Love thy (partisan) neighbor: Brief befriending meditation reduces affective polarization

https://doi.org/10.1177/13684302211020108

Group Processes & Intergroup Relations

2022, Vol. 25(6) 1577 –1593

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/13684302211020108

journals.sagepub.com/home/gpi

Group Processes &

Intergroup Relations

The United States has become increasingly divided

along partisan lines. The discrepancy between how

American adults feel about their own political

party and the political party opposed to their

own—a phenomenon known as affective polariza-

tion—has grown significantly in recent decades.

For example, on a scale ranging from cold (0) to

warm (100), the difference between mean in-party

feeling and mean out-party feeling has risen from

27.0 in 1978 to 45.9 points in 2016 (Boxell et al.,

2020). The widening partisan division has largely

been driven by increasingly negative out-party feel-

ings, which in turn have been fueled by substantive

changes in the political and media environments

(Iyengar et al., 2019). Given that a well-functioning

democracy requires a certain degree of respect and

civility between political opponents, the high levels

of affective polarization in the United States could

have serious implications.

Love thy (partisan) neighbor:

Brief befriending meditation

reduces affective polarization

Otto Simonsson,

Jayanth Narayanan

and Joseph Marks

Abstract

The rising partisan animus between Democrats and Republicans has significant consequences

for American society, both political and nonpolitical. The present study used two preregistered

randomized controlled designs to investigate whether scalable meditation interventions could reduce

affective polarization, relative to baseline scores measured 1 week earlier, in American adults (Study

1: N = 353; Study 2: N = 246) who affiliated with either the Democratic Party or the Republican

Party. The results suggest that a brief befriending meditation can reduce affective polarization between

Democrats and Republicans by increasing positive feelings relatively more for the political outgroup

than the political ingroup.

Keywords

affective polarization, befriending, intergroup relations, meditation, mindfulness

Paper received 15 August 2020; revised version accepted 4 May 2021.

University of Oxford, UK

National University of Singapore, Singapore

University College London, UK

Corresponding author:

Otto Simonsson, Department of Sociology, University of

Oxford, 42–43 Park End Street, Oxford, England OX1 2JD,

UK.

Email: [email protected]

1020108

GPI0010.1177/13684302211020108Group Processes & Intergroup RelationsSimonsson et al.

research-article2021

Article

1578 Group Processes & Intergroup Relations 25(6)

The evidence to date suggests that the rising

partisan animus between Democrats and

Republicans has both political and nonpolitical

consequences (Druckman et al., 2020a; Iyengar

et al., 2019; McConnell et al., 2018). For instance,

affective polarization has been shown to shape

beliefs about ostensibly apolitical issues such as

responses to a pandemic (Druckman et al.,

2020b); increasingly negative out-party feelings

have increased voting along party lines and have

discouraged elected officials from working with

members of the opposing party (Abramowitz &

Webster, 2016); and partisan animosity has under-

mined the level of trust toward the government,

especially when the political outgroup has been in

power (Hetherington & Rudolph, 2015). Hence,

it is vital to better understand the causes of affec-

tive polarization and, perhaps more importantly,

to investigate which interventions can reduce it.

Previous Psychological

Approaches to Reducing

Affective Polarization

Previous research suggests that psychological

interventions can reduce affective polarization

in the United States. For example, Levendusky

(2018) found that priming American national

identity, thereby making a superordinate group

identity more salient than partisan identity,

decreased affective polarization between

Democrats and Republicans (see also Levine

et al., 2005). The priming of national identity

could be used as an effective strategy to reduce

affective polarization among partisans, but it

also has the potential to exacerbate affective

polarization toward nonnationals, such as

undocumented immigrants (Wojcieszak &

Garrett, 2018). It may therefore be valuable to

consider other psychological interventions to

reduce affective polarization.

An alternative approach that holds promise is

imagined intergroup contact (Crisp & Turner,

2009), which involves imagining positive interac-

tions with one or more outgroup members. The

approach has demonstrated effectiveness in

improving intergroup attitudes and has recently

also been shown to reduce affective polarization

between Democrats and Republicans (Miles &

Crisp, 2014; Warner & Villamil, 2017; see also

Bond et al., 2018), with perceived commonality

between the self and the political outgroup as an

important underlying mechanism (Warner et al.,

2020; Wojcieszak & Warner, 2020). It might be

difficult, however, to get partisans with high lev-

els of affective polarization to voluntarily imagine

positive interactions with the political outgroup.

Other psychological interventions that might be

more easily accessible should therefore be

investigated.

Meditation Practices and

Affective Polarization

There has recently been an upsurge of interest in

mindfulness meditation (Simonsson, Martin, &

Fisher, 2020), with mindfulness-based programs

reported in the education system, the workplace,

the military (Creswell, 2017), and parliaments

around the world (Bristow, 2019). The practice of

mindfulness meditation involves bringing an

open and nonjudgmental attention to present-

moment experiences (Kabat-Zinn, 2003), which

facilitates emotion regulation and may reduce

negative outgroup attitudes through lower inter-

group anxiety (Price-Blackshear et al., 2017). It

has been shown to increase prosocial behavior

(Donald et al., 2019) and has also been suggested

as a possible tool to attenuate affective polariza-

tion in the United States (Klein, 2020). While the

effect of mindfulness meditation on affective

polarization remains untested, Petersen and

Mitkidis (2019) found no evidence of an associa-

tion between trait mindfulness and political toler-

ance (a construct that is conceptually linked to

affective polarization) nor an effect of a brief

mindfulness meditation on political tolerance.

The authors suggested that the results may be dif-

ferent with prolonged training in mindfulness

meditation or with other types of meditation that

are more oriented toward facilitating feelings of

compassion for others. Indeed, trait compassion

and similar constructs such as empathic concern

have been associated with more positive and

Simonsson et al. 1579

reconciliatory attitudes in intergroup settings

(Klimecki, 2019), which indicates that compas-

sion could play an important role in reducing

affective polarization (but see also Simas et al.,

2020).

Another type of meditation that has also been

taught in parliaments—as part of an 8-week

mindfulness-based program—is befriending

meditation (Williams & Penman, 2011). The key

difference between befriending meditation and

mindfulness meditation is that the former

involves cultivating a sense of kindness and

goodwill toward oneself and others, whereas the

latter involves paying attention, in a particular

way, to the present moment. Befriending medita-

tion may therefore influence affective polariza-

tion through the same mechanism as imagined

intergroup contact. Like imagined intergroup

contact, befriending meditation engages the

imagination and involves bringing different types

of people to mind, including people who may be

perceived as difficult or challenging. The differ-

ence between these approaches is that befriend-

ing meditation involves directly generating a

sense of kindness and goodwill toward whoever

has been brought to mind, while imagined inter-

group contact interventions involve imagining

positive interactions with one or more outgroup

members. Previous research has shown that

befriending meditation and similar practices such

as loving-kindness and compassion meditation

can strengthen compassion toward others

(Galante et al., 2014), increase perceived com-

monality between the self and others (Kok &

Singer, 2017), and boost positivity toward stran-

gers who were not brought to mind during the

meditation (Hutcherson et al., 2008). Indeed,

positive attitudes toward one outgroup may posi-

tively influence attitudes toward other outgroups

(Pettigrew, 2009; van Laar et al., 2005), and could

therefore make befriending meditation a suitable

intervention to reduce affective polarization.

