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

Design and Participants Participants were assigned to one of nine conditions in a 4 X 2 design with an additional control condition. Factors were attribution information to explain a target's counterstereotypic behavior (no information, internal-stable, internal-unstable, and external) and the representativeness of the target (typical or atypical of other group members). Participants in the control condition did not view any counterstereotypic behavior. Participants were 180 undergraduates enrolled in introductory psychology and social psychology classes at Rutgers University. [...]

Procedure [...] Participants would be presented with information about a group of persons from one of a variety of sources including written statements, audio interviews, and video presentations. They were to examine the information and form an impression of the group. Participants were led to believe that the stimulus persons were members of a real campus group whose name was hidden to maintain confidentiality. [...]

Part 1: Creation of group stereotypes. On the second page they found the first of 12 descriptions of members of a fictitious campus group that met weekly. Each description appeared on a separate sheet of paper; six sheets described women and six described men. [...] This information was constructed to foster a homogeneous impression of the group. All members were either 19 or 20 years old; were business or English majors; and spent their spare time listening to rock music, playing tennis, shopping, or hanging out with friends. Below the demographic information, participants encountered a list of five personality traits that characterized the member along with a behavioral example of that trait. [...]

Part 2: Presentation of target. [...] Then they read a transcript of an interview conducted with a male member of the out-group. [...] He was asked to describe himself briefly (manipulation of typicalness) and to recount his activities of the preceding day. Embedded in his description were behaviors that actively disconfirmed three of the six stereotypes participants had just acquired about his group. The apparent explanation for his counterstereotypic statements differed among experimental conditions (attribution manipulation).

Manipulation of typicalness. On the basis of his brief self-description, the target appeared to be either like the others in the group or quite different from them. In the typical conditions he stated that he was 19; a business major; and liked to listen to rock music, play tennis, and hang out with friends. In the atypical conditions he stated that he was a returning student of 26; a French language major; and enjoyed classical music, gardening, and time alone.

Manipulation of attributions. The interviewer asked the target to recount activities from the preceding day. The target's recollections were designed so that he would appear to be unfriendly, messy, responsible, creative, not persistent, and dull. [...] Overall, the target exhibited four negative actions and two positive ones. [...] The target offered an explanation of his behavior for each of the three incidents that disconfirmed expectations. The reasons he provided differed among conditions. Participants were randomly assigned to one of five attribution conditions.

Internal-stable (dispositional) attribution. When the target described a counterstereotypic action (e.g., unfriendliness), he also mentioned that his action reflected the kind of person he is. [...]

Internal-unstable (motivational) attribution. The target indicated that his actions reflected the effort he chose to put into those activities on that particular day. [...] By contrast, in the internal-stable condition, he attributed the same behavior to his personality ("That's the way I am").

External attribution. The target's explanations for his actions focused on environmental causes. [...] His other behaviors that disconfirmed group stereotypes were similarly accounted for by environmental presses.

No attributions. The target reported the same behaviors as in the preceding attribution conditions. However, he made no attempt to explain or give reasons for his actions. [...]

After participants had read the interview, they completed a questionnaire containing the dependent measures. [...]

Baseline condition. A single control condition was added to the 4 X 2 design in which participants did not encounter a counterstereotypic group member. After reading the profiles that established group expectations (Part 1), they were given an irrelevant task (solving anagrams) to work on for a period of time equal to Part 2. Then they completed the dependent measures.

Dependent Measures On the first page participants were asked to rate the out-group on six 9-point scales: responsible-irresponsible, neat-messy, witty-dull, creative-conventional, friendly-unfriendly, and persistent-not persistent. [...] These measures assessed the extent to which participants endorsed the stereotypes associated with the out-group. [...] Participants also completed a set of manipulation checks that required them to make attributions for the target's behavior. On three measures they indicated the extent to which they thought the actions of the target were due to personality characteristics (internal-stable), motivation or effort (internal-unstable), and environmental aids or obstacles (external). Each of these measures was constructed as a 9-point scale so attributions could range from not at all a reason for his actions (1) to the principle reason for his actions (9). Finally, to assess the effectiveness of the typicalness manipulation, participants rated how typical overall the target appeared to be of others in the out-group. Endpoints of this 9-point scale ranged from not at all typical; very unlike others in the group (1) to very typical; very much like others in the group (9).

Experiment 2

Method

Design and Participants Experiment 2 repeated the 4 X 2 design of Experiment 1. In addition, the control condition was borrowed from Experiment 1. [...] Participants were 160 undergraduates recruited from the same courses as the sample used in Experiment 1.

