What motivates bridge building across pernicious group divides? The effects of regulatory motives, framing, and fit on increasing constructive engagement across political and racial divisions

Introduction

Today, with partisan polarization and enmity in the U.S. reaching historic highs (McCarty, 2019; Kleinfeld, 2021; Pew Research Center, 2023), and Black and White racial sorting of our political parties increasing (Boxell et al., 2020), some of America's top historians and political scientists are warning of the possibility of the U.S. facing another civil war (Shimano, 2021; Meacham, 2022; Walter, 2022). In fact, recent polls show politically motivated hate crimes in the U.S. have spiked to their highest level in over a decade (Buchholz, 2021), with antigovernment hate groups designated as one of the most significant threats to national stability (Southern Poverty Law Center, 2023). Today 23% of Americans agree with the statement, “because things have gotten so far off track, true American patriots may have to resort to violence in order to save our country” (Public Religion Research Institute, 2023). This has led a large swath of Americans, including 56% of Republicans and Republican-leaning Independents, to view the U.S.A., the world's oldest continuous democracy, to be on the verge of collapse (Igielnik, 2023).

One of the core drivers of such forms of runaway polarization is that under these conditions more extreme actors often wield disproportionate influence on political discourse in both traditional and social media, with more moderate citizens tending to disengage politically (More in Common, 2018; Kleinfeld, 2021). For example, one study found that approximately 80% of Twitter (today called X) posts are placed by 10% of Twitter users, and these prolific users were twice more likely to tweet about politics than the remaining 90% less active consumers (Wojcik and Hughes, 2019). The salience of these more extreme voices fuels a perception gap in our politics, where partisans on both sides view the attitudes and actions of their outgroup as significantly more severe than they actually are (More in Common, 2019). Such misperceptions often fuel a hardening of attitudes in the perceiver, contributing to a vicious, self-fulfilling cycle.

The good news is that, today, there are thousands of bridge-building groups in communities across the country (see Bridging Divides Initiative, 2023) and across sectors such as government, journalism, business, education, volunteerism, and more (see Bridge Alliance, 2023), who are working to bring willing members of opposing groups together to meet, dialogue, and at times mobilize together to address shared concerns. Many of these initiatives are informed by contact theory (Allport, 1954), which has found in thousands of studies that facilitated intergroup contact—under the right conditions—can promote better understanding and tolerance (Pettigrew and Tropp, 2013; Tropp and Page-Gould, 2014; Paluck et al., 2019; Esterling et al., 2021; Fishkin et al., 2021). However, in today's political climate, these bridge-building initiatives face significant challenges to sufficient, balanced-partisan involvement due to their lobsided appeal to progressive Democrats vs. conservative Republicans (Lawson, 2021), and in terms of their difficulty engaging significant numbers of disengaged moderates who are exhausted and fed up with politics (More in Common, 2023).

The current set of six studies seeks to apply insights from peacebuilding research and from basic research on motivation and framing effects to address these challenges to broadening citizen engagement in bridge building across political and racial differences. Based on prior research on peacebuilding in Israel/Palestine and on regulatory theories of motivation, these studies sequentially investigated the effects of stronger prevention (mitigating negative outcomes) vs. promotion (seeking ideal outcomes) and assessment (evaluating the best way to proceed) vs. locomotion (just “getting it done”) motivational orientations on people's involvement with political and racial bridge-building activities (BBAs) in the U.S across political, racial, and gender groups. In addition, we examined whether a fit effect between people's orientation and framing (i.e., how the activities were presented) can increase motivation to work across divides. Findings from six studies are presented and the implications for mobilizing bridge building are discussed.

Relevant work

Uncovering what motivates peacebuilding in deeply divided societies

In 2011, a group of researchers set off to investigate the basic underlying motives that drive citizens to engage in peacebuilding efforts in the context of one of the most divided societies on the planet, Israel and Palestine (Coleman et al., 2012). In an attempt to circumvent the often-engrained automatic programming and self-report biases common in such settings of protracted conflict (Bar-Tal et al., 2012), the researchers eschewed traditional survey methods and instead employed rule development experimentation and conjoint analysis, methods developed in consumer market research for identifying resonant, unarticulated, implicit motives for behavior (Moskowitz and Gofman, 2007; Gofman and Moskowitz, 2012).¹

A series of three paired studies were conducted simultaneously in both Israel and Palestine over a 2-year period, which revealed two distinct motives driving engagement with peacebuilding (Coleman et al., 2012).

Peacebuilding motive 1: peace and coexistence are possible and beneficial

The studies found that roughly 56% of the joint-population of Israel/Palestine surveyed was significantly motivated by the implicit belief that peace and coexistence between Israelis and Palestinians are possible, and by the goal of achieving positive outcomes (better education for children, more safety and security, freedom from violence and oppression) from realizing peace. However, this group was also found to be particularly demotivated by the mention of painful losses in their communities (injuries, deaths, and lost revenues or opportunities), particularly in relation to their children. This loss frame had strong negative effects on peace-building motives for this group.

Peacebuilding motive 2: preventing painful losses through responsibility

The same studies also revealed that 44% of the populations sampled were motivated to make peace by the dread of continuing pain and suffering, the great need for safety and security, and by the recognition that leaders on both sides of the conflict share responsibility for violent atrocities committed. This group seemed primarily motivated to work for peace in an attempt to prevent further death, suffering, violence and insecurity. However, again, this group was also significantly demotivated by the mention of a promising future from peace (better future for children, economic opportunity and prosperity, international cooperation). In other words, the positive potential for peace was found to be repulsive to this group.

These findings suggested that two of the most basic motives people harbor to work for peacebuilding in Israel/Palestine are to some degree mirror opposites of one another. In other words, those motives that might mobilize roughly half of the population to work for peace may in fact disincentivize the other half. The results also indicated that the two basic motives identified for promoting peace transcended Palestinian and Israeli ethnic and political affiliations, with 46% of Israelis and 54% of Palestinians being driven by the benefits of peace and 53% of Israelis and 47% of Palestinians being moved by the threat of more hardship. Participants in the peace-benefits group also varied in their political affiliation with 14% indicating they were “rightists,” 23% as “centrist” and 10% as “leftists,” while the preventing-loss participants reported 20% as “rightists,” 22% as “centrists” and 9% as “leftists,” with the remainder not identifying with any political ideology.

Regulatory theories of motivation

The primary distinction identified through our rule development studies in Israel and Palestine are consistent with findings from decades of basic social-cognitive research on motivation. In particular, a similar distinction has been identified in self-regulatory research with the difference between promotion and prevention modes in regulatory focus, or in how people self-regulate (Higgins, 1997, 1998). In the promotion mode of self-regulation, people are especially motivated by experiences that enhance positive outcomes (e.g., attaining money or resources, gaining power, or enhancing self-esteem), while in the prevention mode, people are more responsive to experiences that increase their risks and negative outcomes. This research has identified chronic differences in peoples' propensities to view and approach the world in either promotion or prevention terms (see Higgins et al., 2001). Accordingly, the same decision or situation can be valued to the extent that it promotes one's wellbeing and/or to the extent that it prevents an erosion of one's current level of wellbeing or reduces the intensity of negative outcomes.

In addition to the development of regulatory focus theory on differences in goal orientation in motivation, researchers also developed a complementary theory of regulatory mode, or of differences in how people prefer to achieve their goals (Kruglanski et al., 2000; Higgins, 2012). Individuals have been found to differ in the degree to which they prefer locomotion or assessment modes of action when achieving goals, with those geared toward a locomotion mode being inclined to be more focused on moving quickly and getting things done, while those with a strong preference for assessment tending to take time to analyze different goals and options before acting (Hodis et al., 2017).

Furthermore, additional research investigating regulatory fit theory suggests that when there is a match between a person's more chronic orientation to a goal (prevention or promotion) and how that goal is framed (as preventative or promotive), and when there is a match between the means preferred to approach a goal (assessment or locomotive orientation) and the means presented to achieve the goal (evaluating the best way forward or just acting), it produces a state of regulatory fit that both engenders a feeling of appropriateness about the goal pursuit and increases task engagement (Higgins, 2005, 2006). Higher levels of regulatory fit have been found to intensify responses, such as the valuing of or satisfaction with a chosen object (Kruglanski et al., 2007), while states of non-regulatory fit in comparison have evidenced no such effects.

Taken together, the relative strength of the motivational foci of prevention vs. promotion goals, of the motivational mode preferences of locomotion vs. assessment means of goal attainment, and the added value of fitting vs. non-fitting framing of such goals and processes, offer a promising repertoire of motivational levers to investigate with regard to stimulating bridge building across differences.

The research questions

The current studies attempt to build on previous research to address the following research questions:

• Are higher levels of the self-regulatory motives of prevention goals for mitigating conflict and promotion goals for benefitting from peace associated with increases in reported willingness to engage in bridge-building activities across group divisions?

• Does the fit between more chronic preventive vs. promotive motivational focus orientations with prevention vs. promotion framing conditions for political/racial bridge-building activities increase the degree of reported willingness to actively engage in those activities?

• Does the fit between more chronic assessment vs. locomotion motivational mode orientations with assessment vs. locomotive-framed activities for political/racial bridge-building increase the degree of willingness to engage in those activities?

• Finally, are there meaningful differences in (a) political party affiliation, (b) racial group identification, (c) gender, and (d) racial identity values on the tendency to engage in bridge-building activities and the inclination to be motivated by prevention, promotion, locomotion, or assessment orientations?

We hypothesized that, in general, individuals with higher levels of either prevention or promotion motives for bridge-building would report a greater likelihood of engaging in such activities (Hypothesis 1). However, we also predicted that individuals with stronger prevention motives would be significantly more likely to participate in activities when they were framed as preventing harmful intergroup encounters, compared to individuals with higher levels of promotion motives (Hypothesis 2), while individuals with higher levels of promotion motives would be significantly more likely to engage with such activities when framed as promoting positive peace, compared to individuals with strong prevention motives (Hypothesis 3). We further hypothesized that, in general, individuals with higher levels of assessment mode preferences would be significantly more likely to participate in BBAs when they were framed as assessment activities, compared to individuals with higher levels of locomotion mode preferences (Hypothesis 4), while individuals with higher levels of locomotion preferences would be significantly more likely to engage with BBAs when framed as locomotion activities, compared to individuals with stronger assessment mode preferences (Hypothesis 5).

