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 1a

Participants One hundred and four undergraduate students taking psychology courses (58 females) at UC Berkeley were recruited for compensation of course credit. Forty two of them (40.4%) identified themselves as White/Caucasians (22 females), 36 as East Asians (18 females), and the rest as other ethnicities including Southeast Asians, Latinos/Latinas, African Americans, and others (18 females). [...]

Procedure [...] Participants did the two sets of tasks (one with the physically disabled and the other with African Americans as the target outgroup) in random order. In each set of tasks, participants were first administered the prediction task, and filled out the questionnaire on their outlook of future interaction with the target outgroup, as well as on perceptions of the outgroup. After both sets of tasks, participants were asked about their subjective experience (i.e., perceived difficulty) during each of the prediction tasks. [...]

Prediction Task [...] Participants were asked to imagine themselves in the situations interacting with a person and to guess what the individuals would say in the specific circumstances. In both situations, the contexts were those of open-ended interactions between strangers. [...] Below the situation description were the prediction instructions: in the Few Predictions condition, participants were instructed to think of 4 different sentences, while participant in the Many Predictions condition had to generate 15 sentences. [...]

Questionnaire

As the main dependent variable, outlook of possible future interaction with members of the target outgroups was measured with four constructs: Expected Uncertainty in future interactions, Expected Quality of the interactions, Anticipated Anxiety, and Intention to Avoid the target outgroup members. In addition, I asked participants about their perceptions of the outgroups {Overall Attitude toward the outgroup and Perceived Similarity between the outgroup and participant's ingroup), to explore whether the difficulty manipulation also affects how the target groups are perceived, and whether perceptions correlate with future outlook. Participants also reported how difficult they found the prediction tasks to be. [...]

Outlook of future intergroup interactions. Expected Uncertainty was measured with 5 items from Gudykunst and Nishida's (1986) Attributional Confidence scale. Expected Quality was measured with items used in two studies: 6 items were adapted from scale of communication quality in the Iowa Communication Record questionnaire [...], and the other 6 items were from Gudykunst and 13 Shapiro (1996), measuring positive expectations of future intergroup interaction. [...] For Anticipated Anxiety, participants rated how much they would feel each of a series of positive or negative affects when interacting with a target group person. [...]

Perceptions of outgroups. Overall Attitude toward the outgroup was measured with two items, and Perceived Similarity was measured with an item adapted from a 3- item scale by Stephan and Stephan (1985).

Perceived Difficulty. Participants were asked to recall the experience in the prediction tasks, and rate the Perceived Difficulty of the tasks using 4 items. [...]

Study 1b

Participants One hundred undergraduate students taking psychology courses at UC Berkeley (76 females) were recruited for compensation of course credit or money ($10). Thirty one of them identified themselves as East Asians (22 females), 27 as White/Caucasians (20 females), 22 as Southeast Asians (18 females), and the rest as other ethnicities including Latinos/Latinas, and African Americans. As in Study la, African American participants (four in all) were excluded from analysis on the task with African Americans as the target outgroup. Participants were randomly assigned to either Cognitive Load or No Load condition.

Procedure The procedure was identical to that of Study la, except for the perceived difficulty manipulation. After the initial description of the study, participants were also told: [Verbal Stimulus A]. Then, in each of the sets of tasks for the two outgroups, participants first read the description of the situation. Subsequently, participants were given either a 10-digit (in the Cognitive Load condition) or a 3-digit (in the No Load condition) number and were told they would be asked to regenerate it after the imagination task. [...]

Study 2

Participants One hundred and six UC Berkeley undergraduates taking psychology courses (80 females) participated in the study for course credit. Forty seven of them identified themselves as White/Caucasians (38 females), 30 as East Asians (22 females), 8 as South Asians (4 females), and the rest as other ethnicities including Latinos/Latinas, Southeast Asians, African Americans, and others. It was a 2 (Group Description: Predictable or Unpredictable) by 2 (Prediction or No Prediction) factorial design, and participants were randomly assigned to one of the four conditions. [...]