Research from recent population studies

shows an overall increase in meditation use in the

United States (Black et al., 2018; Clarke et al.,

2018), with millions of Americans voluntarily

choosing to meditate. Given that this is not true

for the other psychological approaches designed

to reduce affective polarization, it does suggest

meditation might be a more scalable intervention.

Here, we therefore investigate the effects of brief

meditation interventions on affective polarization

between Democrats and Republicans.

The Present Research

Our overarching hypothesis was that compas-

sion reduces affective polarization, while mind-

fulness does not. We tested this hypothesis by (a)

examining the naturally occurring relationships

between affective polarization, trait compassion,

and trait mindfulness, and (b) directly influenc-

ing compassion and mindfulness by having par-

ticipants complete either a befriending meditation

or a mindfulness meditation. Building on prior

work (Klimecki, 2019; Petersen & Mitkidis,

2019), we hypothesized that trait compassion but

not trait mindfulness would be negatively associ-

ated with affective polarization in American

adults who affiliated with either the Democratic

Party or the Republican Party. We also hypothe-

sized that a brief befriending meditation would

reduce affective polarization between Democrats

and Republicans.

In Study 1, using an online sample of

American adults, we assessed whether individual

differences in trait compassion and trait mindful-

ness were associated with preintervention varia-

tion in affective polarization. We also investigated

whether participants randomly assigned to com-

plete a brief befriending meditation would exhibit

a greater reduction in affective polarization, rela-

tive to baseline levels measured 1 week earlier,

than participants who completed a brief mindful-

ness meditation or listened to a talk on the topic

of mindfulness meditation. In Study 2, using a

separate online sample of American adults, we

only compared the impact of a brief befriending

meditation on affective polarization with that of

the control task (listening to a talk on the topic of

mindfulness meditation). We did not ask any par-

ticipants to complete a mindfulness meditation in

Study 2. Finally, by pooling the data from Study 1

and Study 2, we conducted additional exploratory

1580 Group Processes & Intergroup Relations 25(6)

analyses to test whether the effects varied across

measurements of affective polarization or self-

identified political affiliation.

Study 1

Materials and Methods

The study (hypotheses, design plan, sampling

plan, variables, and analysis plan) was preregis-

tered on the Open Science Framework (https://

osf.io/qscfk). Sample size was determined using

a power analysis (G*Power Version 3.1.9.2; Faul

et al., 2007). A sample size of 243 participants (81

per condition) achieves 80% power to detect a

small effect (effect size f = 0.10, partial η

= .01)

with an alpha of .05. Hence, the aim was to

recruit 450 participants in total at Time 1, with

150 participants per condition, assuming that not

all would complete the second part of the study.

Participants. Participants were recruited on Pro-

lific Academic (https://app.prolific.co). We used

the custom prescreening function to only recruit

American citizens over 18 years of age who spoke

English and who affiliated with either the Demo-

cratic Party or the Republican Party. Stratified

random sampling was used to ensure we had

approximately equal numbers of Democrats and

Republicans, as Democrats tend to be overrepre-

sented on online platforms (Arechar & Rand,

2020; Paolacci & Chandler, 2014). Two hundred

and twenty-five Democrats and 226 Republicans

completed the first part of the study on March

17, 2020; 81.78% of the Democrats and 77.88%

of the Republicans completed the second part of

the study 1 week later, over a 2-day period from

March 24 to 26, 2020. The participants who failed

the attention checks were removed, leaving the

final number of participants at 353 (172 females,

178 males, and three who preferred not to pro-

vide gender information; 182 Democrats, 171

Republicans; age range: 18–76 years, M = 37.15,

SD = 13.05). All participants gave informed con-

sent and provided demographic information

(age, gender, education, meditation experience,

first language, and whether they have asthma)

through Qualtrics (https://www.qualtrics.com/),

the platform used to collect the data for the study.

The participants were naïve to the purpose of the

experiment and were simply told that the aim was

to “expand the existing knowledge in the fields

of mindfulness, compassion, and political atti-

tudes.” All participants were paid £1.00 for par-

ticipating in the first session and £3.00 for

participating in the second session.

Design and procedure. We utilized a longitudinal

randomized controlled design to assess the causal

effects of meditation practices on affective polar-

ization. Specifically, we examined whether mind-

fulness and/or befriending meditation reduced

affective polarization, relative to baseline scores

measured 1 week earlier. Changes in affective

polarization over time were compared to a con-

trol condition, in which participants listened to a

talk about mindfulness meditation, to ensure that

fluctuations were not due to a temporal con-

found. The current study therefore had three

between-subject factors (mindfulness, befriend-

ing, control) and two within-subject factors (time:

preintervention, postintervention). Randomiza-

tion checks confirmed that the between-subject

conditions were balanced on demographic, politi-

cal, trait, and polarization variables (see online

supplemental material).

Preintervention stage. In the preintervention

stage (Time 1), participants were asked to indicate

their political party affiliation (Democrat, Repub-

lican, independent, other, none), and answered

five items designed to assess how strongly they

identified with that party (e.g., “How important

is your identity as a Democrat to you?”; Druck-

man & Levendusky, 2019). They were also asked

to complete two questionnaires to examine their

levels of trait mindfulness (Cognitive and Affec-

tive Mindfulness Scale-Revised; Feldman et al.,

2007) and trait compassion (Santa Clara Brief

Compassion Scale; Hwang et al., 2008).

Affective polarization was then assessed using

three validated measures that are all strongly

related to one another (Druckman & Levendusky,

2019): a feeling thermometer (asks respondents

Simonsson et al. 1581

to rate how cold or warm they feel toward

Democratic and Republican voters [party candi-

dates, elected officials]); a trait rating measure

(asks respondents to rate how well positive

traits—such as intelligence, honesty, and generos-

ity—and negative traits—such as hypocrisy, self-

ishness, and meanness—describe the two parties’

voters [party candidates, elected officials]); a trust

measure (asks respondents how much of the time

they think they can trust the two parties’ voters

[party candidates, elected officials] to do what is

right for the country). Our preregistration speci-

fied that we would specifically ask participants to

rate each party’s voters on these three measures.

After submitting the preregistration but before

collecting the data, we decided to also assess par-

ticipants’ feelings, trait ratings, and trust toward

party candidates and elected officials, as done by

Druckman and Levendusky (2019). We analyzed

these two sets of data separately and note in the

Results section that the preregistered analysis (i.e.,

the analysis of affective polarization when par-

ticipants are asked to focus on voters) is confirm-

atory, while the analysis of participants’ feelings,

trait ratings, and trust toward party candidates

and officials must be considered exploratory.

Intervention stage. One week later (Time 2),

participants were invited to complete the second

part of the study in which they were randomly

assigned to one of three conditions (mindfulness,

befriending, control). Participants in the mind-

fulness condition listened to a 10-minute guided

mindfulness meditation (focused on sounds and

thoughts). In the recording, they were instructed

to (a) settle into a comfortable posture, (b) bring

awareness to sounds for a few minutes, (c) bring

awareness to thoughts for a few minutes, and (d)

bring awareness to their breath and body in the

last moments of the guided meditation. Partici-

pants in the befriending condition listened to a

10-minute guided befriending meditation. In the

recording, they were instructed to (a) settle into

a comfortable posture, (b) bring kindness and

friendship to themselves by silently saying: “May

I be free from suffering, may I be happy and

healthy, may I have ease of being,” (c) bring to

mind a loved one and wish them well in the same

way, (d) bring to mind a stranger and wish them

well in the same way, (e) bring to mind a difficult

person and wish them well in the same way, (f)

extend kindness and friendship to all living beings

and wish them well in the same way, and (g)

bring awareness to their breath and body in the

last moments of the guided meditation. Partici-

pants in the active control condition listened to

a 10-minute audio recording about mindfulness

meditation. In the recording, they were educated

about mindfulness meditation and the evidence

to date on the neuroscience of mindfulness and

the efficacy of mindfulness-based programs.

