, they were told the team that produced a final product closer to a standard would be eligible for a small monetary reward. All teams were given 20 min to work on the team task, after which dependent measures were administered.

Dependent Measures. In the response packets, subjects were first asked to reward each of the members of their team according to each person’s performance. Subjects also rated the friendliness, sincerity, and competence of each of the members of their team [...].

Finally, as an assessment of intrapersonal complexity of their perception of teammembers, a visualizatiuon task was administered.