Overview of six research studies

To test our hypotheses, we conducted a series of six studies, presented in the subsequent sections. Our first two studies (N = 296, 300) investigated the effects of prevention vs. promotion mindsets and framing effects on engagement with political bridge-building activities across three political groups (i.e., Republican, Democrat, Independent) and three racial groups (i.e., White, Black, non-Black People of Color; Study 2). Applying our learnings from the political studies, Studies 3 (N = 300) and Study 4 (N = 429) examined prevention vs. promotion mindsets, framing effects, and gender, racial, and political differences on engagement with interracial bridge-building activities. Study 5 (N = 300) investigated the effects of assessment vs. locomotion mode preferences on engaging with assessment vs. locomotive-framed interracial bridge-building activities. The final study, Study 6 (N = 425), combined learnings from our prior studies to examine the effects of prevention vs. promotion mindsets, locomotion vs. assessment preferences, framing effects, and differences in racial (White, Black, non-Black People of Color), gender, and political groups (Republican, Democrat, Independent) on engagement with interracial bridge-building activities.

Study 1

Using items obtained through our prior research investigating motives for peacebuilding in Israel and Palestine (Coleman et al., 2012), Study 1 was conducted as an initial test of the extent to which the promotion vs. prevention items identified in our prior research and adapted for this study were applicable in the U.S. context. A second aim was to ascertain whether there were significant correlations between both promotion and prevention motivational orientations and engagement in political bridge-building activities (stronger motives = higher willingness to engage with BBAs; Hypothesis 1). Third, we wanted to test whether there were “fit effects” between the different motivational mindsets and consistent frames for political BBAs. We hypothesized that individuals with higher levels of prevention mindset would be significantly more likely to participate in BBAs if the prompt was framed as preventing harmful intergroup encounters, compared to individuals with promotion mindsets (Hypothesis 2), while individuals with higher promotion mindsets would be more likely to join BBAs framed as promoting bipartisan peace, compared to individuals with prevention mindsets (Hypothesis 3). The data was collected from December 2021 to January 2022.²

Method

Procedure

Participants were instructed to complete a three-part survey that measured their motivations to engage in different political BBAs in the U.S. Participants took 10 min on average to respond and were compensated $2.50 upon completion of the survey. The first part, promotion vs. prevention scale, included 10 questions that asked participants to rate the degree to which they felt motivated by each item. The next section measured their willingness to engage with 10 different BBAs along with one attention check question. This section consisted of two experimental conditions, one with a promotion and the other with a prevention framing stem in the opening prompt. Participants were randomized into one of the two conditions. The survey concluded with demographic items.

Sample and design

Participants were recruited via Prolific, an online participant recruitment platform that verified its sample pool using IP addresses and bank-grade identification checks. We recruited a balanced sample into the six groups from the 2 (framing conditions: prevention, promotion) × 3 (political affiliation: Democrat, Independent, Republican) between-participant design. After removing one participant who did not pass the attention check question, a total of 296 respondents (N = 296) were selected into the final analysis (77.7% White/Caucasian, 4.7% Black/African/Caribbean, 5.4% Hispanic/Latino, 4.4% Asian/Asian American, 6.7% other). In terms of gender, 49.3% identified as male, 49.7% as female, 1% non-binary, 3 missing data points.

Measures and conditions

Promotion/prevention scale

We asked participants to rate the extent to which each of 5 promotion and 5 prevention statements (presented randomly) motivated them to engage in political BBAs (3-point scale: 1 = Not at all, 2 = Maybe, 3 = Definitely). As recommended by Moskowitz and Gofman (2007), each of the 10 items were then dichotomized with “Not at all” coded as 0 and “Maybe”/“Definitely” coded as 1. The final promotion and prevention composite scores were created by averaging the dichotomized promotion items and the dichotomized prevention items separately.

Experimental promotion/prevention framing conditions

There were two versions of the survey, which only differed on the stem statement prior to participants' rating of the BBAs. For promotion framing, participants were presented with 10 BB activities preceded by “In order to promote a healthier, more positive political climate in my community I would be willing to engage in the following…” For prevention framing, the stem read “In order to prevent the chances of political violence becoming worse in my community I would be willing to engage in the following…” Participants were randomized to be presented with only one of the two versions of the survey, promotion framing (n = 150) and prevention framing (n = 146).

Political bridge-building activity scale

We adapted a set of 10 political bridge-building items from a set developed by Harris and Young (2009) as well as from descriptions of existing bridge-building initiatives. We then asked participants to rate their willingness to engage in these 10 activities on a 3-point scale (1 = Not at all, 2 = Maybe, 3 = Definitely) with either a promotive or preventive framing (see above for framing conditions). Exploratory factor analysis (EFA) results categorized these activities into two BBA categories: Dialogue & Education Activities (7 items; α = 0.86; ω = 0.90) and Social Activities (2 items; α = 0.65). The means for these BBA categories and the mean for all ten items were chosen as three dependent variables for subsequent analyses.

Results and discussion

Promotion/prevention scale

In order to examine whether the promotion/prevention scale had structural validity in the U.S. context, we ran EFA on the ten statements (see Table 1 for factor loadings). Our results showed KMO at 0.86, indicating a good level of sampling adequacy for factor analysis. Selecting maximum likelihood with Promax rotation and suppressing loadings smaller than 0.4 in R Studio version 2023.06.01+524, we found that all of the promotion items loaded onto one factor with good reliability (Cronbach's α = 0.80; McDonald's ω = 0.84). Similarly, the prevention items also loaded into a separate factor (α = 0.76; ω = 0.82).

Table 1

Table 1. Factor loadings for exploratory factor analysis.

Our preliminary analyses indicated no significant mean differences in a willingness to engage in BBAs or in individual promotion vs. prevention orientations across the experimental conditions, suggesting a baseline balance between these groups. One-way ANOVA with Bonferroni correction showed no significant differences across political groups in promotion vs. prevention orientations or socio-economic status. This implication that people can have a promotion or a prevention mindset regardless of their social groups was consistent with other research on regulatory focus theory (Higgins, 1997, 1998).

There was also a strong positive correlation between individual promotion and prevention orientations (r = 0.466, p < 0.001). However, the factor analysis categorized the items into two different factors, confirming that promotion and prevention mindsets were not opposite but orthogonal constructs. This was further supported through the normal distribution of the difference between individuals' promotion and prevention scores—the majority of the population had both with a small subset leaning strongly toward promotion or prevention (Figure 1).

Figure 1

Figure 1. Distribution of promotion-prevention difference.

Hypothesis testing

As expected, individuals with stronger promotion and/or prevention motives were found to be more likely to report a willingness to engage in political BBAs (Hypothesis 1). We further hypothesized that individuals with high prevention motives would be significantly more likely to participate in BB activities if the prompt was framed as preventing harmful conflict, compared to individuals with high promotion motives (Hypothesis 2), while individuals with strong promotion motives would be significantly more likely to join BBAs framed as promotive, compared to individuals with high prevention motives (Hypothesis 3). To test for these “fit” effects, we filtered respondents with equal levels of promotion and prevention mindsets. We decided to focus our comparisons on people leaning toward one motive over the other as those possessing higher levels of both motives would “fit” in both framing conditions, potentially confounding the results. In other words, after calculating absolute difference scores between promotion and prevention mindsets, we selected only cases with absolute difference scores above the mean (>0.59). The results supported our fit hypotheses. In the promotion framing condition, higher promotion orientation scores positively correlated with higher levels of engagement with bridge building activities (r = 0.31, p < 0.05 for Total Activities, r = 0.35, p < 0.05 for Dialogue and Education Activities). However, there were no significant correlations for prevention orientation and engagement in the promotion-frame condition (Hypothesis 2). Similarly, in the prevention framing condition, higher prevention scores positively correlated with higher engagement with Dialogue and Education BBAs (r = 0.31, p < 0.05). However, higher promotion scores did not evidence significant correlations in the prevention framing condition (Hypothesis 3).

Demographic differences

A one-way ANOVA also revealed significant differences in willingness to engage in political BBAs across political groups. Those who identified as Democrats (M = 22.48) were significantly more likely to engage in overall BBAs than people who identified as Republicans (M = 20.56); F_(2;286) = 4.24; p = 0.015. Similarly, Democrats (M = 15.72) were significantly more likely to engage in Dialogue & Education BBAs in particular more than Republicans (M = 14.18); F_(2;286) = 4.71; p = 0.01.

To summarize, Study 1 served to validate the prevention and promotion scales derived from our prior research, with the items for each scale being categorized by the EFA in distinct dimensions, showing sufficient structural validity, and being significantly correlated with reported intentions to engage in bridge-building. Further, both fit hypotheses (Hypotheses 1, 2) were supported, suggesting that when participants evidenced strong and distinct prevention or promotion motives and were presented with correspondingly framed activities, they expressed a greater willingness to engage.

Study 2

Applying learning from Study 1, we made several minor revisions to the methods in Study 2 and sought to test the replicability of the results. The revisions included changing the number of choices for the promotion/prevention scale, adding four BBA items, asking participants identified as Independents about their political leanings, and running two-way ANOVAs to further examine our hypotheses. This follow up study was conducted from February to March of 2022.

Method

Procedure

Study 2 had an identical procedure to Study 1.

Sample and design

After removing two participants who failed the attention check, a total of 300 participants (N = 300) were included in subsequent analyses (76.7% White/Caucasian, 3.7% Black/African/Caribbean, 3.7% Hispanic/Latino, 8.3% Asian/Asian American, 6.9% other). In terms of gender, 34.3% identified as male, 62.3% as female, 1.7% non-binary, 1% transgender. With recruitment from Prolific, each of the six groups from the 2 (framing condition: prevention, promotion) × 3 (political affiliation: Democrat, Independent, Republican) had 50 participants. Among participants who identified as Independents (n = 100), 50.5% identified as Democratic leaning, 23.3% identified as Republican leaning, and 25.3% identified as neither.