Prediction Task Participants were given the following instruction: [Verbal Stimulus B]. In each trial, participants read the situation descriptions, wrote the imagined reactions, and clicked on a button on the screen to proceed to the next trial. [...]

Procedure Participants first read the description about the target group. Then participants in the Prediction conditions performed the prediction task and subsequently reported their perceived difficulty of the task; those in the No Prediction conditions skipped these two tasks. [...] The Perceived Difficulty measure consisted of the four items used in the previous studies with one more item added: "How vague or vivid were the imagined reactions in your mind?" [...]. After that, all participants rated the future outlook items, which were the same items as in the previous studies. [...]

Study 3

Participants Participants were 86 UC Berkeley undergraduate students (58 females) taking psychology courses recruited for compensation of course credit. Thirty-six of them (41.9%) identified themselves as White/Caucasians (25 females), 21 as East Asians (14 females), and the rest were of other ethnicities including Southeast Asians, Latinos/Latinas, African Americans, and others (19 females). Participants were randomly assigned to one of the two conditions: one in which the situations in the prediction task were relatively more difficult to make predictions about (Difficult Situations condition), and one in which the situations were easier to make predictions about (Easy Situations condition). [...]

Questionnaire Questionnaire items for both perceived difficulty and future outlook were the same as the ones used in Study 2, and scores were aggregated for each measure in the same way. Positive and negative state affect were measured with the 10-item International Positive and Negative Affect Schedule Short Form [...].

Procedure [...] Subsequently, participants performed the prediction task, which consisted of 8 trials with either difficult or easy situations. Participants were then asked about their experience of difficulty during the prediction task, rated their outlook of future interaction with members of the target group, and reported their affective states. After this, I asked participants to think of the target group once again and report their impression of the group: participants rated their impression {Overall Impression) on a 6-point scale from Very Negative to Very Positive. [...]

Study 4

Participants Participants were 94 undergraduate students (56 females) taking psychology courses at UC Berkeley recruited for compensation of course credit. 31 of them (33.0%) identified themselves as White/Caucasians (18 females), 28 as East Asians (15 females), and the rest as other ethnicities including Southeast Asians, Latinos/Latinas, African Americans, and others (23 females). Participants were randomly assigned to one of three conditions: one in which knowledge about the group was helpful in prediction (the Knowledge Applicable condition), one in which it was not helpful in prediction (the Knowledge Inapplicable condition), and lastly, one in which the provided knowledge was not informative about the target group (the Knowledge Uninformative condition).

Descriptions of Target Group Members [...] Participants then read descriptions about 14 members of the group. The descriptions took the form of direct quotes from the group members describing their own behaviors; this self-description format was used to give an authentic feeling. The descriptions appeared on the screen sequentially, and each description was presented with the name of the person. [...]

Prediction Task In the prediction task, participants made predictions in a succession of trials. Besides the fact that each situation pertained to a personality trait, there were other aspects different from previous studies. One was that each interaction was between a target group member and another third person, that is, participants did not imagine themselves interacting with target group members. Also, the target group members' reactions that participants had to predict included thoughts as well as speech. [...] In the Knowledge Applicable condition, the situations were those that pertained to the traits Thrifty and Perfectionistic, which were used in the group description. In the other two conditions, the situations pertained to the traits of Soft-hearted and Defiant. [...]

Questionnaire Several additional measures were added. Prediction Confidence was measured with three items [...] as another potential mediator. Then participants were asked about their Perceived Knowledge with two items [...]. Lastly, participants were given an open-ended question that asked them to describe the group; this was given to verify that participants in Knowledge Applicable and Inapplicable conditions formed group impressions that pertained to the two personality traits.

Procedure [...] After participants learned about the target group, they performed the prediction task, and answered questionnaire items probing their subjective experience of the prediction task and their outlook of future interactions with the outgroup members. [...]