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

Experiment 1 explored whether emphasizing common American identity, compared to making different racial identities salient, affects Whites' sensitivity to the existence of racial biases against Blacks that constitute barriers to their advancement in society. [...]

Participants Participants were 118 White residents of the United States (41 women, 76 men, and 1 unspecified) drawn from Amazon Mechanical Turk's online pool [...]. Participants' average age was 27.5 years (SD = 7.4).

Procedure [...]

Threat manipulation. First, participants were randomly assigned to read one of two threat conditions. In the threat condition, participants read excerpts from an article in the Huffington Post, which indicated that the U.S. ranks 17th in the world for education. In the no-threat condition, participants read the identical article, except that it indicated that Canada ranks 17th for education, with no mention of the U.S. Group identity manipulation. Next, participants were randomly assigned to one of two group-identity conditions. Participants in the common-identity condition [...] read an ostensible excerpt from USA Today titled “Celebrating our American Identity,” in which the author explained that despite the diversity in America, its people are united by their shared values [...]. In the separate-identities condition, participants read an article titled “Celebrating our Ethnic Identities,” in which the author instead explained that there are many distinctive identities in America, each with their own values and strengths [...].

Discrimination scenario. Participants read that 60% of White Americans pay off their student debt within 2 years of graduation, but only 25% of Black Americans do so. Participants further read that [Text Stimulus A]. Thus, the reason why Black Americans pay off their debt more slowly was left ambiguous.

Perceptions of bias. Participants rated their agreement (1 = strongly disagree to 7 = strongly agree) with four items indicating that Black Americans are discriminated against on the job market. [...] Finally, participants indicated the overall extent to which they thought that “Black Americans pay off their debt more slowly than White Americans because of discrimination against Black Americans” (1 = not at all due to discrimination to 7 = completely due to discrimination). [...]

Experiment 2

[...] In Experiment 2, we systematically vary ambiguity of the discrimination to examine how it moderates the effects of common-identity on perceptions of racial bias. [...]

Participants Participants were 158 White residents of the United States (97 women, 60 men, and 1 unspecified) drawn from Amazon Mechanical Turk's online pool [...]. Participants' average age was 33.5 years (SD = 12.6).

Procedure [...]

Group identity manipulation. First, participants were randomly assigned to one of three group-identity conditions. [...] Participants in the common-identity condition, read an ostensible excerpt from USA Today titled “Celebrating our American Identity,” in which the author extolled the virtues of emphasizing the shared American identity, rather than specific ethnic identities [...]. In the separate-identities condition, participants read an article titled “Celebrating our Ethnic Identities,” in which the author instead extolled the virtues of emphasizing racial/ethnic identities [...]. In the control condition, participants did not read any excerpt.

Discrimination manipulation. [...] Participants reviewed a brief job description for an entry-level Human Resources Consultant and summaries of the resumes of two candidates for the job. [...] the first candidate had a degree in Marine Biology from the College of New Jersey, some experience in retail, and a few extracurricular activities while in college. The second candidate had a degree in African-American Studies from Princeton, several internships, and extracurricular activities. His resume was designed to indicate that he was better qualified for the position. Although the race of the candidates was not mentioned, the second candidate's activities implied that he was African American [...]. Participants then read the evaluations of each candidate made by the Human Resources Officer. In both conditions, the Officer recommended hiring the first candidate, but not the second candidate. Our manipulation of subtle versus blatant discrimination [...] involved the reasons the Officer gave for not hiring the second candidate. In the subtle-discrimination condition, the Officer indicated that the second candidate had no human resources experience and his degree was not in business or economics – credentials that the White candidate also did not possess but were mentioned only for the Black candidate. In the blatant-discrimination condition, the Officer also noted his degree was not in business or economics, but also referenced the candidate's race [...].

Perceptions of bias. Participants rated their agreement (1 = strongly disagree to 7 = strongly agree) with two items indicating that the decision made by the Human Resources Officer was biased [...]. We also asked, “In the evaluation of Candidate 2, to what extent were the comments made by the human resources officer prejudiced?” Participants responded on a 1 (not at all prejudiced) to 7 (extremely prejudiced) scale. [...]

Willingness to protest. Participants rated their likelihood (1 = not at all likely to 7 = completely likely) of protesting the Human Resource Officer's decision [...].

Prejudice. After completing the other measures, participants rated their agreement (1 = strongly disagree to 7 = strongly agree) with four items [...] assessing individual differences in prejudice [...].

Experiment 3

[...] We conducted a study to further examine whether a dual-identity representation would be more effective than a common-identity representation in promoting expressions of willingness to protest blatant discrimination against Blacks. [...]

Participants Participants were 77 White residents of the United States (43 women and 34 men) drawn from Amazon Mechanical Turk's online pool [...]. Participants' average age was 35.1 years (SD = 12.0).

Procedure [...]

Group identity manipulation. As in Experiment 2, participants either did not read a newspaper article (control condition) or read a newspaper excerpt that emphasized the importance of common identity as an American. In a third condition, participants read an article [...] designed to make a dual identity salient. The article stated, [Text Stimulus A]. Participants then read the blatant discrimination hiring situation from Experiment 2.

Prejudice. Participants completed the same 4-item prejudice measure used in Experiment 2 [...].

Dependent measures. Using the same items as Experiment 2, participants reported perceptions of discrimination [...] and willingness to protest the decision [...].