Measures and conditions

Promotion/prevention scale

All of the 10 items from Study 1 were kept but were presented in the online survey on individually in randomized order to control for order effects. The initial 3-point scale items were also modified to be rated on a 4-point scale (1 = Not at all; 2 = Very little; 3 = Somewhat; 4 = Very much) to control for respondents' susceptibility to desirability bias (Johns, 2005), with “Not at all”/“Very little” coded as 0 and “Somewhat”/“Very much” coded as 1. The final promotion and prevention composite scores were created by averaging the dichotomized promotion items (α = 0.80; ω = 0.83) and the dichotomized prevention items (α = 0.77; ω = 0.81) separately. Here, Democrats were found to be more likely to have promotion orientations than Republicans (r = −0.152; p = 0.008).

Experimental conditions

Similar to Study 1, participants were randomized into either the promotion (n = 150) or prevention framing (n = 150) condition.

Political bridge-building activity scale

We added four new social activities inspired by existing bridge-building organizations to the initial ten items. Participants were asked to rate their willingness to engage in these 14 activities on a 5-point scale (1 = Not at all to 5 = Extremely willing) with either a promotive or preventive stem framing. EFA results categorized these activities into two categories: Dialogue & Education (8 items; α = 0.88; ω = 0.91) and Social Activities (5 items; α = 0.79; ω = 0.82). The two mean scores for these categories and the mean for all 14 items were taken as three dependent variables for subsequent analyses.

Results and discussion

Even with the adjustments in the promotion/prevention scale from 3 to 4-points and presenting the items in randomized order, EFA results categorized the ten items into promotion vs. prevention dimensions as expected with good reliability (α = 0.80; ω = 0.83 for promotion and α = 0.77; ω = 0.81 for prevention; KMO = 0.90 indicating factorization adequate for the data). EFA also indicated that Study 2 evidenced the same 3 highest loading items from each scale as Study 1. This finding implied that there is potential for a shorter 6-item scale to measure promotion/prevention mindsets in the U.S. context.

The two motivational mindsets (prevention and promotion) were normally distributed and distributed equally across all 3 political groups. Similar to Study 1, individuals with higher promotion and/or prevention mindsets were found to be more likely to engage in political BBAs (Hypothesis 1). The change in our scales also resulted in stronger effects in Study 2. For example, for Total BBAs, higher promotive scores alone correlated positively with higher engagement with promotive-framed activities (r = 0.55, p < 0.001) and preventive-framed activities (r = 0.48, p < 0.001). Similarly, higher preventive scores alone correlated positively with higher engagement with promotive-framed activities (r = 0.44, p < 0.001) and preventive-framed activities (r = 0.52, p < 0.001; see Table 2 for all correlations).

Table 2

Table 2. Correlations for participants with strong promotion or prevention motives.

Two-way ANOVAs were then run to further examine Hypotheses 2 and 3 (see Supplementary material A for post-hoc power calculations using G^*Power version 3.1; Faul et al., 2009). First, both promotion and prevention motives scores were transformed into factors by dichotomizing them at the mean. Four different two-way interaction terms were conducted: 1. Promotion × prevention motives in the promotion framing condition, 2. Promotion × prevention motives in the prevention framing condition, 3. Promotion motive × framing conditions, and 4. Prevention motive × framing conditions. Prior to conducting these analyses, a series of assumption tests were assessed for each interaction term. All of the terms satisfied Levene's test for homogeneity of variances. However, most of the ANOVA models violated the assumptions of normality of residuals (Shapiro-Wilk test and qqPlot). With the sample size and F-test's robustness against normality assumptions (Schmidt and Finan, 2018), we decided to proceed with the analyses.

Two-way ANOVA reported significant effects for promotion and prevention motives in the promotion framing condition. From the ANOVA results (Table 3; see Figure 2 for interaction plots), we can see that promotion motive was the only significant factor in predicting willingness to engage in Dialogue & Education Activities [F_{(1, 60)} = 6.90; p = 0.01] and Total BBAs [F_{(1, 60)} = 6.45; p = 0.01], partly supporting Hypothesis 2 (promotion-promotion match). For Social Activities, only prevention motive had a moderately significant main effect [F_{(1, 60)} = 3.89; p = 0.05].

Table 3

Table 3. Two-way ANOVA with promotion x prevention motives in promotion framing condition.

Figure 2

Figure 2. Study 2 interaction plots for promotion × prevention motive in promotion framing condition.

In the prevention framing condition, Hypothesis 3 was strongly supported using two-way ANOVA with promotion and prevention motives. All three models for Dialogue & Education Activities [F_{(1, 56)} = 15.33; p < 0.001], Social Activities [F_{(1, 56)} = 5.19; p = 0.03], and Total BBAs [F_{(1, 56)} = 17.63; p < 0.001], showed that only prevention motive was a significant predictor for willingness to engage (see Table 4 for all statistics).

Table 4

Table 4. Two-way ANOVA with promotion x prevention motives in prevention framing condition.

To examine framing effects and promotion motives, the third round of two-way ANOVAs was conducted using promotion motive and framing manipulation as the predictors. Even though the ANOVA models with interactions did not produce significant results (p < 0.05), they indicated that there might be significant group differences (see Table 5 for all statistics). Post-hoc analyses using simple contrasts revealed patterns that partly supported Hypothesis 2. Specifically, people with high promotion motive were more willing to engage in Dialogue & Education Activities [t₍₁₂₀₎ = −2.45; p = 0.02] and Total BBAs [t₍₁₂₀₎ = −2.45; p = 0.02] under the promotion framing (Condition 1; match) than people with low promotion motive. There was no significant differences in willingness to engage in BBAs under the prevention framing condition (mismatch).

Table 5

Table 5. Two-way ANOVA with promotion motive x framing condition.

The fourth and final interaction examined in this study was between prevention motive and framing manipulations. All ANOVA models revealed significant main effects for prevention motive (see Table 6 for all statistics). Results from post-hoc analyses were congruent with Hypothesis 3. In the prevention framing condition (match), people with higher prevention motive were more willing to engage in Dialogue & Education Activities [t₍₁₂₀₎ = −3.33; p = 0.001], Social Activities [t₍₁₂₀₎ = −2.20; p = 0.03], and Total BBAs [t₍₁₂₀₎ = −3.60; p < 0.001] than people with lower prevention motive. Similar to the promotion motive, there was no significant effect when the motive and framing were mismatched (prevention motive and promotion framing). Lastly, people with high prevention motive were significantly more willing to engage in Dialogue & Education Activities [t₍₁₂₀₎ = −1.94; p = 0.05] and Total BBAs [t₍₁₂₀₎ = −1.95; p = 0.05] when the framing was prevention (match) than when the framing was promotion (mismatch; see Figure 3 for interaction plots). The framing did not have a significant effect for people with low prevention motives for both of these cases.

Table 6

Table 6. Two-way ANOVA with prevention motive x framing condition.

Figure 3

Figure 3. Study 2 interaction plots for prevention motive × framing manipulation.

Demographic differences

Consistent with findings from Study 1, ANOVA results with Bonferroni indicated participants who identified as Democrats were significantly more likely to engage in Dialogue & Education activities (M = 29.2) than people who identified as Republicans (M = 26.28); F_(2;289) = 3.98; p = 0.02. In addition, independent t-test results also indicated differences in gender effects for BBA engagement. Because of the sample imbalance, weighted analysis was employed to control for 34.3% male (weighted at 1.458) vs. 62.3% female (weighted at 0.803). The results showed that women were significantly more likely to engage in BBAs than men (M = 3.33; SD = 0.88), t₍₂₈₈₎ = −3.76; p < 0.001.

In summary, the findings from our first two studies supported our hypothesis that a fit between promotion vs. prevention motives and consistent framing motivated people to engage more in political BBAs.

Study 3

With our hypotheses supported in political bridge-building from the first two studies, Study 3 sought to test our approach in another challenging context of societal divisiveness, racial bridge-building in the U.S. Here we focused on studying members of three racial groups: White, Black, and non-Black People of Color (non-Black POC).³

Method

Procedure

The procedures used in Study 3 were the same as those employed in Studies 1 and 2, with the exception of the types of BBAs presented and the inclusion of racial panels in recruiting of participants.

Sample and design

Study 3 included 300 participants (N = 300; 4 removed due to failed attention check). Focusing on racial bridge-building, we were able to achieve an equal distribution for 6 groups from the 2 (framing condition: prevention, promotion) × 3 (racial group: White, Black, non-Black POC) with 50 participants in each. Among our participants, 52.3% were Democrat, 34% Independent, and 7% Republican. In terms of gender, our respondents were 48.3% male, 49% female, 3% non-binary, and 0.3% transgender.

Measures

Promotion/prevention scale

We adapted all 10 items from the first two studies to focus on motives for engaging in BBAs regarding race relations in the U.S. Similar to Study 2, the items were presented in randomized order and were rated on a 4-point scale (1 = Not at all; 2 = Very little; 3 = Somewhat; 4 = Very much), with “Not at all”/“Very little” coded as 0 and “Somewhat”/“Very much” coded as 1. The final promotion and prevention composite scores were created by averaging the dichotomized promotion items and of the dichotomized prevention items separately.

Experimental conditions

Participants were randomized into either the promotion (n = 150) or prevention framing (n = 150) version of the survey. The framing stems read: “In order to promote a healthier, more positive racial climate in my community I would be willing to engage in the following...” (promotion framing) or “In order to prevent the chances of race-related violence becoming worse in my community I would be willing to engage in the following...” (prevention framing).

Racial bridge-building activity scale

The 14 political BB activities were converted into racial BBAs for this study. Participants were asked to rate their willingness to engage in these 14 activities on a 5-point scale (1 = Not at all to 5 = Extremely willing) with either a promotive or preventive stem framing. EFA results categorized these activities into two dimensions. However, at this point we interpreted the dimensions differently, realizing that they were best essentialized as those focused on interrogating racial differences (Political Activities; 6 items; α = 0.88; ω = 0.92) vs. those oriented toward gathering people across groups for community building activities (Community Activities; 4 items; α = 0.77; ω = 0.81). The two mean scores for these categories and the mean for all 14 items were taken as three dependent variables for subsequent analyses.

