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

On January 27, 2017, President Donald Trump issued a controversial executive order temporarily banning individuals from several predominantly Muslim countries from traveling to the United States. [...] In line with our model, we hypothesized ingroup interconnectedness would be positively associated with support for the executive order through less empathy toward immigrants, whereas outgroup interconnectedness would be negatively associated with support for the executive order through greater empathy toward immigrants.

Participants and procedure. Three hundred Americans recruited from Amazon Mechanical Turk participated in an online study about public opinion and social perception for $0.70. [...] Three participants were excluded for failing an attention check (e.g., “This is an attention check item. Please select ‘at least once’ for your answer”), leaving 297 participants for analyses (54% female, 72% White, M_age = 31.10 years old, SD = 9.68). After providing informed consent, participants completed randomized measures (within and between block; see following lines), along with several additional measures unrelated to the present study.

Measures

Intergroup interconnectedness. To represent our conceptualization of intergroup interconnectedness as the extent to which various ingroups and outgroups are included in the self-concept, we used a modified version of the Inclusion of Other in the Self Scale. Ingroup interconnectedness was measured via ratings of perceived connectedness between the self and five ingroups (“your family,” “your friends,” “students at your alma mater university,” “your ethnic group,” “your nationality”), whereas outgroup interconnectedness was measured via ratings between the self and four outgroups (“students at other universities,” “ethnic groups different from your own,” “cultures different from your own,” “nationalities different from your own”). [...]

Empathy toward immigrants. [...] This measure included nine items [...] primarily taken and modified from the Empathic Concern subscale of the Interpersonal Reactivity Index. All responses were made on 5-point scales (1 = not at all like me, 5 = very much like me) [...].

Support for executive order on immigration. [...] After being told the executive order has been “controversial in recent weeks,” we asked participants to share their thoughts regarding the extent to which they disapproved/approved of the executive order (1 = strongly disapprove, 9 = strongly approve). [...]

Study 2

In Study 2, to test the generalizability of our findings and provide a stronger test of statistical mediation, we broadened our operationalizations of empathy and political ideology. Here, we examined whether ingroup and outgroup interconnectedness predict liberal and conservative ideology through generalized empathy. [...]

Participants and procedure. [...] By the time data collection ended, 296 undergraduates at the University of Arkansas had completed an online study about public opinion and social perception [...]. [...] This left 261 participants for analyses (69% female, 84% White, M_age = 19.93 years old, SD = 3.89). After providing informed consent, participants completed randomized measures (within and between block; see following lines), along with several other measures unrelated to this study. [...]

Measures

Intergroup interconnectedness. With the exception of changing one item to “students at the University of Arkansas” (to reflect the undergraduate sample), we used the same measures of ingroup [...] and outgroup [...] interconnectedness from Study 1 [...].

Empathy. To provide a broader operationalization of empathy, we used two different measures. The first was Davis’s (1983) Empathic Concern subscale of the IRI [...]. The second scale referenced six adjectives [...] argued to represent the core of empathy, that were framed as traits representing the self [...]. Both measures were scored on 5-point scales (1 = not at all like me, 5 = very much like me). [...]

Political ideology. Political ideology was measured using a modified version of Kerlinger’s (1984) Social Attitudes Statements Scale (SASS), consisting of separate Liberalism [...] and Conservatism subscales [...].

Studies 3a and 3b

[...] In our final two studies, we directly manipulate perceptions of intergroup interconnectedness and measure its effect on empathy and political ideology. [...] The primary goal of Study 3b was to attempt to replicate the effects of Study 3a with a different operationalization of political ideology. Besides this difference, Studies 3a and 3b were nearly identical.

Study 3a 

Participants and procedure. [...] 170 undergraduates from the University of Arkansas participated in exchange for partial fulfillment of an introductory psychology course requirement. [...] This left 145 participants for analyses (68% female; 86% White; M_age = 19.15 years old, SD = 0.96). [...] Half of participants were randomly assigned to an outgroup interconnectedness condition, whereas the rest (the control condition) proceeded directly to the dependent measures presented in the following order: outgroup empathy, liberal and conservative ideology, manipulation check, and demographics. [...]

Manipulating outgroup interconnectedness. To manipulate the perception of outgroup interconnectedness, we first familiarized participants with the concept of intergroup interconnectedness using descriptions and a picture of two overlapping circles labeled “Self ” and “Group.” We then guided participants through concrete examples of outgroups and asked them to select and write about why they perceived a sense of interconnectedness with three outgroups from their own life. For those in the control condition, the concept of interconnectedness was not mentioned, and the pictorial representation was not provided; participants were merely told we were interested in learning how people think about outgroups. Control participants were also asked to select three outgroups from their own life, but they did not write about how they perceived a sense of interconnectedness with these groups. [...]

Outgroup empathy. Empathy was measured using a condensed, 10-item version of the empathy scale from Study 2, but the items were framed to be about the outgroups participants selected on the previous page [...]. The 10 items were each answered on a 7-point scale (1 = strongly disagree, 7 = strongly agree) [...].

Study 3b

Participants and procedure. We posted 215 slots, for which 212 undergraduates from the University of Arkansas participated in exchange for partial fulfillment of a course requirement. [...] Because three participants both failed the attention check item and did not follow instructions, a total of 27 participants were removed, leaving 183 participants for analyses (70% female; 85% White; M_age = 19.32 years old, SD = 1.73). [...] Participants were randomly assigned to the same control or outgroup interconnectedness condition as in Study 3a, and completed measures presented in the following order: outgroup empathy, political policy support, manipulation check, and demographics. [...]

Manipulating outgroup interconnectedness. With the exception of removing gender as an outgroup category, the control and outgroup interconnectedness conditions were identical to Study 3a.

Outgroup empathy. The same 10-item measure of outgroup empathy from Study 3a was used [...].

Political policy support. We consulted previous research for a measure of political policy support of welfare programs, an issue that tends to be favored more by liberals. Participants read a fictional news article [...] discussing two different welfare options, a limited coverage plan proposed by Republicans (the Comprehensive Assistance Plan) and a broader coverage plan proposed by Democrats (the Umbrella Aid Plan). After reading the news article, participants were asked to indicate which plan they preferred on a 7-point scale [...]. A single-item measure of political self-identification (1 = very liberal, 7 = very conservative) [...].

Manipulation check. The same four items from Study 3a [...].