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.

EXPERIMENT 1

Method

Subjects Subjects were 62 female undergraduates; 30 from Douglass College and 32 from Rutgers College. [...]

Procedure [...] The experimenter asked the subject and accomplice to wait in an adjacent room and introduce themselves while materials for the study were being assembled. The manipulation of typicalness occurred at this point.

Typical contact person. In half of the sessions the confederate dressed and presented herself in a manner designed to fit subjects’ stereotypes of the out-group [...]. [...] Thus, the typical Douglass student exhibited three characteristics (conservatism, neatness, studiousness) that were found in pretesting and earlier research to be attributed to Douglass women by Rutgers women. When the confederate posed as a typical Rutgers student, [...] compared to the typical Douglass student, the typical Rutgers woman was portrayed to be more liberal, less neat, and less studious.

Atypical contact person. The confederate identified herself as a biological science major. When she hailed from Douglass College, the confederate dressed and behaved like the typical member from Rutgers College (as described above). When ostensibly a Rutgers student, she dressed and acted like the typical student from Douglas. [...] Once the introductions were completed, the experimenter led the participants to an adjacent room and asked them to sit in separate booths. [...] The subjects were in the “two-persons-heterogeneous condition” because each was a student from a different college. He encouraged subjects to cooperate and do their best because their performance would be compared with that of other sized groups. [...]. Then the subject and confederate were given a set of problem-solving tasks that required 40 min to complete (short essays, anagrams, riddles). [...]

Pleasant contact. The confederate gave the correct or most popular responses to 60% of the exercises. [...] In addition, the confederate wrote brief comments of approval about the subject’s performance when she succeeded or agreed (e.g., “I agree,” “good idea”) and encouraging comments in response to failure or disagreement (e.g., “We’ll do better next time,” “ I understand your opinion but feel differently”).

Unpleasant contact. The confederate succeeded on the same number of tasks. [...] But written reactions to the subject’s performance were unhelpful and denigrating (e.g., “poor idea,” “You blew it”).

After completing the tasks, subjects were scheduled to return for a second session the following day. The tasks were similar but fewer, requiring a total of 20 min. The purpose of the second session was to reinforce the subject’s initial impression of the out-group member. The accomplice’s behavior was the same in both sessions, so it would be difficult for subjects to dismiss her behavior on the first day to some transitory mood or chance event. [...] At the conclusion of the second session, subjects completed a questionnaire designed to assess their impressions of the out-group member and their reactions to the study. Then the experimenter asked them if they would like to earn an extra dollar by completing a short questionnaire about college life (“Survey of College Life”) for a colleague who taught at Livingston College (a third college in the Rutgers archipelago). In this manner the second questionnaire was dissociated from the subject’s and confederate’s colleges as well as the experimenter’s study. Acting as a positive model, the confederate quickly agreed to do so, and all subjects followed suit. The second questionnaire contained items soliciting opinions about the five colleges in the Rutgers system. [...] Finally, a fifth condition was included as a control or baseline. [...]

Dependent Measures Three sets of measures were of particular interest: ratings of the contact person (confederate), evaluations of the out-group as a whole, and stereotypes of the out-group. [...] On 7-point bipolar scales they rated the helpfulness of the out-group person, their desire to have the out-group member as a partner again, and how typical the out-group member appeared to be of her college. The second questionnaire, administered under the guise of a separate research project, contained three items evaluating the out-group as a whole. These were constructed as 11-point bipolar scales. [...]

EXPERIMENT 2 

A second experiment examined the effectiveness of contact with an out-group member who appeared to be very typical by confirming beliefs directly involving the subject’s in-group.

Method

Subjects Subjects were 33 female undergraduates from Rutgers College. [...]

Procedure

Typical-relevant condition. This condition was identical to the typical-pleasant condition of Experiment 1 with one qualification. In addition to displaying the “typical” characteristics described above, the contact person confirmed the belief that Douglass women thought themselves superior to Rutgers women. [...]

Typical-irrelevant condition. The procedure was the same as in the preceding condition with one modification. The target of the confederate’s unfavorable comparisons was Livingston College, not Rutgers College. Thus, the confederate’ behavior was the same in both the relevant and irrelevant conditions. [...]

Atypical-unpleasant-irrelevant condition (control). This condition was identical to the atypical-unpleasant condition of Experiment I. It was included to assess the effectiveness of the typicalness and pleasantness manipulations.

Dependent measures were the same as those employed in Experiment 1. [...]

EXPERIMENT 3

Method

Subjects Twenty female undergraduates were recruited from Rutgers College. [...]

Procedure The purpose of Experiment 3 was to examine why the typical contact manipulation had a greater impact on evaluations of the out-group in the pleasant condition than the atypical manipulation. Subjects were randomly assigned to either of those two conditions from Experiment 1: typical-pleasant or atypical-pleasant conditions. The procedure of Experiment 1 was followed to the end of the first day’s session. At that point subjects completed a questionnaire containing the dependent measures.

Dependent Measures The first item required subjects to list as much information about the other person (confederate) as they could remember [...]. Subjects were provided with 20 numbered blank lines to list information. [...] After completing this item, subjects turned to the second page and answered 9 questions constructed as 9-point scales. [...]