Results and discussion

Despite the change from politics to race, Study 3's results generally mirrored many findings from the first two studies. First, EFA results indicated that the promotion vs. prevention scale could be translated to the racial bridge-building context with good reliability (α = 0.84; ω = 0.85 for promotion and α = 0.70; ω = 0.77 for prevention). However, distinct from the findings from Studies 1 and 2 where individuals with higher promotion and/or prevention mindsets were found to be more likely to engage in political BBAs (Hypothesis 1), Study 3 found that only stronger promotion motives significantly predicted higher likelihood to engage in racial BBAs.

Second, our fit hypothesis was also partially supported. Participants with higher promotion orientations reported being more motivated to engage in political (r = 0.34; p < 0.05) and community (r = 0.39; p < 0.01) racial BBAs when presented with promotive framing compared to preventive framing (r = 0.28, p < 0.05 for political; no significant results for community activities; Hypothesis 3). However, there were no significant fit findings for prevention motives (Hypothesis 2). Finally, independent t-tests for gender revealed that women (M = 4.29; SD = 0.74) were significantly more likely to engage in community activities than men (M = 4.07; SD = 0.83), t₍₂₉₀₎ = −2.73; p < 0.05. ANOVA reported no significant differences among racial groups in willingness to engage in racial BBAs.

The findings from Study 3 on racial bridge building mostly replicated those of the first two studies on the effects of prevention and promotion orientations, with one major difference. Promotion motives, especially when matched with promotion framing of the BBAs, were the only conditions that evidenced significantly higher levels of engagement with racial BBAs. This suggests that there is something inherent to the more idealized promotive framing about the benefits of enhancing race relations that is—at least in the current context—especially motivating for racial BBAs.

Study 4

Study 4 sought to replicate and expand our findings from Study 3 in several ways. First, we added equal panels for three political groups in addition to racial groups, forming a 2 (experimental conditions) × 3 (racial groups) × 3 (political groups) design. Second, we selected the top three highest loading items from the prevention and promotion subscales to shorten the measure from 10 to 6 items. Third, 10 social media-related activities were added to broaden the range of BBAs. Finally, we added a Cross-Ethnic Identity Scale-Adult (CERIS-A; Worrell et al., 2019) to assess different values underlying attitudes regarding engagement across racial differences.

Method

Procedure

Participants were recruited through Prolific and were instructed to complete a four-part survey that measured their motivations to engage in different racial BBAs in the U.S. Participants took around 15 min on average and were compensated $3.50 upon finishing the survey. The first part, promotion/prevention scale, included 6 items that asked participants to rate whether they felt motivated by each in a series of six statements. The next section measured their willingness to engage with 24 different racial BBAs along with one attention check question. Participants were randomized into either promotion or prevention framing conditions for the survey. The third section presented the CERIS-A scale. Finally, the survey concluded with demographic items.

Sample and design

Except for Black Republicans due to recruitment challenges (n = 29), all other demographic combination groups (e.g., White Democrats, White Republicans) had 50 participants, equally divided into promotion or prevention experimental conditions (2 × 3 × 3 = 18 groups). Out of the 429 participants in this study (N = 429), 48.8% identified as male, 49.3% female, 1.2% non-binary, and 0.5% transgender.

Measures and conditions

Promotion/prevention scale

The three highest loading items from each subscale used in Study 3 were selected, presented in randomized order, and rated on a 4-point scale (1 = Not at all; 2 = Very little; 3 = Somewhat; 4 = Very much), with “Not at all”/“Very little” coded as 0 and “Somewhat”/“Very much” coded as 1. The final promotion and prevention composite scores were created by averaging the dichotomized promotion items and of the dichotomized prevention items separately.

Experimental conditions

Similar to Studies 1–3, participants were randomized into either the promotion (n = 217) or prevention framing (n = 210) version of the survey.

Racial bridge-building activity scale

Participants were asked to rate their willingness to engage in the 24 racial BBAs on a 5-point scale (1 = Not at all to 5 = Extremely willing) with either a promotive or preventive stem framing. EFA results categorized these activities into three categories: Political Activities (10 items; α = 0.93; ω = 0.94), Community Activities (8 items; α = 0.88; ω = 0.91), and Social Media Activities (4 items; α = 0.87; ω = 0.90). The mean scores for each category were taken as three dependent variables for subsequent analyses.

Cross-ethnic racial identity scale-adult

The Cross-Ethnic Racial Identity Scale-Adult (Worrell et al., 2019) sought to measure cross-ethnic racial identity attitudes in the U.S. It consisted of seven subscales: Assimilation (i.e., “If I had to put a label on my identity, it would be ‘American,' and not a specific ethnic/racial group;” α = 0.93), Miseducation (i.e., “I think many of the stereotypes about my ethnic/racial group are true;” α = 0.79), Self-hatred (i.e., “Privately, I sometimes have negative feelings about being a member of my ethnic/racial group;” α = 0.88), Anti-dominant (i.e., “I hate people from the dominant racial/ethnic group;” α = 0.85), Ethnocentricity (i.e., “I believe that only people who accept a perspective from their ethnic/racial group can truly solve the race problem in America;” α = 0.76), Multiculturalist Inclusive (i.e., “I believe it is important to have a multicultural perspective which is inclusive of everyone;” α = 0.87), and Ethnic-Racial Salience (i.e., “During a typical week in my life, I think about ethnic and cultural issues many, many times;” α = 0.79). Each of the items were rated on a 7-point scale. The mean scores were taken from the four items in each subscale for subsequent analyses.

Results and discussion

Confirmatory Factor Analysis (CFA) was conducted using the lavaan package in R to test the two-factor model for the promotion/prevention scale. The CFA results revealed that the two-factor model provided a good fit to the data, with CFI = 0.99 and TLI = 0.98. The p-value for RMSEA ≤ 0.05 is not significant, indicating a close-fitting model. All factor loadings were statistically significant (p < 0.05) and ranged from 0.49 to 0.78, indicating moderate relationships between the observed items and their respective latent constructs of promotion or prevention motivational orientations.

Similar to Study 2, we ran two-way ANOVAs in Study 4 to further examine Hypotheses 2 and 3 for bridge-building across racial divides (see Supplementary material A for post-hoc power calculations. Both promotion and prevention scores were transformed into binary factors by cutting at the mean. Four rounds of two-way ANOVAs were conducted: (1) Promotion × prevention motives in promotion framing condition, (2) Promotion × prevention motives in prevention framing condition, (3) Promotion motive × framing conditions, and (4) Prevention motive × framing conditions. All of the models passed Levene's test of homogeneity of assumptions but did not satisfy normality of residuals. However, we decided to proceed with the analyses due to F-test's robustness against the latter assumption (Schmidt and Finan, 2018).

In the promotion framing condition, promotion, prevention motives, and interaction between the two variables were statistically significant. However, promotion motive was the most significant factor in predicting willingness to engage in Political Activities [F_{(1, 213)} = 25.71; p < 0.001], Community Activities [F_{(1, 213)} = 41.85; p < 0.001], and Social Media Activities [F_{(1, 213)} = 23.37; p < 0.001; Table 7], supporting Hypothesis 2. However, findings in the prevention framing condition did not support Hypothesis 3. Despite the mismatch, promotion motive was again the most significant factor in predicting willingness to engage in Political Activities [F_{(1, 207)} = 28.36; p < 0.001], Community Activities [F_{(1, 207)} = 33.58; p < 0.001], and Social Media Activities [F_{(1, 213)} = 29.60; p < 0.001; Table 8]. Aligned with Study 3 findings, this might indicate that promotion motive generally was found to be more predictive of people's willingness to engage in racial BBAs, regardless of framing.

Table 7

Table 7. Two-way ANOVA with promotion x prevention motives in promotion framing condition.

Table 8

Table 8. Two-way ANOVA with promotion x prevention motives in prevention framing condition.

Although findings from running two-way ANOVAs with promotion motive and framing condition did not provide further insights to Hypothesis 2, they were in line with our predictions for Hypothesis 1. Post-hoc analysis revealed people with higher promotion scores were significantly more willing to engage in Political, Social, and Social Media Activities than people with lower promotion scores regardless of framing conditions (see Figure 4 for interaction plots). Similarly, running two-way ANOVAs with prevention motive and framing condition indicated that people with higher prevention scores were more likely to engage in racial BBAs than those with lower promotion motive regardless of framing conditions.

Figure 4

Figure 4. Study 4 interaction plots for promotion motive × framing manipulation.

Out of all the CERIS-A subscales, Multiculturalist Inclusive values had the strongest effect. In general, for all participants, the higher their multiculturally inclusive scores (welcoming diverse perspectives) were, the more they were willing to engage in racial BBAs (r = 0.62, p < 0.001).

Finally, we utilized independent t-test and one-way ANOVA and found significant differences between demographic groups. Similar to prior findings, women (M = 3.86; SD = 0.78) were significantly more likely to engage in overall racial bridge-building activities than men (M = 3.50; SD = 0.99), t₍₄₀₈₎ = −4.105; p < 0.001. For race, one-way ANOVA results with Bonferroni post-hoc revealed that Black participants were significantly more likely to engage in overall bridge-building activities than White or non-Black POC, F_{(2, 413)} = 4.00, p = 0.019. For political affiliation, similar to Study 2, Republicans scored significantly lower in the promotion mindtype than Democrats, F_{(2, 409)} = 5.45, p = 0.005. Democrats (M = 4.00; SD = 0.65) scored highest in willingness to engage in BBAs, followed by Independents (M = 3.65; SD = 0.91), and finally Republicans (M = 3.31; SD = 0.98), F_{(2, 09)} = 21.63, p < 0.001.

To summarize, our findings from Studies 3 and 4 again offered support for the fit hypothesis, although only for promotion motive-frame matches—which seemed to particularly resonate with racial divisions today. Results also indicated that many of the racial bridge-building activities employed in this study were best categorized into political-difference reckoning vs. community-uniting activities. While political activities focused more on examining contrasting experiences from members of various racial groups, community activities aimed to gather people across groups and focus on common interests. Community activities were found to be the most likely to elicit engagement across both promotion and prevention framing. There were also significant demographic differences in the findings, with White Democrats more likely to have promotion mindsets compared to Black and non-Black POC Democrats, while Black Republicans, compared to White and non-Black POC Republicans, tending to have more promotion mindset and were more willing to engage in BBAs. Lastly, Democrats were more likely to engage in racial BB activities in all categories.