Procedure The procedure and measures were identical to those of Experiment 1, but with one major change. In both experiments, after participants had been exposed to the information that created the six out-group stereotypes, they heard another group member describe a day's activities in which his actions disconfirmed three of the six expectations. In Experiment 1 the exception behaved in a manner that was unfriendly, messy, and responsible (disconfirmed stereotypes) and creative, not persistent, and dull (confirmed stereotypes). In Experiment 2 he behaved in a manner that was friendly, neat, and irresponsible (confirmed stereotypes) and uncreative, persistent, and witty (disconfirmed stereotypes). Thus, the target in Experiment 2 supported the traits that were disconfirmed in Experiment 1 and contradicted those that were confirmed in the first study. Dependent measures were carried over from Experiment 1. Data from all trait ratings were coded so that larger means indicated a response more in agreement with out-group stereotypes.

Experiment 3

Method

Design and Participants One hundred and seventeen participants (60 men and 57 women) were recruited from introductory psychology classes. [...] Participants were randomly assigned to 1 of 10 conditions in a 5 X 2 design. Factors were Attribution Information (internal-stable typical, internal-stable atypical, internal-unstable, external, and no information) and Participant Gender.

Procedure The experimenter informed participants that he was collecting norms of student performance on a range of academic issues. [...] After completing the tasks, the experimenter would provide feedback about their performance. [...] The experimenter crossed into an adjoining room and turned on a tape recording allegedly of an interview with another participant in the project. Although he closed the door between the rooms, participants were able to hear the content of the interview because an intercom in their room had been left on. [...] After participants had overheard the taped interview for 2 min, the assistant entered the room carrying an armful of questionnaires. He showed surprise at the "live" intercom, switched it off, and then distributed the first task to the participants. The first task contained the dependent measures. [...] The attribution manipulation was contained in the content of the overheard interview. In all conditions the recording began with the experimenter's voice reviewing background information about the interviewee or target, including a fictitious case number ("46"), year in college ("sophomore"), and college affiliation ("Princeton University"). The experimenter proceeded to render an evaluation of the interviewee's performance on the academic tasks. At this point participants in each of the five conditions heard a different dialogue.

Internal-stable typical condition. The experimenter noted that the interviewee appeared to be a fairly typical Princeton student based on a survey of interests and activities he had completed prior to participation in the study. [...] After a short pause during which he rustled a few sheets of paper, the experimenter stated that the interviewee had performed quite poorly on most of the problem-solving and verbal reasoning tasks. [...] The interviewee responded that these results did not surprise him. He had never performed well on those sorts of tasks. Past performance had convinced him that he had low ability in those academic areas. [...]

Internal-stable atypical condition. In this condition the interviewee was presented as a fairly atypical Princeton student. [...] The experimenter continued in the manner described in the internal-stable condition.

Internal-unstable condition. This condition was identical to the internal-stable typical condition up to the interviewee's explanation of his performance. He stated that he was quite surprised because poor performance on these sorts of tasks was unusual for him. He added that he had difficulty maintaining high effort because he had not been feeling well during the experiment. [...]

External condition. In this condition the interviewee attributed his poor performance to environmental obstacles. He stated that noise from an adjacent room had distracted him repeatedly while performing the tasks. [...]

No-attribution condition. The first part of the audiotape was the same as in the internal-stable typical, internal-unstable, and external conditions in that the experimenter gave the interviewee negative feedback about his performance. However, the target declined when asked if he had any comments to make. [...]

Dependent Measures The questionnaire was entitled "Survey of Opinions of New Jersey College Students" and included 15 items assessing the participants' opinions on a range of issues from state funding for higher education to perceptions of fellow students at Rutgers and other colleges in the state. Embedded in these measures were three questions concerning Princeton University. Each of these items was constructed as a 9-point scale. The first one asked participants to rate the intelligence of students at Princeton in comparison with college students in general. Endpoints were labeled much less (1) and much more (9), with the midpoint labeled same (5). The second items asked them to estimate the wealth of Princeton students in comparison with other college students. Endpoints were labeled much poorer (1) and much wealthier (9), with the midpoint labeled about the same (5). For the third item, participants predicted the future success of a Princeton graduate in comparison with other college graduates. Endpoints were labeled not as successful (1) and much more successful (9), with the midpoint labeled as successful (5). [...]