The word mindfulness was never mentioned in

the mindfulness condition, but it was mentioned

repeatedly in the talk about mindfulness medita-

tion, which was focused on the concept (rather

than the practice) of mindfulness. The audio

recording for the control condition was a com-

bination of talks by Professor Mark Williams,

who has delivered mindfulness teachings to Brit-

ish politicians together with Chris Cullen from

the Oxford Mindfulness Centre (Bristow, 2019).

The audio recordings used in the mindfulness

and befriending conditions were also recorded by

Professor Mark Williams and were derived from

Mindfulness: A Practical Guide to Finding Peace in a

Frantic World (Williams & Penman, 2011), which

forms the basis of an eight-week mindfulness

course that has been taught in the UK Parlia-

ment. After listening to the audio recording,

participants were assessed on the three affective

polarization measures (the feeling thermometer,

the trait rating measure, and the trust measure).

Attention checks. In the preintervention stage,

participants were presented with one attention

check question embedded in the trait mindfulness

questionnaire, “Please select the disagree option

for this question,” and one embedded in one

of the trait rating measures, “Please click very

well.” Participants who did not give the desired

responses to these questions (n = 32) were not

invited to complete the second part of the study.

In the second part of the study, participants

were presented with an attention check that

1582 Group Processes & Intergroup Relations 25(6)

followed the audio recording to which they had

been assigned: mindfulness condition (“During

the audio recording, I was instructed to. . .”; 1 =

focus on sounds and thoughts, 2 = recite a mantra, 3 =

stretch my body), befriending condition ( “During

the audio recording, I was instructed to. . .”; 1 =

generate feelings of kindness toward myself and others, 2

= memorize numbers and dates, 3 = stretch my body),

control condition (“During the audio recording, I

learnt about. . .”; 1 = mindfulness and meditation, 2

= sports and gymnastics, 3 = politics and law).

Participants who gave the wrong answer on these

questions (n = 5) were excluded from data analy-

ses, as were those who gave the same answer for

every item on the trait mindfulness questionnaire

(as this included reverse-coded items).

Manipulation check. After listening to the audio

recording to which they had been assigned,

all participants also completed a manipulation

check: “How much did you generate feelings

of kindness and goodwill toward others during

the recording you listened to?” and “How much

did you focus on the present moment during the

recording you listened to?” Responses were rated

on a 5-point Likert scale (1 = not at all, 5 = very

much). Participants in the befriending condition

were expected to provide higher scores on the

kindness question than participants in the other

two conditions, while those in the mindfulness

condition were expected to provide higher scores

on the presence question.

Statistical analyses. As specified in our preregistra-

tion, we computed composite measures of affec-

tive polarization for each time point. Aggregating

numerous closely linked measures of the same

latent construct can reduce measurement noise

(Rushton et al., 1983), increase test–retest reliabil-

ity (Eisenberg et al., 2019), and alleviate the need

for researchers to conduct multiple statistical

tests. We based our decision to aggregate affec-

tive polarization scores on the feeling thermom-

eter, trait rating, and trust measures on previous

research showing that these three measures are

strongly correlated with one another (Druckman

& Levendusky, 2019).

To compute the composite affective polariza-

tion scores, we first created a net trait rating for

each group of voters per participant by subtract-

ing the sum of the negative traits from the sum

of the positive traits, consistent with earlier stud-

ies (Druckman & Levendusky, 2019; Iyengar

et al., 2012). Next, ratings on the feeling ther-

mometer, net trait rating, and trust measure were

standardized. Note that we did not state that we

would do this in our preregistration, however,

standardization was necessary as the ratings were

provided on different scales. Specifically, we line-

arly scaled scores on each measure onto a 0 to

100 scale. Affective polarization scores on each

measure at each time point were then computed

by calculating the difference between partici-

pants’ rating of their own party’s voters and their

rival party’s voters. We then aggregated these

three measures separately for each time point to

produce a Time 1 and Time 2 composite affective

polarization measure. A reliability analysis per-

formed on the Time 1 data revealed that this

composite measure has good internal consistency

(Cronbach’s alpha = .86, average interitem cor-

relation = .67; all interitem correlations are

reported in the supplemental material). We

repeated these steps using participants’ feelings,

trait ratings, and trust toward each party’s candi-

dates and elected officials to create a second com-

posite measure of affective polarization with a

different target (Cronbach’s alpha = .85, average

interitem correlation = .66).

To test our first preregistered hypothesis—

that trait compassion, but not trait mindfulness,

negatively correlates with preintervention affec-

tive polarization scores—we assessed the bivari-

ate (Pearson) correlations between baseline (Time

1) composite affective polarization scores, trait

mindfulness scores, and trait compassion scores.

We also assessed the partial correlation between

baseline composite affective polarization scores

and trait mindfulness, controlling for trait com-

passion, and the partial correlation between base-

line affective polarization scores and trait

compassion, controlling for trait mindfulness.

To test our second preregistered hypothesis—

that befriending meditation reduces affective

Simonsson et al. 1583

polarization more than mindfulness meditation

or listening to a talk about mindfulness medita-

tion—we tested whether there was an effect of

the meditation practice condition on composite

affective polarization change scores. In a slight

deviation from the analysis plan set out in our

preregistration, we calculated the change in affec-

tive polarization by subtracting the Time 1 affec-

tive polarization composite scores from the Time

2 scores. We then performed a one-way ANOVA

on these change scores, with the meditation prac-

tice condition entered as the between-subject fac-

tor. The rationale for deviating from the

preregistered analysis plan as well as the results of

the preregistered analysis are reported in the sup-

plemental material.

Results

Confirmatory analyses

Correlations between baseline affective polarization,

trait mindfulness, and trait compassion. Our over-

arching hypothesis was that feeling compassion

reduces affective polarization, while being mind-

ful does not. We therefore investigated the natu-

rally occurring relationship between reported

trait compassion and preintervention composite

affective polarization scores (based on ratings

of Democratic and Republican voters), to test

whether people with higher trait compassion

scores exhibit lower levels of affective polariza-

tion. Contrary to our hypothesis, a one-tailed

bivariate (Pearson) correlation did not reveal a

significant relationship between trait compassion

and baseline levels of affective polarization, r(353)

= .04, p = .257. There was also a nonsignificant

relationship between reported trait mindfulness

and baseline affective polarization, r(353) =

−.06, p = .140. A one-tailed partial correlation

between trait compassion and baseline affective

polarization revealed that this null finding held

when controlling for trait mindfulness, r

partial

(350)

= .04, p = .237; and a one-tailed partial correla-

tion between trait mindfulness and baseline affec-

tive polarization, controlling for trait compassion,

also showed no significant association, r

partial

(350)

= −.06, p = .132.