Study 5

Findings from Studies 3 and 4, which revealed two main categories of racial BB activities—political or community—highlighted the importance of different types of racial BBAs in motivating people to work across the divides. Building on this, Study 5 investigated two distinct regulatory modes of activity, locomotion and assessment (Kruglanski et al., 2000), as additional predictive dimensions. Using the Locomotion and Assessment Scales (Kruglanski et al., 2000), we sought to test for additional “fit effects” between these dimensions of self-regulation and different types of BBAs. While people with high assessment orientations tend to be “thinkers” that prefer to critically evaluate and deliberate before committing to a particular course of action, people with high locomotion forms of self-regulation tend to be “doers” who act with less deliberation or delay. Applying this to Study 5, we adapted and sorted our BBAs into more assessment-based (i.e., exchanging ideas, thinking, planning) vs. more locomotion-based (i.e., taking actions to achieve goals, physical activities). We hypothesized that individuals with higher levels of assessment mode preferences would be significantly more likely to participate in activities when they were framed as assessment activities, compared to individuals with higher levels of locomotion preferences (Hypothesis 4), while individuals with higher levels of locomotion preferences would be significantly more likely to engage with activities when framed as locomotion activities, compared to individuals with assessment preferences (Hypothesis 5). In order to test for these effects, we replaced the promotion and prevention scales with the locomotion and assessment scales in a correlational study.

Method

Procedures

Participants were asked to complete a 3-part survey with questions about engaging in different racial BBAs in the U.S. It took participants approximately 10 min to complete, and they were compensated $2.5 upon completion. The survey began with questions that measured their assessment and locomotion mode preferences, follow by a measure of their willingness to engage with different BBAs. The survey ended with a set of demographic questions.

Sample and design

A total of 300 participants (N = 300) were selected for this study, divided equally into three political groups (100 Democrats, 100 Independents, and 100 Republicans). The sample consisted of 52.2% male, 44.2% female, 2.7% non-binary, and 0.3% transgender individuals (one missing date point). For race, 52.8% identified as White/Caucasian, 13.3% as Black/African American, 15.3% as Hispanic/Latino, 15.3% as Asian/Asian American, 0.7% as Middle Eastern/North African (MENA), 0.3% as Native/Indigenous people, and 0.7% as belonging to other racial/ethnic backgrounds.

Measures

Locomotion/assessment scales

Participants completed the 24-item locomotion assessment scale (Kruglanski et al., 2000) designed to assess people's self-regulatory mode. The scales were rated on a 6-point Likert scale ranging from 1 (strongly disagree) to 6 (strongly agree). Each scale included 12 items: 12 for assessment (i.e., “I often critique work done by myself or others”; α = 0.87; ω = 0.89) and 12 for locomotion items (i.e., “I am a go-getter”; α = 0.85; ω = 0.85).

Locomotion/assessment activities

After adapting and revising our race-relation BBAs according to locomotion vs. assessment, we ran a preliminary Q-sort study to help categorize and validate our activity items into locomotion or assessment types. Seven graduate students who were briefed on the locomotion vs. assessment distinction were asked to sort a set of 24 racial BBAs into either assessment or locomotive activities. The results were then analyzed to assess whether our activities were framed accurately—all activities that had <75% agreement were modified. EFA results identified two factors: Assessment Activities (7 items; α = 0.85; ω = 0.92) and Locomotion Activities (5 items; α = 0.84; ω = 0.87). All of the items from each factor were aligned with the categories as we initially intended.

Results and discussion

In general, higher locomotion mindsets were found to be more positively associated with a general willingness to engage in BBAs, for both assessment and locomotion activities. This suggested that individuals who preferred to “get things done” were naturally inclined to actively address interracial challenges. However, we also found support for our fit hypothesis. Specifically, higher assessment scores significantly correlated with higher engagement with assessment-based activities (Hypothesis 4; r = 0.160, p = 0.005) but not for locomotion activities (mismatch). In addition, higher locomotion scores predicted higher engagement with locomotion-based activities (Hypothesis 5; r = 0.285, p < 0.001), compared to assessment-based activities (mismatch; r = 0.152, p = 0.008).

Furthermore, an independent t-test revealed significant demographic differences. Overall, Black, Indigenous, and People of Color (BIPOC) respondents exhibited higher inclination toward participation in assessment-based racial BB activities (M = 3.20; SD = 1.11) compared to their White counterparts (M = 3.00; SD = 0.88), t₍₂₉₆₎ = 15.79, p = 0.04. In terms of gender, women (M = 3.89; SD = 0.86) displayed a significant willingness to engage in locomotion-oriented activities compared to men (M = 3.43; SD = 1.03), t₍₂₈₈₎ = 5.96, p < 0.001. One-way ANOVA results indicated that Republicans (M = 4.44) had higher locomotion mindsets than Democrats (M = 4.10) and Independents (M = 4.13); F_(2;291) = 6.56; p = 0.002. Conversely, Democrats (M = 3.32) were more willing to engage in assessment-based racial BB activities than Republican participants (M = 2.77); F_(2;291) = 7.76; p < 0.001.

To summarize, our first investigation into the relationship between self-regulatory modes of Assessment and Locomotion and racial bridge-building activities that differed in terms of their assessment vs. locomotion natures, again evidenced the predicted fit effects. However, high locomotion individuals also showed a higher general willingness to engage in all types of racial BBAs. This may reflect the basic difference between preferring to simply move on with the future of race relations vs. needing to reckon with past grievances.

Study 6

In Study 6, our final study in this current series, we integrated insights from this sequence of studies to examine bridge-building motives regarding race-relations in the U.S. Employing a 2 × 3 × 3 experimental design (N = 425), we aimed for balanced representation across two conditions (promotion vs. prevention frame), three political groups (Republican/Democrat/Independent), and three racial groups (White/Black/non-Black POC), totaling 18 groups (see Figure 5 for a visualization of the study design). However, similar to Study 4, we ran into significant challenges in recruiting Black Republicans for this study, a limitation to our findings. To reduce cognitive demands on participants taking the survey, we employed the brief 6-item promotion/prevention scale, and similarly selected the top-loading three items each for locomotion and assessment from the full 24-item scale, standardizing them into four 3-item scales for each motivational orientation. Additionally, based on our findings from Study 4, Multiculturalist Inclusive and Self-hatred subscales from the CERIS-A were included as control variables.

Figure 5

Figure 5. Study 6 independent and dependent variables by condition.

Methods

Procedure

Study 6 had similar procedures to Study 4.

Sample and design

The sample of 425 participants was comprised of 189 females (40.0%), 220 males (51.3%), 11 non-binary (2.6%), and 1 transgender respondent. Due to intentional sampling from the study design, 38.2% identified as White, followed by Black or African American (31.7%), Asian or Native Hawaiian or Pacific Islanders (10.0%) and Hispanic or Latino (6.3%). Regarding political affiliations, 36.1% identified as Democrat, 33.3% as Independents, and 29.1% as Republicans.

Measures and conditions

Promotion/prevention scales

Participants were asked to respond to the validated 6-item promotion and prevention scales from Study 4 on how much each statement increased the likelihood that they would engage in bridge-building activities with others across racial differences. To standardize the measures for participants, we increased the response options from 4-point to 6-point Likert scale (1 = Not at all to 6 = Extremely) to be consistent with the locomotion/assessment scale.

Assessment/locomotion scales

We selected the 3 highest loading items from each of the Assessment/Locomotion scales from Study 5 to measure assessment and locomotion orientations for this study. Participants were asked “How much do you agree with the following statement?” and rated each statement on a 6-point Likert scale (1 = Strongly disagree to 6 = Strongly agree).

Racial bridge-building activity scale

A total of 24 racial BBAs were included to measure participants' willingness to engage. Similar to prior studies, they were measured on a 5-point scale (1 = Not willing, 3 = Undecided, 5 = Very willing). The 24 items were modified to sort into 4 categories (2 × 2; assessment/locomotion x political/community type). We conducted a preliminary Q-Sort survey with graduate students (N = 12) to assure the 24 activities were adequately categorized, with 6 items in each category (see Supplementary material B for all activities in each category).

Cross-ethnic racial identity scale-adult

In order to shorten our survey for participants, we selected the most relevant subscale from the CERIS-A from Study 4, Self-hatred (i.e., disliking the racial/ethnic group one belongs to) and Multiculturalist Inclusive (i.e., welcoming diverse perspectives). The subscales each contained 4 items, rated on a 7-point scale (1 = Strongly disagree to 7 = Strongly agree).

Experimental conditions

Respondents were randomized into either the promotion (n = 225) or prevention framing (n = 204) version of the survey.

Results and discussion

Factors

We ran CFA to test the four-factor model for promotion (α = 0.85; ω = 0.85), prevention (α = 0.65; ω = 0.71), locomotion (α = 0.73; ω = 0.74), and assessment mindsets (α = 0.80; ω = 0.82). The Comparative Fit Index (CFI) was found to be 0.99 and the Tucker-Lewis Index (TLI) yielded a value of 0.99, indicating a good fit to the data. RMSEA was smaller than 0.5, further supporting the model's adequacy. Table 9 presents the factor loading for each item in the model. All factor loadings were significant (p < 0.05) and ranged from 0.42 to 0.87, demonstrating moderate associations between the latent constructs and their corresponding indicators.

Table 9

Table 9. Factor loadings for confirmatory factor analysis results.

With KMO = 0.96 suggesting factorization adequate for the data, EFA was run to categorize the bridge-building activities, which resulted in three types of activities: Assessment Activities (11 items; α = 0.95; ω = 0.96), Locomotion-Political (5 items; α = 0.89; ω = 0.91), and Locomotion Community Activities (6 items; α = 0.90; ω = 0.93). It was striking that all assessment activities—even those Q-sorted into Assessment-Political vs. Assessment-Community activities, loaded onto one dimension. This may reflect the focus of such actions, which request people to reflect critically on and evaluate racial differences in communities, a finding we will return to in the discussion.

Fit hypotheses

Hierarchical Regression was conducted to examine the predicted relationships between Promotion/Prevention motives, framing, and willingness to engage in racial BBAs (i.e., mean score of all 24 racial BBAs). In order to test these relationships, we separated the responses in the promotion framing condition (n = 225) from the prevention framing (n = 204). For each set of data, Racial groups, Political groups, Multiculturalist Inclusive and Self-hatred attitudes, and Locomotion and Assessment motives were entered as control variables in Step 1. In Step 2, Promotion and Prevention motives were added as predictor variables.

