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 

Participants. Participants were [...] college students enrolled in introductory psychology courses [...]. Participants were preselected to have either high- or low-prejudice beliefs toward African Americans on the basis of their scores on the Pro-Black Scale [...], as assessed at a mass testing session at the beginning of the semester. High scores on the scale indicated that participants were high in favorableness toward African Americans and, thus, low in prejudice. High-prejudice participants were those scoring in the lowest 30% of the distribution, whereas low-prejudice participants were those scoring in the highest 30% of the distribution. [...]

[...]

Procedure. Participants were informed that the purpose of the study was to address how people think and talk informally about social groups [...] In addition, we mentioned that participants might be interested in the beliefs held by other college students from their university [...]. Specifically, we told participants that they would be given the average percentage of students that agreed with their responses [...]. After reading and signing a consent form, participants were asked to complete, by computer, the Pro-Black Scale [...].

After the participants had completed the last item [...] participants were randomly assigned to receive one of two possible types of feedback. Participants in the high-consensus condition were given information that 81% of [...] students agreed with their judgments. In the low-consensus condition, participants were given information that 19% of [...] students agreed with their judgments. [...]

[...] The experimenter reentered the experimental room and told the participants that they now would complete the procedure again but would make ratings pertaining to a different social group. While the experimenter supposedly prepared the computer program for the ratings of the second social group, a female African American confederate knocked on the door to the laboratory room and asked to talk to the experimenter about participating in the study. The researcher asked the student to wait in the hallway. The experimenter then began to start the second computer program but pretended that a malfunction occurred that would not allow the program to run. Participants were asked to have a seat in the hall until the experimenter was able to reload the program onto the computer.

The experimenter then led the participant to the hallway, where seven chairs were placed side by side; this part of the experiment allowed us to evaluate the behavior of the participants toward an African American confederate. The African American confederate, who was unaware of the experimental condition of the participants, sat in the seat closest to the door of the experimental room. The confederate recorded where the participant sat. Previous research has demonstrated that seating preference is a valid, unobtrusive measure of racial attitudes [...]. The behavioral measure was computed in terms of how many empty seats were between the African American confederate and the participant.

The experimenter reentered the hallway after 1 min to tell the participant that the program was reloading and informed the confederate that they could now speak about participating in the experiment. The confederate followed the experimenter into the laboratory room and left after 2 min. The experimenter then informed the participant that the experiment would continue. After returning to the experimental room, participants were told that the other part of the program was not working but that they would complete several more measures before being given their experimental credit.

The first questionnaire assessed participants' attitudes toward African Americans by asking participants to rate their perceptions of the percentage (0-100%) of African Americans who possess particular stereotypical traits. These traits included nine favorable traits [...] and nine unfavorable traits [...]. These traits were selected as stereotypical traits of African Americans on the basis of previous research [...].

Participants then completed a manipulation check, in which they were asked to indicate the percentage range of other [...] students who had agreed with their responses on the Pro-Black Scale. Participants made this decision by choosing one of five options: "0-25%," "25-50%," "50-75%," "75-100%," and "I do not remember." Participants also indicated how similar they perceived that their beliefs were to the beliefs of other [...] students on a scale that ranged from 1 (not at all similar) to 9 (extremely similar). [...]

Experiment 2

Method 

Participants. Forty-eight [...] students enrolled in an introductory psychology course [...].

Stimulus traits. On the basis of pilot testing in which European American participants were asked to indicate the percentage of African Americans who possessed each of a number of traits, we developed a list of 24 traits that were stereotypical of African Americans for use in the lexical decision task. The 24 stereotypical traits had a mean percentage rating of 62.80%. Of these 24 traits, 18 were randomly selected as stereotype feedback traits, and 6 were selected as stereotype nonfeedback traits. An additional 12 traits were selected as stereotypes that are primarily attributed to European Americans and not to African Americans [...]. [...]

Procedure. All participants were tested individually. Participants were told that the purpose of the experiment was to examine the attitudes and beliefs that people have about themselves and different social groups. Participants were told that the experiment they were participating in consisted of four smaller studies and that they would be completing all four parts. In Part 1, [...] participants were asked to answer a set of questions presented to them by computer. Participants were provided with nine favorable and nine unfavorable stereotypical traits of African Americans and were asked to rate the percentage of African Americans who possess each of the traits [...]. These traits [...] were pretested as being stereotypical of Blacks and were taken in large part from prior research. Both favorable and unfavorable traits were included to assess possible differences in responding due to the valence of the traits.

After completing this measure, participants were randomly assigned to receive either high- or low-consensus feedback about the beliefs of other [...] students. Using the same procedure as in Experiment 1, high- and low-consensus feedback were created by providing participants with the information that either 81% or 19% of [...] students, respectively, agreed with their responses. 

In Part 2 of the study, participants were asked to complete a short questionnaire, supposedly provided by another graduate student who was pretesting materials for a future experiment. Participants were given a list of 20 names that were stereotypical of African Americans or European Americans and were asked to indicate whether the individuals were Black or White on the basis of their names. Half of the names in the list were stereotypical African American male and female names [...]  and half the names were stereotypical European American male and female names [...]. The purpose of this task was to be certain that participants would associate the prime of black to be used in the subsequent lexical decision task with African Americans [...]. The name-rating task also served to provide a delay, such that there were approximately 7 min between receiving the consensus feedback and the beginning of the lexical decision task.

In the third part of the experiment, students participated in a lexical decision task. The purpose of this task was to examine the extent to which trait terms are linked to the associated category label [...]. Participants were told that the purpose of this part of the study was to examine how people recognize words visually. Participants were presented with a series of subliminal primes, each of which was followed by a target stimulus that was either a word or a series of letters that did not make a word. Three priming stimuli were used, each on one third of the trials: The word black, a neutral word [...], or a neutral nonword [...].

Each of the three primes was followed equally often by 1 of 72 target stimuli. [...] These target stimuli included the 18 stereotypical traits that had previously been rated, 6 additional stereotypical traits of African Americans for which consensus feedback had not been provided, 12 traits that were not stereotypic of African Americans, and 36 nonwords. Each prime-stimulus pair was repeated 3 times, for a total of 216 trials. Order of presentation was randomly determined for each participant. On each trial, the priming stimulus was presented for 15 ms and was immediately followed by a mask [...] for 2,000 ms. The target stimulus then appeared on the screen until the participant indicated his or her response. Participants were asked to indicate as quickly and as accurately as possible whether the sequence of letters appearing on the screen was a word. The D key indicated a no response, and the K key indicated a yes response.

For the fourth task in the study, participants completed manipulation checks and a suspicion check. The manipulation and suspicion checks were similar to those used in Experiment 1 but also included checks of whether participants could perceive the subliminal primes. Specifically, participants were asked whether they saw any words or letter sequences besides the stimuli that they were asked to respond to and, if so, to list the words and letter sequences they saw. Participants were also asked to identify, from a list of 16 items that included the prime words as well as distractors, the words or letter sequences that they might have seen. [...]