Effects of meditation practices on affective polari-

zation. As a manipulation check, we assessed

whether the befriending meditation induced

feelings of compassion more than the mindful-

ness meditation and the control task. A one-

way ANOVA revealed a significant effect of

the meditation practice condition on how much

participants generated feelings of kindness and

goodwill toward others while listening to the

audio recording, F(2, 352) = 20.44, p < .001,

= .11. Post hoc Tukey’s honestly significant

difference (HSD) tests demonstrated that par-

ticipants in the befriending condition reported

generating more compassion (M = 3.73, SD =

1.01) than participants in the mindfulness condi-

tion (M = 2.94, SD = 1.15, p < .001) and those

in the control condition (M = 3.01, SD = 1.00,

p < .001). We also tested whether the mindful-

ness meditation was more effective at inducing

a mindful state than the befriending medita-

tion and the control task. A one-way ANOVA

revealed that the meditation practice condition

did not have a significant effect on how much

participants reported focusing on the present

moment while listening to the audio recording,

F(2, 352) = 0.56, p = .570, η

= .003, sug-

gesting that participants who completed the

mindfulness meditation were not more likely to

have entered a mindful state than those in the

befriending and control conditions.

We then tested whether the meditation prac-

tices reduced levels of affective polarization.

Specifically, we assessed whether pre- to postint-

ervention changes in the composite affective

polarization scores (based on ratings of

Democratic and Republican voters) varied

between meditation practice conditions. A one-

way ANOVA revealed a marginal but not signifi-

cant effect of meditation practice condition on

these change scores, F(2, 352) = 2.59, p = .076,

= .02 (see Table 1).

Exploratory analyses. We repeated the aforemen-

tioned analyses using the composite affective

polarization scores comprised of participants’

ratings of Democratic and Republican candi-

dates and elected officials, as a robustness check.

1584 Group Processes & Intergroup Relations 25(6)

Again, we found no relationship between trait

compassion and baseline levels of affective

polarization, r(353) = .06, p = .135, even when

controlling for trait mindfulness, r

partial

(350) =

.06, p = .124. Nor was there a correlation

between trait mindfulness and baseline affective

polarization, r(353) = −.05, p = .199, or a par-

tial correlation between these two variables

when controlling for trait compassion,

partial

(350) = −.05, p = .182.

We did, however, find a significant effect of

meditation practice condition on the pre- to

postintervention change in composite affective

polarization scores. A one-way ANOVA revealed

a significant effect of meditation practice condi-

tion on affective polarization change scores, F(2,

352) = 3.66, p = .027, η

= .02. Post hoc Tukey’s

HSD tests indicated that participants in the

befriending condition exhibited a greater reduc-

tion in affective polarization (M = −4.95, SD =

12.69) than those in the control condition (M =

−0.86, SD = 14.30, p = .038) and a marginally

greater reduction than those in the mindfulness

condition (M = −1.35, SD = 10.91, p = .075).

There was no difference in change scores between

the mindfulness condition and the control condi-

tion (p = .953). Thus, although our confirmatory

analyses did not produce significant results, we

did find some evidence to support the hypothesis

that befriending meditation influences affective

polarization.

Closer inspection of the change scores sug-

gested that befriending meditation may not affect

all measures of affective polarization equally.

Table 1 suggests that there was little change

between Time 1 and Time 2 affective polarization

scores on the trust measure, both when partici-

pants rated parties’ voters and when they rated

parties’ candidates and elected officials, while

there was pre- to postintervention movement in

the expected direction on the feeling thermome-

ter and the net trait rating measure. Moreover,

although the three Time 1 affective polarization

measures showed good internal consistency and

were strongly related to one another, the internal

consistency of the change scores (Time 2 scores

minus Time 1 scores) was low and the interitem

correlations were weak, both when participants

rated each party’s voters (Cronbach’s alpha = .47,

average interitem correlation = .23) and when

they rated each party’s candidates and elected

officials (Cronbach’s alpha = .40, average inter-

item correlation = .18).

Discussion

In Study 1, confirmatory analyses found that

affective polarization was not significantly associ-

ated with trait compassion or trait mindfulness,

and that a brief befriending meditation had a

marginal but not significant effect on affective

polarization when Democratic and Republican

Table 1. Mean change scores on each measure of affective polarization for participants in the befriending,

mindfulness, and control conditions.

Condition Mean change in affective polarization (standard deviation)

Feeling

thermometer

Net trait rating Trust

Voters

Befriending −6.85 (23.00) −3.65 (14.86) −2.10 (19.14)

Mindfulness −2.22 (18.66) −2.15 (13.39) −1.47 (20.40)

Control −0.26 (21.50) −0.57 (18.13) 0.22 (23.30)

Party candidates and elected officials

Befriending −6.29 (20.84) −6.88 (15.98) −1.68 (18.04)

Mindfulness −3.69 (16.79) −0.58 (15.64) 0.21 (19.12)

Control −1.60 (22.66) 0.11 (19.47) −1.09 (20.78)

Simonsson et al. 1585

voters were the target of the ratings. However,

exploratory analyses found that a brief befriend-

ing meditation significantly reduced affective

polarization when Democratic and Republican

candidates and elected officials were the target of

the ratings, which provided some evidence to

support the hypothesis that befriending medita-

tion influences affective polarization.

In addition to the confirmatory and explora-

tory analyses, change scores in Table 1 indicate

that the effects of a brief befriending meditation

varied across the measurements of affective

polarization (feeling thermometer, net trait rating,

trust measure), with the largest changes observed

on the feeling thermometer. In Study 2, we there-

fore tested whether the befriending meditation

caused a pre- to postintervention reduction in

affective polarization, relative to an active control

condition, on the feeling thermometer only.

Study 2

Materials and Methods

The study was preregistered on the Open Science

Framework (https://osf.io/sd3uf). Sample size

was determined with a power analysis (G*Power

Version 3.1.9.2; Faul et al., 2007). We entered the

mean change scores on the feeling thermometer

for participants in the befriending condition and

the control condition of Study 1 (M

befriending

−6.85, M

control

= −0.26), as well as the standard

deviations of these change scores (SD

befriending

23.00, SD

control

= 21.50), into the power analysis

and found that a sample size of 284 participants

would achieve 80% power to detect an effect size

d = 0.30, with an alpha of .05. We therefore

recruited 350 participants in total at Time 1,

assuming that the dropout and exclusion rates

would be similar to those in Study 1.

Participants. The recruitment procedure was the

same as that for Study 1, except that participants

who completed Study 1 were prevented from also

taking part in Study 2. Participants were paid

£0.50 for participating in the first session and

£2.00 for participating in the second session (the

compensation was slightly less than in Study 1

because questions were removed, thus the time it

took to complete the study was shortened). One

hundred and seventy-five Democrats and 175

Republicans completed the first part of the study

on December 2, 2020; 74.29% of the Democrats

and 79.43% of the Republicans who completed

the first part also completed the second part of

the study 1 week later, over a 3-day period from

December 9 to 11, 2020. Due to a technical error

in the recruitment process, 18 participants who

had not completed the first session took part in

the second session of the study. These partici-

pants, as well as those who failed attention checks,

were excluded from the analysis, leaving the final

number at 246 (123 females, 120 males, two iden-

tified as “other” when asked about gender, and

three preferred not to provide gender informa-

tion; 125 Democrats, 121 Republicans; age range:

18–96 years, M = 34.18, SD = 12.77).