For the promotion framing condition in Step 1, the control variables explained a significant proportion of variance in willingness to engage in BBAs, adjusted R² = 0.50, F_{(8, 208)} = 27.55, p < 0.001. Race (β = 0.36, p < 0.001 for Black respondents), Multiculturalist Inclusive (β = 0.42, p < 0.001), Self-Hatred (β = 0.08, p = 0.01), and Locomotion (β = 0.15, p = 0.001) emerged as additional significant positive predictors, while Political Group differences and Assessment orientation were not found to be significant predictors.

Upon adding Promotion and Prevention in Step 2, the model accounted for an additional significant proportion of variance, F_{(2, 206)} = 14.15, p < 0.001. Both Prevention (β = 0.13, p = 0.003) and Promotion motives (β = 0.14, p < 0.001) exhibited positive and significant relationships with engagement in BBAs. This means that higher promotion and prevention scores were both positively correlated with higher willingness to engage in racial BBAs—but only with a promotion framing—regardless of racial, political groups, self-hatred, and multicultural inclusiveness score. In addition, the relationship was stronger when there was a promotion-promotion match compared to a promotion/prevention mismatch.

Applying the same steps to the prevention framing data, in Step 1, the control variables explained a significant proportion of variance in willingness to engage in BBAs, adjusted R² = 0.47, F_{(8, 187)} = 22.88, p < 0.001. Slightly different from the promotion framing, Political Group (β = −0.05, p = 0.43 for Republicans), Multiculturalist Inclusive (β = 0.41, p < 0.001), Locomotion (β = 0.12, p = 0.03), and Assessment (β = 0.11, p = 0.03) were also found to be significant predictors. The model from Step 2 accounted for additional variance, F_{(2, 185)} = 4.31, p = 0.01, however only Promotion motives was a significant predictor (β = 0.11, p = 0.04). This aligned with findings from Study 3 and 4 that Promotion motives and Promotion Framing were more effective in motivating people to bridge build across racial divides.

Furthermore, we used Hierarchical Regression to examine Assessment and Locomotion Mindsets against different types of racial BBAs: Assessment, Locomotion-Political, and Locomotion-Community on all data (N = 429). Racial Group, Political Group, Multiculturalist Inclusive, Self-hatred, Promotion, and Prevention Mindsets were entered as control variables for the first model. For the second model, Assessment and Locomotion were added as predictor variables.

For Assessment Activities, the control variables in the first model explained a significant proportion of variance in willingness to engage, adjusted R² = 0.52, F_{(8, 404)} = 55.72, p < 0.001. Multiculturalist Inclusive (β = 0.48, p < 0.001), Promotion (β = 0.12, p < 0.001), and Prevention (β = 0.08, p = 0.01) were significant positive predictors, while Racial, Political Group, and Self-hatred were not. The second model accounted for a substantial proportion of variance in willingness to engage in Assessment Activities, F_{(2, 402)} = 3.21, p = 0.04. Similar to Study 5, our proposition for assessment mindsets and assessment activities match was not supported. Only Locomotion (β = 0.09, p = 0.01) had a significant positive relationship with willingness to engage in Assessment Activities.

For Locomotion-Political Activities, Racial (β = 46, p < 0.001 for Black respondents), Political Groups (β = −0.35, p = 0.008 for Republicans), Multiculturalism Inclusive (β = 0.10, p = 0.008), Self-Hatred (β = 0.10, p = 0.008), and Promotion (β = 0.10, p = 0.008) were significant among the control variables. The second model significantly predicted additional variance for Locomotion-Political Activities, adjusted R² = 0.42, F_{(2, 402)} = 8.27, p < 0.001. Only Locomotion (β = 0.20, p < 0.001) was a significant predictor for engagement in Locomotion-Political Activities.

For the last type of activities, Locomotion-Community Activities, the first Hierarchical Regression model revealed that Race (β = 0.40, p < 0.001 for Black participants), Multiculturalist Inclusive (β = 0.30, p < 0.001), Promotion (β = 0.19, p < 0.001), and Prevention (β = 0.10, p = 0.04) were significant predictors. In the second model, Locomotion (β = 0.20, p < 0.001) predicted significant additional variance in engagement in Locomotion-Community Activities, F_{(2, 402)} = 9.78, p < 0.001, but not Assessment.

Demographic differences

Independent t-tests and one-way ANOVAs revealed some significant demographic differences. First, for race, Black respondents had higher multiculturalist inclusive scores [F_{(3, 420)} = 4.50, p = 0.01] and were more willing to engage in different types of racial BBA (Assessment, Locomotion Political, and Locomotion Community) than White and non-Black POC. White respondents were more likely to have higher self-hatred scores (i.e., disliking their own racial/ethnic group) than Black and non-Black POC, F_{(2, 420)} = 2.84, p = 0.05. In terms of political affiliation, Democrats were likely to have more Promotion [F_{(2, 420)} = 3.49, p = 0.02], higher multiculturalist inclusive [F_{(2, 420)} = 14.17, p < 0.001], self-hatred scores [F_{(2, 420)} = 6.27, p < 0.001], and were more willing to engage in Assessment [F_{(2, 420)} = 13.44, p < 0.001] and Locomotion-Political Activities [F_{(2, 420)} = 15.38, p < 0.001] than Republicans. However, there was no significant difference for Locomotion-Community activities. Lastly, for gender, women were more likely to have higher multiculturalist inclusive scores (M = 5.81, SD = 1.11) and more willing to engage in Assessment activities (M = 4.05, SD = 0.92) than men (M = 5.29, SD = 1.48 and M = 3.83, SD = 1.06, respectively).

Study 6, the final study in the current series, was an attempt to integrate the theoretical distinctions investigated in studies 1–5 on Prevention, Promotion, Assessment and Locomotion regulatory motives into one 2 × 2 model while also testing for fit, racial-identity value (multiculturalism and self-hatred) and demographic group effects on categorically different types of BBAs (Assessment, Locomotion-Political, Locomotion-Community). CFA found that the items for the four distinct types of motivation loaded cleanly onto four dimensions, each resulting in brief 3-item scales with suitable reliabilities. The findings from the hierarchical regressions on motives, fit, and BBAs were largely consistent with the results from the prior studies, especially regarding the positive effects of both Promotion and Locomotion mindsets and frames for increasing reported willingness to engage in racial bridge building.

General discussion

Given the pernicious forms of polarization that many Western democracies are battling today (McCoy and Somer, 2019), and the tendency for more extreme voices to dominate in these times (Kleinfeld, 2021), this series of studies sought to apply insights from prior research on motivation and peacebuilding to mobilize depolarization efforts at bridge-building across political and racial divides in the U.S. The six studies presented in this article were designed to address the following research questions:

• Are higher levels of the self-regulatory motives of prevention goals for mitigating conflict and promotion goals for benefitting from peacefulness associated with increases in reported willingness to engage in bridge-building activities across group divisions?

• Does the fit between more chronic preventive vs. promotive motivational focus orientations with prevention vs. promotion framing conditions for political/racial bridge-building activities increase the degree of willingness to actively engage in those activities?

• Does the fit between more chronic assessment vs. locomotion motivational mode orientations with assessment vs. locomotive-framed activities racial bridge-building increase the degree of willingness to engage in those activities?

• Finally, are there meaningful differences in (a) political party affiliation, (b) racial group identification, (c) gender, and (d) racial identity values on the tendency to engage in bridge-building activities and the inclination to be motivated by prevention, promotion, locomotion, or assessment orientations? (Exploratory questions).

These questions resulted in the following five hypotheses:

Hypothesis 1: Individuals with higher levels of either prevention or promotion motives for bridge-building will report a greater likelihood of engaging in such activities.

Hypothesis 2: Individuals with stronger prevention motives would be significantly more likely to participate in activities when they were framed as preventing harmful intergroup encounters, compared to individuals with higher levels of promotion motives.

Hypothesis 3: Individuals with higher levels of promotion motives will be significantly more likely to engage in bridge-building activities when framed as promoting positive peace, compared to individuals with strong prevention motives.

Hypothesis 4: Individuals with higher levels of assessment mode preferences will be significantly more likely to participate in bridge-building activities when they were framed as assessment activities, compared to individuals with higher levels of locomotion mode preferences.

Hypothesis 5: Individuals with higher levels of locomotion preferences will be significantly more likely to engage with bridge-building activities when framed as locomotion activities, compared to individuals with stronger assessment mode preferences.

These hypotheses were tested over six studies. Studies 1 and 2 investigated the effects of prevention vs. promotion mindsets and framing effects on engagement with political bridge-building activities across three political groups (i.e., Republican, Democrat, Independent) and three racial groups (i.e., White, Black, non-Black People of Color; Study 2—supporting Hypotheses 1–3). Applying our learnings from the political studies, Studies 3 and 4 examined prevention vs. promotion mindsets, framing effects, and gender, racial, and political differences on engagement with interracial bridge-building activities (Hypotheses 2–3). Study 5 investigated the effects of assessment vs. locomotion mode preferences on engaging with assessment vs. locomotive-framed interracial bridge-building activities (Hypotheses 4–5). Finally, Study 6 combined learnings from our prior studies to examine the effects of prevention vs. promotion mindsets, locomotion vs. assessment preferences, framing effects, and differences in racial (White, Black, non-Black POC), gender, and political groups (Republican, Democrat, Independent) on engagement with interracial bridge-building activities (Hypotheses 1–5).

The results from this series of six studies – which covered a lot of ground regarding the effects of four types of self-regulatory motives on two types of bridge building (political and racial) with other important moderators (political, racial, gender, and value differences)—generally offered support for the five hypotheses, with some important qualifiers. The major empirical findings from these studies include the following:

Higher levels of prevention and promotion motives were found to be generally good for motivating political bridge building. In support of Hypothesis 1, stronger prevention and promotion motivational orientations were both found to be generally associated with a greater willingness to engage in political bridge-building activities. These findings were largely consistent with prior research on motivation and peacebuilding in the Middle East (Coleman et al., 2012) and on regulatory focus theory (Higgins, 1997, 1998).

