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.

Study 1
We proposed that White participants confronted for using negative Black stereotypes would use fewer negative Latino stereotypes 1 week later compared to nonconfronted participants.
Method

Procedure

T1 involved a 30-min session, while T2 involved an unrelated 5-min online survey that they would receive 1 week after T1 and have 48 hr to complete. [...] During T1, participants completed a stereotyping task used to elicit negative stereotypes about Black Americans. [...] Participants saw 16 trials in which images of White and Black, men and women [...] were paired with a descriptive sentence, and participants’ task was to make an inference about the individual. Three critical trials presented Black men with descriptive sentences intended to evoke a stereotypical response (e.g., “This person can be found behind bars”; stereotypical response: “criminal”) but could evoke a neutral response (“bartender”).
Participants responded aloud for the experimenter to record. After all trials, based on random assignment, participants were confronted for their stereotypical answers by the experimenter or not [...]. In the confrontation condition, the experimenter said, “I thought some of your answers seemed a little offensive. The Black guy behind bars could be a bartender. People shouldn’t use stereotypes, you know?” The second sentence was tailored to participant’s stereotypical responses, and experimenters were trained to deliver the confrontation in a spontaneous, genuine, and neutral manner. If participants responded, experimenters were instructed to say “Okay” and continue the study.
Next, participants completed filler tasks followed by an affect measure and received T2 instructions. [...]
One week later, participants received the T2 survey via email. T2 included a modified version of the T1 stereotyping task with 15 randomized trials of White, Black, and Latino male faces and descriptive sentences. Eight trials included neutral descriptive sentences. The remaining consisted of three negative Latino stereotype trials (e.g., This person spends a lot of time at shelters; stereotypical response: homeless; neutral response: volunteer), two new negative Black stereotype trials (e.g., This person takes peoples’ cars; stereotypical response: car thief; neutral response: valet driver), and two negative Black stereotype trials from the baseline task. [...]

Study 2

Study 2 procedures were nearly identical to Study 1 although included one time point. After completing the Study 1 confrontation task, participants completed filler inference tasks (about 15–20 min) and Study 1 affect measures [...]
Finally, participants completed the stereotyping task administered in Study 1 at T2. This task included 24 trials (7 critical): 3 gender role stereotype trials with images of White women (e.g., This person works at a hospital; stereotypical response: nurse; nonstereotypical response: doctor; M. Burns, 2017), 2 positive Black stereotype trials (e.g., This person is on a team; stereotypical response: athlete; neutral response: debate team), and 2 positive Asian stereotype trials (e.g., This person has a scholarship; stereotypical response: smart student; neutral response: athlete). [...]

Study 3

Participants were informed they would complete a survey on moral decisions by communicating with a randomly selected MTurk worker. Participants learned their partner was a 35-year-old White man, and they would discuss four moral scenarios. During the first two scenarios, their partner would recommend action after an immoral act occurred, and they would respond to their partner’s recommendation, and vice versa for the last two scenarios (based on Mallett & Wagner, 2011). Unknown to participants, their partner’s responses were preprogrammed.
The scenarios involved moral indiscretions (e.g., a student cheating on an exam), and the critical fourth scenario was: “A nurse discovers a hospital patient has been given blood contaminated with the AIDS virus” (Mallett & Wagner, 2011). After submitting a recommendation, participants randomly assigned to the confrontation condition received, “I noticed you referred to the nurse as a ‘she.’ The nurse could also be a man. We shouldn’t use stereotypes, you know?” In the control condition, participants received, “I think the nurse should take care of the patient right away before worrying about who to blame.”  [...]
Next, participants were asked a series of open-ended questions probing for suspicion and responded to 3 items assessing how much they liked their partner ( [...], e.g., “Did you enjoy the interaction with the other participant?,” [...] On the same scale, participants responded to, “Was the other participant rude?” and “Did the other participant anger you?” as a brief measure of neg-other,  [...], and a modified 9-item measure of neg-self  [...], 1 (does not apply to me at all) to 7 (very much applies to me).
After 24 hr, the T2 survey was made available on MTurk to participants who completed T1. T2 was presented as examining inferences and did not mention T1. The only connection between T1 and T2 was the HITs (Human Intelligence Tasks) were posted by the same requester account. During T2, participants completed 1 practice trial followed by the stereotyping task (Studies 1–2) with 24 trials including 3 negative Black and 3 negative Latino stereotype trials (Study 1). Later, participants were asked, “Over the last 24 hours, how often did you find yourself . .” followed by 2 racial egalitarian motivation items, for example, “Focusing on being egalitarian toward people regardless of their race,”  [...], and 2 racial rumination items, for example, “Thinking about how often you use racial stereotypes,”  [...], on a scale from 1 (not at all) to 7 (very much).  [...]