Design and procedure. As in Study 1, we used a lon-

gitudinal randomized controlled design to investi-

gate whether reductions in affective polarization

over time were greater for participants who com-

pleted a brief befriending meditation than for

those who listened to a talk about mindfulness

meditation. We did not include a mindfulness

meditation condition in this study and therefore

had a mixed design with two between-subject fac-

tors (befriending, control) and two within-subject

factors (time: preintervention, postintervention).

Preintervention stage. In the preintervention

stage (Time 1), participants were asked to indicate

their political party affiliation (Democrat, Repub-

lican, independent, other, none), and answered

five items designed to assess how strongly they

identified with that party (e.g., “How important is

your identity as a Democrat to you?”; Druckman

& Levendusky, 2019). Trait mindfulness and trait

compassion were not recorded in this study.

Participants then rated Democratic and

Republican voters as well as party candidates

and elected officials on the feeling thermometer,

trait rating measure, and trust measure, as in

Study 1. Our preregistration specified that we

1586 Group Processes & Intergroup Relations 25(6)

would focus the confirmatory analysis on par-

ticipants’ ratings of each party’s voters on the

feeling thermometer, however, we decided to

collect data on all three measures (with both

voters as well as party candidates and elected

officials specified as the target) to allow for fur-

ther exploratory analyses.

Intervention stage. One week later (Time 2),

participants were invited to complete the second

part of the study in which they were randomly

assigned to one of two conditions (befriending,

control). Participants in the befriending condi-

tion listened to a 10-minute guided befriending

meditation, while those in the control condition

listened to a 10-minute talk about mindfulness

meditation. The audio clips were the same as

those played to participants in the befriending

and control conditions in Study 1.

Attention checks. In the preintervention stage,

participants were presented with an attention

check question embedded in one of the trait

rating measures, “Please click very well.” In the

second session, participants were presented with

an attention check immediately after the inter-

vention to ensure they had listened to the audio

clip. The question asked and the response options

were the same as those for Study 1. Participants

who gave the wrong answer on these questions

(n = 3) were excluded from data analyses, as were

those who failed the attention check question in

the first session (n = 21).

Manipulation check. After listening to the audio

recording to which they had been assigned, all

participants also completed a manipulation check.

The question asked and the response options

were the same as those for Study 1.

Statistical analyses. As specified in our preregistra-

tion, in this study our focal measure of affective

polarization was built using participants’ ratings

on the feeling thermometer only. Specifically, we

subtracted participants’ ratings of outgroup vot-

ers from their ratings of ingroup voters, sepa-

rately for each time point. Change scores were

then computed by calculating the difference

between Time 1 and Time 2 affective polarization

scores. An independent samples t test was con-

ducted to test our preregistered hypothesis that

participants in the befriending condition would

exhibit a greater reduction in affective polariza-

tion than those in the control condition.

Results

The results of the independent samples t test

indicated that there was a significant difference

in change scores between conditions, t(244) =

2.47, p = .014, d = 0.31. As hypothesized, there

was a greater reduction in affective polarization

for participants who completed the befriending

meditation (befriending condition: M = −6.55,

SD = 19.74) than for those who listened to the

talk about mindfulness meditation (control con-

dition: M = −0.65, SD = 17.83; see Figure 1).

This finding held when participants who failed

attention checks were not excluded (see supple-

mental material). This suggests that befriending

meditation does influence affective polarization

when the difference in feelings toward ingroup

and outgroup voters is used as the dependent

variable.

Discussion

In Study 2, our confirmatory analysis found that

a brief befriending meditation significantly

reduced affective polarization, relative to an

active control condition, as measured by the feel-

ing thermometer only. While the findings are

novel, the results are broadly in line with those

from studies that have used similar practices to

reduce prejudice and intergroup bias in other

contexts (Berger et al., 2018; Kang et al., 2014).

The results from these two studies still leave

many questions unanswered. Given that Study 1

and Study 2 used the same befriending and con-

trol conditions, it was possible to pool the data

from the two studies to increase the sample size

and allow additional exploratory analyses to be

conducted beyond the overall intervention effect

(Bangdiwala et al., 2016).

Simonsson et al. 1587

Pooling Data from Study 1 and

Study 2

We pooled the data from Study 1 and Study 2 to

answer two additional questions. The first is

whether the effect of befriending meditation on

affective polarization varies depending on the

instrument used to measure it. Study 2 showed

that befriending meditation reduces the differ-

ence in feelings toward ingroup and outgroup

voters, while (as shown in Table 1) it may have

less (or no) effect on how much members of rival

groups trust each other. The second so far unan-

swered question is whether the effect of befriend-

ing meditation on affective polarization differs

for Democrats and Republicans. While political

affiliation has not yet been investigated as a mod-

erator in meditation research, there are several

neurocognitive and personality differences across

the political spectrum that could possibly influ-

ence individual responses to meditation interven-

tions (Amodio et al., 2007; Carney et al., 2008;

Jost et al., 2003). For example, conservatives tend

to value ingroup loyalty more than liberals

(Graham et al., 2009) and may therefore respond

differently to the instructions to bring kindness

and friendship toward the various target catego-

ries (oneself, a loved one, a stranger, a difficult

person, all living beings) in a befriending

meditation.

Methods

Data from Study 1 and Study 2 were pooled to

conduct additional exploratory analyses. We

excluded participants who failed attention checks,

using the same criteria as used in the respective

study, and those from Study 1 who were assigned

to the mindfulness condition, as there was no

mindfulness condition in Study 2.

We computed change scores on each measure

of affective polarization. As in Study 1, we built a

net trait rating score for each group of voters per

participant by subtracting the sum of the negative

traits from the sum of the positive traits. We then

standardized the feeling thermometer, net trait

rating, and trust measures by linearly scaling the

ratings on each measure onto a 0 to 100 scale.

Affective polarization scores on each measure at

each time point were computed by calculating the

difference between participants’ rating of their

own party’s voters (as well as candidates and

elected officials) and their rival party’s voters (and

candidates and elected officials). Lastly, we com-

puted the change in affective polarization on each

measure by subtracting Time 1 scores from Time

2 scores.

Results

Effects of befriending meditation on different measures of

affective polarization. To test whether befriending

meditation reduced affective polarization on

some measures more than others, we entered the

change scores from all six affective polarization

measures (feeling thermometer, net trait rating,

and trust differences for voters, and candidates

and elected officials) into a 2 (target: voters, can-

didates and elected officials) x 3 (measure: feeling

thermometer, net trait rating, trust) x 2 (medita-

tion practice condition: befriending, control) x 2

(study number: 1, 2) mixed ANOVA. The target

and measure were entered as within-subject fac-

tors, while meditation practice condition and

study number were entered as between-subject

factors.

The results revealed a main effect of medita-

tion practice condition, F(1, 476) = 11.71, p =

.001, η

= .02, a main effect of the measure used

Figure 1. Change in affective polarization over time

by condition.

Note. Affective polarization is calculated by subtracting

participants’ ratings of outgroup voters on the feeling ther-

mometer from their ratings of ingroup voters. The reduction

in affective polarization was greater in the befriending condi-

tion than in the control condition.