We found strong “fit” effects, or aligned motive-framing effects, especially for prevention and promotion with political bridge-building. More extreme differences between prevention vs. promotion mindtypes were associated with significant mindset-framing “fit” effects. These motives were more conducive to bridge-building when the framing of the bridge-building activities was consistent with (fit) the motive. This finding is again consistent with other research on regulatory fit theory (Higgins et al., 2001; Higgins, 2005), and can have significant implications for the practical application of this research for bridge-building purposes (see below).

However, promotion motives were found consistently to be best for encouraging interracial bridge-building engagement in the U.S. In the context of interracial bridge-building, we observed a somewhat different pattern regarding the promotion/prevention motivational distinction. Here, we found that a promotion mindset, especially when coupled with promotion framing, proved to be particularly effective in motivating individuals to bridge racial divides in the U.S. The prevention motive and framing did not evidence such positive interracial bridge-building effects. This difference is similarly reflected in other studies on prevention and promotion orientations and conflict management, which has shown that compared to prevention-focused individuals, promotion-focused people achieve superior outcomes in negotiations (Appelt and Higgins, 2010; Galinsky et al., 2015), prefer gain-maximizing strategies (Winterheld and Simpson, 2011) and are more creative and open to change (Friedman and Forster, 2001).

Another possible interpretation of these one-sided effects of promotion on racial bridge building is the threat to validity known as history—in this case the effects of the era in which these studies were conducted. The studies presented here were conducted between December 2021 and December 2022, in the wake of the murder of George Floyd in 2020 and during a spike in protests and public attention to the Black Lives Matter movement. The heightened levels of public awareness of the racial grievances of members of traditionally marginalized racial communities during this period may have provided a ceiling effect for these studies, limiting the impact of preventive types of threat framing on motivating bridge-building—as this type of motivation was already at high levels for many Americans. In such a context, it is likely that the hopefulness provided by promotion mindsets and framing may have had more potential to add additional motivation for engaging in interracial bridge-building activities (see Jasko et al., 2019 for related findings).

Qualitatively different types of bridge-building activities offer important considerations for motivating interracial bridge-building engagement. Our factor analyses of the 24 interracial bridge-building activities gleaned from the literature and offered in Studies 3–5 unearthed an important distinction in terms of categorically different types of BBAs—those with a focus on interrogating political intergroup differences (Political activities) vs. those focused on uniting groups based on commonalities (Community activities). Difference-focused activities aimed to promote recognition and understanding of diverse experiences of members of different identity groups, while community-oriented activities aimed to foster a sense of solidarity by emphasizing shared traits and experiences. These qualitatively distinct types of bridge-building activities were found to be associated with different predictors and were more or less popular among different demographic groups (discussed below). These findings are largely consistent with the considerable body of research on the effects of competitive vs. cooperative goals on intergroup dynamics and conflict resolution (Deutsch, 1973, 2014).

Locomotion mindsets and frames were found to be most effective in motivating interracial bridge-building engagement. Although stronger Assessment orientations were found to motivate some degree of racial bridge-building for Assessment-related activities only, Locomotion orientations and framing were by far the most motivating for the various types of interracial bridging engagement. Future studies should investigate the relative effects of assessment and locomotion for addressing political divisions as well.

Moving (locomotion) together for unity is highly attractive in interracial bridge building. Across most groups, Locomotive-Community gathering activities, or those that encouraged joint actions like “Volunteer for a community garden where people of all races, creeds, and political affiliations can visit, care for, and get access to clean and fresh vegetables” or “Volunteer for a local clean-up day that brings volunteers from different religious and political backgrounds together to pick up trash around their neighborhoods” were found to be most popular and least problematic for increasing engagement.

Demographic and value differences findings. Our research also identified distinct patterns associated with demographic differences and personal values. In general, women consistently demonstrated a stronger inclination for engaging with bridge-building activities compared to men. For political affiliations, Democrats displayed significantly higher promotion and assessment mindsets and levels of bridge-building engagement, whereas Republicans leaned toward being more locomotion-oriented. In addition, BIPOC participants showed greater willingness to engage in Assessment activities, such as listening to a presentation on the history of slavery and discrimination of racial groups in America, than their White counterparts. In terms of racial-identity values, high Inclusive Multiculturalist values emerged as a predictor of higher interracial bridge-building involvement. To address the limitation in our recruitment challenges for Black Republicans, future research should extend their outreach to achieve a well-represented sample.

Implications for practice. In practical terms, the findings from these studies suggest that one immediate means to increase engagement would be for bridge-building organizations to consider using the brief 3-item measures developed for these studies to assess the relative strength of potential participant motives on the four types of regulatory motives (prevention, promotion, assessment, and locomotion), and then offer messaging regarding their activities (via website or written materials) framed in a manner fitting with the different orientations. These types of tailored messaging campaigns have been used effectively with various service organizations, including hospital emergency rooms (see Rotondo and Palazzole, 2013), with public health messaging for youth during COVID-19 (Cheng et al., 2021), and with other internet website health campaigns (Bennett and Glasgow, 2009).

Bridging organizations might also wish to reflect carefully on the specific types of BB activities they offer (reckoning, uniting, assessing, and locomoting), and consider offering different types of activities for distinct demographic groups (e.g., Black vs. White Republicans). Finally, it seems clear that under the current conditions of race relations in the U.S., interracial bridge building should focus on promotion framing for community-gathering activities if the primary goal is to broaden outreach.

As the U.S. heads into yet another increasingly contentious political campaign cycle, a majority of more moderate Americans are feeling awash in a sense of exhaustion and dread regarding our politics (More in Common, 2023). The good news is that such acute, widespread states of dissatisfaction with the status quo have been found to motivate citizens and societies to pivot away from contentious means of political engagement and seek more constructive forms (Zartman, 2000). However, in order for such acute levels of misery to lead to more constructive types of intergroup engagement, leaders and citizens must see a clear alternative path forward. It is our hope that the findings from our research can help to increase engagement with the myriad of political and racial bridge-building that is offered today across our country.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement

The studies involving humans were approved by the Institutional Review Board, Teachers College, Columbia University. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their digital informed consent to participate in this study.

Author contributions

PC: Conceptualization, Funding acquisition, Investigation, Methodology, Resources, Supervision, Writing – original draft, Writing – review & editing. LP: Data curation, Formal analysis, Methodology, Project administration, Software, Visualization, Writing – original draft, Writing – review & editing.

Funding

The author(s) declare that financial support was received for the research, authorship, and/or publication of this article. The authors would like to thank One Million Truths for providing financial and outreach support for the data collection of Studies 3, 4, and 6.

Acknowledgments

The authors would like to thank Mark Eckhardt for being a thought partner in the research and for assisting with the outreach to Black Republican participants.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/frsps.2024.1352284/full#supplementary-material

Footnotes

1. ^This method is based on the assumption that consumers often cannot articulate exactly what they need, want, or like if they are asked directly. In response, it offers an alternative to focus groups, interviews, and surveys in the form of a systematic, structured method of experimentation with consumers - presenting them with a set of experimentally-designed products or concepts and soliciting their preferences. Studies have revealed that it is much easier for consumers to choose a preferred option from a set of discreet choices rather than from pure self-reflection.

2. ^Data for all six studies can be found at osf.io/qw5nb.

3. ^Non-Black People of Color were included due to the fact that after the murder of George Floyd, many racial bridge-building initiatives had been focused only on White and Black narratives, excluding other People of Color, who are also often targets of bias, discrimination and exclusion.

References

Allport, G. W. (1954). The Nature of Prejudice. Boston, MA: Addison-Wesley.

Google Scholar

Appelt, K. C., and Higgins, E. T. (2010). My way: how strategic preferences vary by negotiator role and regulatory focus. J. Exper. Soc. Psychol. 46, 1138–1142. doi: 10.1016/j.jesp.2010.05.010

Crossref Full Text | Google Scholar

Bar-Tal, D., Sharvit, K., Halperin, E., and Zafran, A. (2012). Ethos of conflict: the concept and its measurement. Peace Confl. 18, 40–61. doi: 10.1037/a0026860

Crossref Full Text | Google Scholar

Bennett, G. G., and Glasgow, R. E. (2009). The delivery of public health interventions via the Internet: actualizing their potential. Ann. Rev. Public Health 30, 273–292. doi: 10.1146/annurev.publhealth.031308.100235

PubMed Abstract | Crossref Full Text | Google Scholar

Boxell, L., Gentzkow, M., and Shapiro, J. M. (2020). Cross-country trends in affective polarization. Rev. Econ. Stat. 2020, 1–60. doi: 10.3386/w26669

Crossref Full Text | Google Scholar

Bridge Alliance. (2023). All member organizations. Available online at: https://www.bridgealliance.us/member-organizations (accessed December 5, 2023).

Google Scholar

Bridging Divides Initiative. (2023). Building resilience ecosystem map. Available online at: https://bridgingdivides.princeton.edu/ecosystem-map (accessed December 5, 2023).

Google Scholar

Buchholz, K. (2021). U.S. hate crimes at new decade high. Statista. Available online at: https://www.statista.com/chart/16100/total-number-of-hate-crime-incidents-recorded-by-the-fbi/ (accessed December 5, 2023).

Google Scholar

Cheng, T., Horbay, B., Nocos, R., Lutes, L., and Lear, S. A. (2021). The role of tailored public health messaging to young adults during COVID-19: “There's a lot of ambiguity around what it means to be safe”. PLoS ONE 16:e0258121. doi: 10.1371/journal.pone.0258121

PubMed Abstract | Crossref Full Text | Google Scholar

Coleman, P. T., Moskowitz, H., and Marion, T. (2012). “Mapping peaces: Rule development experimentation and ethos of peace and conflict in Israel/Palestine,” in 25^th Annual Conference of the International Association of Conflict Management. doi: 10.1037/e532162013-125

Crossref Full Text | Google Scholar

Deutsch, M. (1973). “Group formation,” in The Resolution of Conflict, 48–123.

Google Scholar

Deutsch, M. (2014). “Cooperation, competition, and conflict,” in The Handbook of Conflict Resolution: Theory and Practice, eds. P. T. Coleman, M. Deutsch, and E. C. Marcus (New york: ProQuest Ebook Central), 3–27.

