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

We manipulated White Canadians’ sense of social connectedness to a Chinese Canadian confederate using nonverbal mimicry, an established manipulation of social connectedness. We tested whether the nonverbal mimicry would increase participants’ interest in Chinese culture.

Method

Participants. Fifty-one (28 female, 23 male) White Canadian undergraduates participated [...].

Prestudy survey. Several weeks prior to the experiment, in an ostensibly unrelated survey, participants indicated how positively or negatively they viewed people from China, Australia, and Britain; responses were made using feeling thermometers from 0 to 100. [...] Participants also reported how interested they were in Chinese culture and in the two Western (Australian and British) cultures, using scales from 1 (not at all) to 7 [...].

Procedure. Participants took part individually in a study that was ostensibly on interactions and worldviews; each participant was paired with one of two extensively trained Chinese Canadian confederates posing as another participant. The experimenter gave the dyad a series of questions to ask each other and then left the room. The questions and the confederate’s responses were scripted so that the confederate described both mainstream experiences (e.g., academic interests) and his or her Chinese cultural background (e.g., favorite Chinese foods). [...] The confederate either mimicked the participant’s body posture, leg position, and arm position following a lag of several seconds (mimicry condition) or held a neutral position (no-mimicry condition). Confederates were unaware of the hypothesis and purpose of the manipulation. [...]

Dependent measures. After the conversation, the experimenter separated the dyad and gave the participant a questionnaire. First, as a manipulation check, we assessed perceived similarity, a key basis of interpersonal liking. Participants reported how much they had in common with the confederate, using a scale from 1 (nothing) to 7 (very much), and the degree to which they were similar to the confederate, using a scale from 1 (not at all) to 7 [...]. Next, participants reported how interested they were in Australian culture, British culture, and Chinese culture, using scales from 1 (not at all) to 7 (very), and how much they liked Chinese art and Chinese movies, using a scale from 1 (not at all) to 7 (very much). [...] Finally, participants were invited to complete lottery tickets to win Chinese cultural products (e.g., Chinese films) ostensibly left over from a prior study. [...]

Experiment 2

The results of Experiment 1 showed that a small cue of social connectedness with a member of an out-group caused people to express greater interest in that group’s culture. Does enacting such cultural interests reduce prejudice? We tested this question in Experiment 2, in which non– Latino American participants freely took part in a cooperative cultural activity with a Latino American (Mexican American) peer. [...]

Method

Participants. Fifty-eight non–Latino American female undergraduates (31 White American, 23 Asian American, 4 other race or ethnicity) participated [...]. We recruited only women to create same-sex pairs with female confederates [...].

Prestudy survey. Several weeks before the experiment, participants reported five idiosyncratic interests (e.g., their favorite book) in an ostensibly unrelated mass survey.

Procedure. [...] Each participant was paired with one of two extensively trained Latina American confederates posing as another participant. First, the experimenter orchestrated a brief getting-to-know-you conversation by alternately asking each member of the pair questions. In all conditions, the confederate revealed that she was Mexican American. Then, following a personalized script for each participant, the confederate was queried about three idiosyncratic interests (e.g., favorite book). In the social-connection condition, the confederate described having one interest that the participant had reported in the prestudy survey [...]. In the no-social-connection condition, each participant was yoked to a social-connection-condition participant, and the confederate described having an interest reported by this participant. Thus, across conditions, the confederates expressed the same types of interests; what was manipulated was whether one interest did or did not match one of the participant’s interests. [...] Next, the experimenter told the dyad that they would design a music video for a new pop song and that they could choose what band to use music from. To reinforce the sense of free choice, we presented the pair with a fictitious Canadian band, which the confederate and all participants rejected. The experimenter then read the name of a second band, Camila, an actual Mexican band, and was interrupted by the confederate, who excitedly said that she liked Camila and that Camila was popular in either Mexico (Mexican condition) or Portugal (Portuguese condition). The experimenter confirmed the group’s origin and asked the confederate and participant if they wanted to use Camila. All participants agreed to do so. The dyad then read about Camila’s current (fictitious) tour in either Latin America or Europe, viewed a Camila music video, listened to a new Camila song, and read a synopsis of the song’s theme. [...] The confederate’s contributions were semiscripted so that in all conditions, she provided relevant cultural information— for instance, suggesting that the video include a traditional Mexican or Portuguese (depending on condition) dance [...].

Dependent measures. After the dyad had completed the music task, the experimenter separated the dyad and asked the participant to complete “poststudy” measures. The first measure was a White-Latino lexical Implicit Association Test. Participants sorted words into four categories (bad, good, White surnames, and Latino surnames) by pressing one of two keys. [...] Finally, as a manipulation check, we had participants report how much they had in common with their interaction partner and how similar they were to their interaction partner, using 7-point scales [...].

Experiment 3

To further explore the role of consistency processes in prejudice reduction, we manipulated whether the choice to design a music video for a Mexican band was represented as free or required. Consistency theories predict that only behaviors that seem freely chosen should create the feeling of behavioral-attitudinal inconsistency that causes attitude change. [...]

Method

Participants. Fifty-eight non–Latino American female undergraduates (26 White American, 25 Asian American, 7 other race or ethnicity) participated [...].

Procedure. In the free-choice condition, the procedure, including the social-connection manipulation, was identical to that in the Mexican condition in Experiment 2. The procedure was the same in the no-free-choice condition except that after the confederate expressed enthusiasm for Camila, the experimenter paused and, instead of asking the dyad if they would like to use Camila, said that the study protocol required them to do so. The music task and dependent measures were identical to those in Experiment 2 except for the inclusion of measures assessing participants’ cultural-task engagement. After each session, the confederate rated how enthusiastic and interactive the participant had been during the music task, using scales from 1 (not at all) to 7 (very) [...]. In addition, two coders, who were unaware of the conditions and hypotheses, rated participants’ contributions to the music-task worksheet [...] on six measures (e.g., “detailed/vivid,” “enthusiastic”), using scales from 1 (not at all) to 5 [...].

Experiment 4: Long-Term Follow-Up

Does the experience created in Experiments 2 and 3 improve intergroup outcomes along broader measures and in the future? [...]

Method

Participants. Because some participants from Experiments 2 and 3 had graduated, we estimated that we had valid contact information for 91 remaining participants. Of these, 60.44% (n = 55) responded to our communications and completed an online survey [...].

Dependent measures. Embedded among various ostensibly unrelated surveys were items assessing how much participants would “enjoy,” “feel enthusiastic about,” “want to learn from,” and “try to avoid” (reverse-coded) talking with a Mexican American student about (a) Mexican cultural traditions and practices [...] and (b) U.S. immigration policy and undocumented Mexican migrants [...]. Participants responded to all items using scales from 1 (not at all) to 7 (very much). [...]