1588 Group Processes & Intergroup Relations 25(6)

to record affective polarization, F(2, 475) = 9.86, p

< .001, η

= .04, and an interaction between

meditation practice condition and measure, F(2,

475) = 3.31, p = .037, η

= .01. The main effect

of meditation practice condition indicated that,

when collapsing the data across the other inde-

pendent variables, participants who completed the

befriending meditation exhibited a greater reduc-

tion in affective polarization (M = −4.31, SE =

0.74) than those who completed the control task

(M = −0.77, SD = 0.73). The main effect of the

measure demonstrated that, when collapsing the

data across the other independent variables, the

pre- to postintervention change scores varied

depending on which instrument was used to

record affective polarization. Post hoc Bonferroni-

corrected pairwise comparisons revealed that

change scores on the trust measure (M = −0.26,

SE = 0.76) were smaller than those on the feeling

thermometer (M = −4.29, SE = 0.81, p < .001)

and the net trait rating measure (M = −3.08, SE =

0.64, p = .003), while there was no difference

between the change scores on the feeling ther-

mometer and the net trait rating (p = .584). The

interaction between meditation practice condition

and measure was characterized by differences

between the befriending and control conditions on

the feeling thermometer change scores, F(1, 476)

= 13.64, p < .001, η

= .03, and the net trait rat-

ing change scores, F(1, 476) = 8.01, p = .005, η

= .02, but not on the trust change scores, F(1, 476)

= 0.48, p = .487, η

= .001. This suggests that

befriending meditation reduces affective polariza-

tion on the feeling thermometer and the net trait

rating but does not influence how much Democrats

and Republicans trust outgroup voters as well as

party candidates and elected officials relative to the

same ingroup targets. There was no main effect of

target (i.e., voters vs. party candidates and elected

officials), F(1, 476) = 0.93, p = .336, η

= .002,

or study number, F(1, 476) < 0.01, p = .980, η

.001, on affective polarization change scores. All

other two- and three-way interactions were not sig-

nificant (p values > .05). There was a marginal

four-way interaction between meditation practice

condition, measure, target, and study number, F(2,

475) = 3.02, p = .050, η

= .01.

Effects of befriending meditation on affective polarization

in Democrats and Republicans. In all previous analy-

ses, we combined data from participants who

affiliated with the Democratic Party or the

Republican Party. Here, we tested whether

befriending meditation impacts these two groups

to different extents. If so, we should observe an

interaction between meditation practice condi-

tion and participants’ political affiliation.

As the previous analysis showed that befriend-

ing meditation influenced affective polarization

on the feeling thermometer and the net trait rat-

ing but not on the trust measure, we did not

include the change scores on the trust measure in

this analysis. Entering the four change scores of

interest (pre- to postintervention changes in

affective polarization on the feeling thermometer

and the net trait rating when participants rated

voters as well as candidates and elected officials)

into a 2 (meditation practice condition: befriend-

ing, control) x 2 (political affiliation: Democrat,

Republican) x 2 (study number: 1, 2) MANOVA

revealed a main effect of meditation practice

condition, F(4, 469) = 4.69, p = .001, η

= .04,

no main effect of political affiliation, F(4, 469) =

1.06, p = .374, η

= .01, and no interaction

between meditation practice condition and politi-

cal affiliation, F(4, 469) = 0.13, p = .972, η

.001. This suggests that the effects of befriending

meditation on affective polarization did not vary

by political affiliation. There was also no main

effect of study number, F(4, 469) = 1.27, p =

.281, η

= .01, a marginal but not significant

interaction between meditation practice condi-

tion and study number, F(4, 469) = 2.04, p =

.088, η

= .02, and no three-way interaction,

F(4, 469) = 0.23, p = .919, η

= .002.

Discussion

Using pooled data from Study 1 and Study 2,

exploratory analyses found that the effects of a

brief befriending meditation on affective polariza-

tion did not differ between Democrats and

Republicans. The effects did, however, vary across

the measurements of affective polarization;

befriending meditation only reduced affective

Simonsson et al. 1589

polarization more than the control task on the

feeling thermometer and the net trait rating meas-

ure. The trust change scores did not vary between

conditions, which suggests that a brief befriend-

ing meditation does not necessarily decrease the

difference in trust toward the political outgroup

and ingroup. Befriending meditation does, on the

other hand, increase positive feelings toward, and

perceptions of, the political outgroup more than

the political ingroup.

General Discussion

The present research builds on previous literature

by testing whether meditation interventions can

reduce affective polarization in American adults

who affiliate with either the Democratic Party or

the Republican Party. Across two studies, we

found evidence that a brief befriending medita-

tion reduces affective polarization between

Democrats and Republicans by increasing positive

feelings and trait perceptions relatively more for

the political outgroup than the political ingroup.

While exploratory analyses found varying effects

of the befriending meditation across measure-

ments of affective polarization, the effects did not

differ between Democrats and Republicans.

Taken together, the findings in the present study

are the first to identify a causal relationship

between a brief befriending meditation and affec-

tive polarization.

While affective polarization was not signifi-

cantly associated with baseline levels of trait

compassion, a brief befriending meditation sig-

nificantly reduced affective polarization between

Democrats and Republicans. The measurement

of trait compassion evaluates compassion toward

others more generally and does not measure

intergroup feelings of compassion. It may there-

fore be less useful in understanding affective

polarization than a variable that is more specific

about intergroup feelings. For example, previous

research has shown that different types of medi-

tation can increase perceived commonality

between the self and others (Kok & Singer,

2017), which has recently also been shown to

mediate the effects of imagined intergroup

contact on affective polarization (Wojcieszak &

Warner, 2020). Future research should investigate

whether perceived commonality between the self

and the political outgroup, rather than general

feelings of compassion, might mediate the effects

of a brief befriending meditation on affective

polarization.

The results from the two studies are particu-

larly promising considering the increasing popu-

larity of meditation and the scalability of

meditation interventions. For example, between

2012 and 2017, the percentage of meditation

practitioners in the United States increased nine-

fold for children (0.6% to 5.4%) and threefold for

adults (4.1% to 14.2%; Black et al., 2018; Clarke

et al., 2018). Recent evidence also suggests that

meditation-based apps have become the most

common way of learning to meditate (Simonsson,

Fisher, & Martin, 2020), which demonstrates the

promise of meditation as a scalable intervention.

The present research project utilized a rigor-

ous research design to assess the effects of mind-

fulness and befriending meditation on affective

polarization, but there are several important

points to consider when interpreting the results.

First, the sample was not representative of the

American adult population, participants self-

selected into the study, and the study took place

during the coronavirus disease (COVID-19) out-

break. The same effects might therefore not be

present in a different sample or at a different time

point, which limits the generalizability of the

findings. Second, this research assessed the

immediate effects of 10-minute meditation inter-

ventions. As such, further empirical work is

needed to test whether changes in affective polar-

ization caused by befriending meditation are sus-

tainable over longer periods of time. We speculate

that longer, more regular practice would be

required for this intervention to generate deeper,

more lasting change. Third, both studies in this

paper utilized within-subject, repeated-measures

designs, which may introduce order effects and/

or increase demand characteristics. However, by

introducing a week-long gap between the pre-

intervention stage (in which we recorded baseline

measures) and the intervention stage (in which

1590 Group Processes & Intergroup Relations 25(6)

we applied the intervention and measured affec-

tive polarization once more), we reduced the pos-

sibility of demand characteristics that may have

resulted from participants providing responses

on the same measures twice. It is important to

note that participants may still have been able to

discern the link between the befriending medita-

tion instructions and the measures of affective

polarization. Future studies should therefore uti-

lize research designs that further minimize the

effect of demand characteristics (e.g., lagged

designs that consider the lasting effects of medi-

tation interventions).