Google Scholar

Esterling, K. M., Fung, A., and Lee, T. (2021). When deliberation produces persuasion rather than polarization: measuring and modeling small group dynamics in a field experiment. Br. J. Polit. Sci. 51, 666–684. doi: 10.1017/S0007123419000243

Crossref Full Text | Google Scholar

Faul, F., Erdfelder, E., Buchner, A., and Lang, A.-G. (2009). Statistical power analyses using G^*Power 3.1: Tests for correlation and regression analyses. Behav. Res. Methods 41, 1149–1160. doi: 10.3758/BRM.41.4.1149

PubMed Abstract | Crossref Full Text | Google Scholar

Fishkin, J., Siu, A., Diamond, L., and Bradburn, N. (2021). Is deliberation an antidote to extreme partisan polarization? Reflections on “America in One Room.” Am. Polit. Sci. Rev. 115, 1464–1481. doi: 10.1017/S0003055421000642

Crossref Full Text | Google Scholar

Friedman, R. S., and Forster, J. (2001). The effects of promotion and prevention cues on creativity. J. Person. Soc. Psychol. 81:1001. doi: 10.1037//0022-3514.81.6.1001

PubMed Abstract | Crossref Full Text | Google Scholar

Galinsky, A. D., Rucker, D. D., and Magee, J. C. (2015). “Power: past findings, present considerations, and future directions,” in The Handbook of Personality and Social Psychology, eds. M. Mikulincer, and P. R. Shaver (London: American Psychological Association), 421–460. doi: 10.1037/14344-016

Crossref Full Text | Google Scholar

Gofman, A., and Moskowitz, H. R. (2012). Rule Developing Experimentation: A Systematic Approach to Understand and Engineer The Consumer Mind. Sharjah: Bentham Science.

Google Scholar

Harris, M., and Young, P. (2009). Bridging Community Divides: The Impact of Grassroots Bridge Building Activities. Institute for Voluntary Action Research.

Google Scholar

Higgins, E. T. (1997). Beyond pleasure and pain. Am. Psychol. 52, 1280–1300. doi: 10.1037/0003-066X.52.12.1280

Crossref Full Text | Google Scholar

Higgins, E. T. (1998). Promotion and prevention: regulatory focus as a motivational principle. Adv. Exper. Soc. Psychol. 30, 1–46. doi: 10.1016/S0065-2601(08)60381-0

Crossref Full Text | Google Scholar

Higgins, E. T. (2005). Value from regulatory fit. Curr. Direct. Psychol. Sci. 14, 209–213. doi: 10.1111/j.0963-7214.2005.00366.x

Crossref Full Text | Google Scholar

Higgins, E. T. (2006). Value from hedonic experience and engagement. Psychol. Rev. 113, 439–460. doi: 10.1037/0033-295X.113.3.439

PubMed Abstract | Crossref Full Text | Google Scholar

Higgins, E. T. (2012). “Regulatory focus theory,” in Handbook of Theories of Social Psychology, eds. P. A. M. Van Lange, T. E. Higgins, and A. W. Kruglanski (London: Sage), 483–504. doi: 10.4135/9781446249215.n24

Crossref Full Text | Google Scholar

Higgins, E. T., Friedman, R. S., Harlow, R. E., Idson, L. C., Ayduk, O. N., and Taylor, A. (2001). Achievement orientations from subjective histories of success: promotion pride vs. prevention pride. Eur. J. Soc. Psychol. 31, 3–23. doi: 10.1002/ejsp.27

Crossref Full Text | Google Scholar

Hodis, F. A., Hattie, J. A. C., and Hodis, G. M. (2017). Investigating student motivation at the confluence of multiple effectiveness strivings: a study of promotion, prevention, locomotion, assessment, and their interrelationships. Person. Indiv. Differ. 109, 181–191. doi: 10.1016/j.paid.2017.01.009

Crossref Full Text | Google Scholar

Igielnik, R. (2023). These voters share almost no political beliefs, but they agree on one thing: We're failing as a nation. The New York Times. Available online at: https://www.nytimes.com/2023/08/19/upshot/poll-republicans-democrats-pessimism.html (accessed December 5, 2023).

Google Scholar

Jasko, K., Szastok, M., Grzymala-Moszczynska, J., Maj, M., and Kruglanski, A. W. (2019). Rebel with a cause: personal significance from political activism predicts willingness to self-sacrifice. J. Soc. Issues 75, 314–349. doi: 10.1111/josi.12307

Crossref Full Text | Google Scholar

Johns, R. (2005). One size doesn't fit all: selecting response scales for attitude items. J. Elect. Public Opin. Parties 15, 237–264.

Google Scholar

Kleinfeld, R. (2021). The rise of political violence in the United States. J. Democracy 32, 160–176. doi: 10.1353/jod.2021.0059

Crossref Full Text | Google Scholar

Kruglanski, A. W., Pierro, A., and Higgins, E. T. (2007). Regulatory mode and preferred leadership styles: how fit increases job satisfaction. Basic Appl. Soc. Psychol. 29, 137–149. doi: 10.1080/01973530701331700

Crossref Full Text | Google Scholar

Kruglanski, A. W., Thompson, E. P., Higgins, E. T., Pierro, A., Shah, J. Y., and Spiegel, S. (2000). To ‘do the right thing' or to ‘just do it': locomotion and assessment as distinct self-regulatory imperatives. J. Person. Soc. Psychol. 79, 793–815. doi: 10.1037/0022-3514.79.5.793

PubMed Abstract | Crossref Full Text | Google Scholar

Lawson, A. (2021). Building trust across the political divide. Available online at: https://comment.org/building-trust-across-the-political-divide/ (accessed December 5, 2023).

Google Scholar

McCarty, N. (2019). Polarization: What Everyone Needs to Know. Oxford: Oxford University Press. doi: 10.1093/wentk/9780190867782.001.0001

Crossref Full Text | Google Scholar

McCoy, J., and Somer, M. (2019). Toward a theory of pernicious polarization and how it harms democracies: comparative evidence and possible remedies. Ann. Am. Acad. Polit. Soc. Sci. 681, 234–271. doi: 10.1177/0002716218818782

Crossref Full Text | Google Scholar

Meacham, J. (2022). Lincoln saved American democracy. We can too. TIME. Available online at: https://time.com/6221209/lincoln-saved-american-democracy-we-can-too/ (accessed December 5, 2023).

Google Scholar

More in Common. (2018). Hidden tribes. Available online at: https://hiddentribes.us/ (accessed December 5, 2023).

Google Scholar

More in Common. (2019). The perception gap. Available online at: https://perceptiongap.us/

Google Scholar

More in Common. (2023). SOTU 2023: Speaking to the exhausted majority. Available online at: https://www.moreincommon.com/media/4p5os3ww/sotu-2023.pdf (accessed December 5, 2023).

Google Scholar

Moskowitz, H. R., and Gofman, A. (2007). Selling Blue Elephants: How to Make Great Products That People Want Before They Even Know They Want Them. London: Pearson Education.

Google Scholar

Paluck, E. L., Green, S. A., and Green, D. P. (2019). The contact hypothesis re-evaluated. Behav. Public Policy 3, 129–158. doi: 10.1017/bpp.2018.25

Crossref Full Text | Google Scholar

Pettigrew, T. F., and Tropp, L. R. (2013). When Groups Meet: The Dynamics of Intergroup Contact (1st ed.). New York: Psychology Press. doi: 10.4324/9780203826461

Crossref Full Text | Google Scholar

Pew Research Center. (2023). Americans' dismal views of the nation's politics. Available online at: https://www.pewresearch.org/politics/2023/09/19/americans-dismal-views-of-the-nations-politics/ (accessed December 5, 2023).

Google Scholar

Public Religion Research Institute (2023). Threats to American democracy ahead of an unprecedented presidential election. Available online at: https://www.prri.org/research/threats-to-american-democracy-ahead-of-an-unprecedented-presidential-election/ (accessed December 5, 2023).

Google Scholar

Rotondo, K., and Palazzole, S. (2013). The Mind and Health: Getting to Fewer Hospital Readmissions. iNovum.

Google Scholar

Schmidt, A. F., and Finan, C. (2018). Linear regression and the normality assumption. J. Clin. Epidemiol. 98, 146–151. doi: 10.1016/j.jclinepi.2017.12.006

PubMed Abstract | Crossref Full Text | Google Scholar

Shimano, M. (2021). ‘It's a fight that can be fought and has to be fought': Boston historian Doris Kearns Goodwin sounds alarm on U.S. democracy. Boston. Available online at: https://www.boston.com/news/politics/2021/12/10/doris-kearns-goodwin-alerts-alarm-us-democracy/ (accessed December 5, 2023).

Google Scholar

Southern Poverty Law Center. (2023). Hate map. Available online at: https://www.splcenter.org/hate-map (accessed December 5, 2023).

Google Scholar

Tropp, L. R., and Page-Gould, E. (2014). “Contact between groups,” in APA Handbook of Personality and Social Psychology, Volume 2: Group Processes, eds. M. E. Mikulincer, P. R. Shaver, J. F. Dovidio, J. A. Simpson (London: American Psychological Association), 535–560. doi: 10.1037/14342-020

Crossref Full Text | Google Scholar

Walter, B. F. (2022). How Civil Wars Start: And How to Stop Them. New York: Crown.

Google Scholar

Winterheld, H. A., and Simpson, J. A. (2011). Seeking security or growth: a regulatory focus perspective on motivations in romantic relationships. Association 101, 935–954. doi: 10.1037/a0025012

PubMed Abstract | Crossref Full Text | Google Scholar

Wojcik, S., and Hughes, A. (2019). Sizing up Twitter users. PEW Res. Center 24, 1–23.

Google Scholar

Worrell, F. C., Mendoza-Denton, R., and Wang, A. (2019). Introducing a new assessment tool for measuring ethnic-racial identity: the cross ethnic-racial identity scale–adult (CERIS-A). Assessment 26, 404–418. doi: 10.1177/1073191117698756

PubMed Abstract | Crossref Full Text | Google Scholar

Zartman, W. (2000). “Ripeness: the hurting stalemate and beyond,” in International Conflict Resolution After the Cold War (National Research Council), 225–250.

Google Scholar