In sum, the findings suggest that even a brief

online befriending meditation can reduce affec-

tive polarization between Democrats and

Republicans by increasing positive feelings and

trait perceptions relatively more for the outgroup

than the ingroup. The results contribute to the

existing scientific literature and provide support

for the benefits of befriending meditation in

political contexts, which opens the possibility of

a new subfield at the intersection of the contem-

plative and political sciences.

Author contributions

The first author conceptualized the study and developed

the hypotheses; The first and third authors designed and

preregistered the study; The third author analyzed the

data with assistance from the first author; the first and

third author wrote the manuscript, with inputs from the

second author; the third author supervised the study.

Ethical approval

All procedures performed involving human partici-

pants were in accordance with the ethical standards of

the 1964 Helsinki Declaration and its later amend-

ments or comparable ethical standards. The study was

approved by the Research Ethics Committee of the

Department of Sociology (DREC) at the University of

Oxford.

Funding

The author(s) disclosed receipt of the following finan-

cial support for the research, authorship, and/or publi-

cation of this article: This study was funded by

Kindness.org and the Singapore Ministry of Education

Academic Research Fund Tier 1 (T1 17/2019/115),

and by the National University of Singapore Research

Fund to J. N.

Informed consent

Informed consent was obtained from all individual

participants included in the studies.

ORCID iDs

Otto Simonsson https://orcid.org/0000-0003-4197

-7566

Joseph Marks

https://orcid.org/0000-0002-9644

-2967

Data accessibility statement

Research data and code supporting the results reported

in this paper are available on the Open Science

Framework (https://osf.io/u9m3c/).

Supplemental material

Supplemental material for this article is available online.

Note

1. Note that the attention checks in the preinterven-

tion stage were added after submission of the

preregistration.

References

Abramowitz, A. I., & Webster, S. (2016). The rise of

negative partisanship and the nationalization of

U.S. elections in the 21st century. Electoral Stud-

ies, 41, 12–22. https://doi.org/10.1016/j.elects-

tud.2015.11.001

Amodio, D. M., Jost, J. T., Master, S. L., & Yee, C.

M. (2007). Neurocognitive correlates of liberalism

and conservatism. Nature Neuroscience, 10, 1246–

1247. https://doi.org/10.1038/nn1979

Arechar, A. A., & Rand, D. G. (2020). Turking in the time

of COVID. PsyArXiv. https://doi.org/10.31234/

osf.io/vktqu

Bangdiwala, S. I., Bhargava, A., O’Connor, D. P., Rob-

inson, T. N., Michie, S., Murray, D. M., Stevens,

J., Belle, S. H., Templin, T. N., & Pratt, C. A.

(2016). Statistical methodologies to pool across

multiple intervention studies. Translational Behavio-

ral Medicine, 6, 228–235. https://doi.org/10.1007/

s13142-016-0386-8

Simonsson et al. 1591

Berger, R., Brenick, A., & Tarrasch, R. (2018). Reduc-

ing Israeli-Jewish pupils’ outgroup prejudice with

a mindfulness and compassion-based social-

emotional program. Mindfulness, 9, 1768–1779.

https://doi.org/10.1007/s12671-018-0919-y

Black, L. I., Barnes, P. M., Clarke, T. C., Stussman, B.

J., & Nahin, R. L. (2018). Use of yoga, meditation, and

chiropractors among US children aged 4–17 years. U.S.

Department of Health and Human Services, Cent-

ers for Disease Control and Prevention, National

Center for Health Statistics. https://www.cdc.gov/

nchs/products/databriefs/db324.htm

Bond, R. M., Shulman, H. C., & Gilbert, M. (2018).

Does having a political discussion help or hurt

intergroup perceptions? Drawing guidance

from social identity theory and the contact

hypothesis. International Journal of Communication,

12, 4332–4352.

Boxell, L., Gentzkow, M., & Shapiro, J. M. (2020). Cross-

country trends in affective polarization (NBER Working

Paper No. w26669). National Bureau of Economic

Research. https://doi.org/10.3386/w26669

Bristow, J. (2019). Mindfulness in politics and public

policy. Current Opinion in Psychology, 28, 87–91.

https://doi.org/10.1016/j.copsyc.2018.11.003

Carney, D. R., Jost, J. T., Gosling, S. D., & Potter, J.

(2008). The secret lives of liberals and conserva-

tives: Personality profiles, interaction styles, and

the things they leave behind. Political Psychology,

29, 807–840. https://doi.org/10.1111/j.1467-

9221.2008.00668.x

Clarke, T. C., Barnes, P. M., Black, L. I., Stussman, B.

J., & Nahin, R. L. (2018). Use of yoga, meditation,

and chiropractors among US adults aged 18 and over.

U.S. Department of Health and Human Services,

Centers for Disease Control and Prevention,

National Center for Health Statistics. https://

www.cdc.gov/nchs/products/databriefs/db325.

htm?mod=article_inline

Creswell, J. D. (2017). Mindfulness interventions.

Annual Review of Psychology, 68, 491–516. https://

doi.org/10.1146/annurev-psych-042716-051139

Crisp, R. J., & Turner, R. N. (2009). Can imagined

interactions produce positive perceptions?:

Reducing prejudice through simulated social con-

tact. American Psychologist, 64, 231–240. https://

doi.org/10.1037/a0014718

Donald, J. N., Sahdra, B. K., van Zanden, B., Duin-

eveld, J. J., Atkins, P. W., Marshall, S. L., & Ciar-

rochi, J. (2019). Does your mindfulness benefit

others? A systematic review and meta-analysis

of the link between mindfulness and prosocial

behaviour. British Journal of Psychology, 110, 101–

125. https://doi.org/10.1111/bjop.12338

Druckman, J. N., Klar, S., Krupnikov, Y., Levendusky,

M., & Ryan, J. B. (2020a). Affective polarization,

local contexts and public opinion in America.

Nature Human Behaviour. Advance online publi-

cation. https://doi.org/10.1038/s41562-020-

01012-5

Druckman, J. N., Klar, S., Krupnikov, Y., Leven-

dusky, M., & Ryan, J. B. (2020b). How affective

polarization shapes Americans’ political beliefs: A

study of response to the COVID-19 pandemic.

Journal of Experimental Political Science. https://doi.

org/10.1017/XPS.2020.28

Druckman, J. N., & Levendusky, M. S. (2019). What

do we measure when we measure affective polari-

zation? Public Opinion Quarterly, 83, 114–122.

https://doi.org/10.1093/poq/nfz003

Eisenberg, I. W., Bissett, P. G., Enkavi, A. Z., Li, J.,

MacKinnon, D. P., Marsch, L. A., & Poldrack, R.

A. (2019). Uncovering the structure of self-reg-

ulation through data-driven ontology discovery.

Nature Communications, 10, Article 2319. https://

doi.org/10.1038/s41467-019-10301-1

Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A.

(2007). G*Power 3: A flexible statistical power

analysis program for the social, behavioral, and

biomedical sciences. Behavior Research Methods, 39,

175–191. https://doi.org/10.3758/BF03193146

Feldman, G., Hayes, A., Kumar, S., Greeson, J., & Lau-

renceau, J. P. (2007). Mindfulness and emotion

regulation: The development and initial valida-

tion of the Cognitive and Affective Mindfulness

Scale-Revised (CAMS-R). Journal of Psychopathology

and Behavioral Assessment, 29, Article 177. https://

doi.org/10.1007/s10862-006-9035-8

Galante, J., Galante, I., Bekkers, M.-J., & Gallacher, J.

(2014). Effect of kindness-based meditation on

health and well-being: A systematic review and

meta-analysis. Journal of Consulting and Clinical Psy-

chology, 82, 1101–1114. https://doi.org/10.1037/

a0037249

Graham, J., Haidt, J., & Nosek, B. A. (2009). Liberals

and conservatives rely on different sets of moral

foundations. Journal of Personality and Social Psychol-

ogy, 96, 1029–1046. https://doi.org/10.1037/

a0015141

Hetherington, M. J., & Rudolph, T. J. (2015). Why

Washington won’t work: Polarization, political trust, and

the governing crisis. University of Chicago Press.

Hutcherson, C. A., Seppala, E. M., & Gross, J. J. (2008).

Loving-kindness meditation increases social con-

1592 Group Processes & Intergroup Relations 25(6)

nectedness. Emotion, 8, 720–724. https://doi.

org/10.1037/a0013237

Hwang, J. Y., Plante, T., & Lackey, K. (2008). The

development of the Santa Clara Brief Compas-

sion Scale: An abbreviation of Sprecher and

Fehr’s Compassionate Love Scale. Pastoral Psy-

chology, 56, 421–428. https://doi.org/10.1007/

s11089-008-0117-2

Iyengar, S., Lelkes, Y., Levendusky, M., Malhotra,

N., & Westwood, S. J. (2019). The origins and

consequences of affective polarization in the

United States. Annual Review of Political Science,

22, 129–146. https://doi.org/10.1146/annurev-

polisci-051117-073034

Iyengar, S., Sood, G., & Lelkes, Y. (2012). Affect,

not ideology: A social identity perspective on

polarization. Public Opinion Quarterly, 76, 405–431.

https://doi.org/10.1093/poq/nfs038

Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway,

F. J. (2003). Political conservatism as motivated

social cognition. Psychological Bulletin, 129, 339–375.

https://doi.org/10.1037/0033-2909.129.3.339

Kabat-Zinn, J. (2003). Mindfulness-based interven-

tions in context: Past, present, and future. Clinical

Psychology: Science and Practice, 10, 144–156. https://

doi.org/10.1093/clipsy.bpg016

Kang, Y., Gray, J. R., & Dovidio, J. F. (2014). The non-

discriminating heart: Lovingkindness meditation

training decreases implicit intergroup bias. Journal

of Experimental Psychology: General, 143, 1306–1313.

https://doi.org/10.1037/a0034150

Klein, E. (2020). Why we’re polarized. Simon & Schuster.

Klimecki, O. M. (2019). The role of empathy and compas-

sion in conflict resolution. Emotion Review, 11, 310–

325. https://doi.org/10.1177/1754073919838609

Kok, B. E., & Singer, T. (2017). Effects of contem-

plative dyads on engagement and perceived social

connectedness over 9 months of mental training:

A randomized clinical trial. JAMA Psychiatry, 74,

126–134. https://doi.org/10.1001/jamapsychia-

try.2016.3360

Levendusky, M. S. (2018). Americans, not partisans:

Can priming American national identity reduce

affective polarization? The Journal of Politics, 80,

59–70. https://doi.org/10.1086/693987

Levine, M., Prosser, A., Evans, D., & Reicher, S. (2005).

Identity and emergency intervention: How social

group membership and inclusiveness of group

boundaries shape helping behavior. Personality and

Social Psychology Bulletin, 31, 443–453. https://doi.

org/10.1177/0146167204271651

McConnell, C., Margalit, Y., Malhotra, N., & Leven-

dusky, M. (2018). The economic consequences

of partisanship in a polarized era. American

Journal of Political Science, 62, 5–18. https://doi.

org/10.1111/ajps.12330

Miles, E., & Crisp, R. J. (2014). A meta-analytic test

of the imagined contact hypothesis. Group Pro-

cesses & Intergroup Relations, 17, 3–26. https://doi.

org/10.1177/1368430213510573

Paolacci, G., & Chandler, J. (2014). Inside the Turk:

Understanding Mechanical Turk as a participant

pool. Current Directions in Psychological Science, 23, 184–

188. https://doi.org/10.1177/0963721414531598

Petersen, M. B., & Mitkidis, P. (2019). A sober second

thought? A pre-registered experiment on the effects of

mindfulness meditation on political tolerance. PsyArXiv.

https://psyarxiv.com/ksy37/

Pettigrew, T. F. (2009). Secondary transfer effect of

contact: Do intergroup contact effects spread to

noncontacted outgroups? Social Psychology, 40, 55–

65. https://doi.org/10.1027/1864-9335.40.2.55

Price-Blackshear, M. A., Kamble, S. V., Mudhol, V.,

Sheldon, K. M., & Bettencourt, B. A. (2017).

Mindfulness practices moderate the association

between intergroup anxiety and outgroup atti-

tudes. Mindfulness, 8, 1172–1183. https://doi.

org/10.1007/s12671-017-0689-y

Rushton, J. P., Brainerd, C. J., & Pressley, M. (1983).

Behavioral development and construct validity: The

principle of aggregation. Psychological Bulletin, 94, 18–

38. https://doi.org/10.1037/0033-2909.94.1.18

Simas, E. N., Clifford, S., & Kirkland, J. H. (2020).

How empathic concern fuels political polariza-

tion. American Political Science Review, 114, 258–269.

https://doi.org/10.1017/S0003055419000534

Simonsson, O., Fisher, S., & Martin, M. (2020).

Awareness and experience of mindfulness in

Britain. Sociological Research Online. https://doi.

org/1360780420980761

Simonsson, O., Martin, M., & Fisher, S. (2020). Soci-

odemographic characteristics and health status of

mindfulness users in the United States. Mindful-

ness, 11, 2725–2729. https://doi.org/10.1007/

s12671-020-01486-4

Van Laar, C., Levin, S., Sinclair, S., & Sidanius, J.

(2005). The effect of university roommate contact

on ethnic attitudes and behavior. Journal of Experi-

mental Social Psychology, 41, 329–345. https://doi.

org/10.1016/j.jesp.2004.08.002

Warner, B. R., Horstman, H. K., & Kearney, C. C.

(2020). Reducing political polarization through

Simonsson et al. 1593

narrative writing. Journal of Applied Communication

Research, 48, 459–477. https://doi.org/10.1080/0

0909882.2020.1789195

Warner, B. R., & Villamil, A. (2017). A test of imagined

contact as a means to improve cross-partisan feel-

ings and reduce attribution of malevolence and

acceptance of political violence. Communication

Monographs, 84, 447–465. https://doi.org/10.108

0/03637751.2017.1336779

Williams, M., & Penman, D. (2011). Mindfulness: A prac-

tical guide to finding peace in a frantic world. Piatkus.

Wojcieszak, M., & Garrett, R. K. (2018). Social iden-

tity, selective exposure, and affective polarization:

How priming national identity shapes attitudes

toward immigrants via news selection. Human

Communication Research, 44, 247–273. https://doi.

org/10.1093/hcr/hqx010

Wojcieszak, M., & Warner, B. R. (2020). Can interparty

contact reduce affective polarization? A system-

atic test of different forms of intergroup contact.

Political Communication, 37, 789–811. https://doi.

org/10.1080/10584609.2020.1760406