University of Southern California Dissertations and Theses

Socio-ecological psychology of moral values

(USC Thesis Other)

Socio-ecological psychology of moral values

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content SOCIO-ECOLOGICAL PSYCHOLOGY OF MORAL VALUES
by
Mohammad Atari
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PSYCHOLOGY )
December 2021
Copyright 2021 Mohammad Atari

ii

Table of Contents
Dedication……………………………………………………………………………………...…iii
Acknowledgments………...…………..…………...………………………………………………v
List of Tables………………………...…………..…….………………………………………...vii
List of Figures………………………….………………………………………………………….x
Abstract………………………………...…...…….……………………………………………...xii
Chapter 1…………………………………………………………………………….…………….1
Chapter 2……………………………….………………………………………………………….4
Chapter 3………………………………......……………………………………….…………….39
Chapter 4……………………………………..…………………………………….…………….95
Appendix……………………………….……………………………………………………….177
Supplementary Materials………..……………….…..….………..…………………………….180
References………………………………………….…………………………………..……….208

iii

Dedication
Family is everything.
To my everything.

iv

Acknowledgements
I stand, and have been standing, on the shoulders of many giants. It feels impossible to do them
all justice in this short Acknowledgments section. First and foremost, I thank my family for
being there, always. My mom, Mehri, never stopped believing in me. Although she has been
calling me “doktor” since I was 18 years old, she can now more confidently call me that, after I
submit this dissertation. My dad, Nemat, made me a stronger person and taught me to be humble.
Thank you for believing in me. I appreciate all the sacrifices my parents have made for me. My
older sister, Maryam, has been an absolute source of inspiration, teaching me to pursue my
dreams since I was a kid. My younger sister, Matin, has always been the person I go to when I
am feeling down; she has an infinite amount of compassion and the biggest heart. I thank her for
believing in me and for supporting me since the day I was born (literally). My nephew, Amir
Mohammad, is and has been the star of my life, I am proud of him, I thank him for being in my
life, and for pushing me to learn how to be better at playing FIFA on XBOX.
I want to give a special shout-out to my lovely friends. Aida is truly such a wonderful
human being, in addition to being one of the most intelligent people I have ever met (someday, I
will beat you in some sort of board game, inshallah). Her partner/husband, Mohamad, has been
there, time and time again. I would like to take some credit for all the great pizzas we baked
together through the pandemic (but really, he did 95% of the work, I just chopped mushrooms).
It’s been a privilege to be friends with both of you. Brendan is such an amazing person, and a
true friend in hardships. Also, thank you for acknowledging and validating my love for hyphens
(and parentheses). I look forward to co-writing with you for many years to come. I appreciate
him, and his wife (and my friend), Dela. I am immensely thankful to my friend/brother, Sahand,
for his wit, warmth, and support through (many) years. He is one of the most creative people in

v

my life, and a superstar artist. Finally, I am hugely thankful to Farzan for his caring about me, for
being my biggest fan, and for encouraging me to keep going when things got rough. I can’t thank
you enough. As I said in the “Dedication” section, family is everything, and these lovely people
are indeed family.
I am quite-very-extremely grateful to Morteza for being the best adviser I could have
hoped for. To date, your reply to my first inquiry in 2016, “Dear Mr. Atari, Thank you for your
email. You have an impressive CV. We should chat. I have time on Wed at 10am, 11 and 11:30
(pacific standard time)”, is the best and most life-changing email I ever received. Right after our
first interview, I was confident that I had found the best adviser in the world, and I was right
(although I was anxious until I received the offer because I was sure you had more competent
applicants). Morteza trusted me, when many chose not to. Joining the Computational Social
Science Lab was the best decision I have made in my life. He taught me that the greatest
characteristic of a scientist is not publishing in big-name journals but being intellectually
humble. Morteza has been more than an academic mentor; he is a life mentor for me, and I look
forward to learning from him for the rest of my life.
I am thankful to my dissertation committee members. Daphna has been a truly inspiring
figure in my academic life. Through her mentorship, I learned how to think like an experimental
social psychologist, and I also learned why it is crucial to think carefully about culture no matter
what psychological phenomena I want to investigate. She enthusiastically provided me feedback
on whatever ideas I had, supported me in developing ideas, conducting research, writing, and
presenting findings to a broad audience. I am inordinately thankful to her, as well as my lab-
mates in the Identity-Based Motivation Lab. I thank Mark for being absolutely supportive and
kind. He patiently helped me think about the best modern statistical model to use in addressing

vi

my research questions. I want to thank Haj for her very useful feedback in my dissertation, and
for wonderful conversations in the last few years. Finally, I want to give very special thanks to
Jon. When I was reading your books and papers early in grad school, never would I have thought
that I could have you on my dissertation committee. Thank you for allowing me (and
encouraging me) to criticize your theory, and thank you for being such a compassionate,
supportive, genius person who lifts junior researchers up every single day. I look forward to
being inspired by you in many years to come. Working with my heroes has given me so much
energy and enthusiasm, and I hope I can someday pay the favor forward to the next generation of
social psychologists.

vii

List of Tables
Table 2.1…………………………………………………………………...…………………….14
Table 2.2.……………………………………………………………………..………………….22
Table 2.3…………………………………………………………………………………………34
Table 3.1…………………………………………………………………………………………41
Table 3.2…………………………………………………………………………………………54
Table 3.3…………………………………………………………………………………………56
Table 3.4…………………………………………………………………………………………57
Table 3.5…………………………………………………………………………………………62
Table 3.6…………………………………………………………………………………………71
Table 3.7…………………………………………………………………………………………77
Table 3.8…………………………………………………………………………………………79
Table 3.9…………………………………………………………………………………………81
Table 3.10………………………………………………………….………….…………………86
Table 3.11………………………………………………………….………….…………………87
Table 4.1…………………………………………………………….………………………….116
Table 4.2…………………………………………………………….………………………….117
Table 4.3……………………………………………………………………………..…………126

viii

Table 4.4……………………………………………………………………………..…………129
Table 4.5…………………………………………………………………………..……………133
Table 4.6………………………………………………………………………..………………135
Table 4.7……………………………………………………………………..…………………137
Table 4.8…………………………………………………………………..……………………154
Table 4.9………………………………………………………………..………………………156
Table S.1……………………………………………………………..…………………………180
Table S.2…………………………………………………………..……………………………182
Table S.3………………………………………………………………………………………..184
Table S.4……………………………………………………………………………………..…186
Table S.5…………………………………………………………………………………..……188
Table S.6………………………………………………………………………………….…….190
Table S.7…………………………………………………………………………………….….192
Table S.8…………………………………………………………………………………..……194
Table S.9………………………………………………………………………………..………196
Table S.10………………………………………………………………………………………198
Table S.11………………………………………………………………………………………200
Table S.12………………………………………………………………………………………202

ix

Table S.13………………………………………………………………………………………204
Table S.14………………………………………………………………………………………206

x

List of Figures
Figure 2.1……………………………………………………………………….……….…….…13
Figure 2.2.……………………………………………………………………….…….…………19
Figure 2.3……………………………………………………………………………...…………21
Figure 2.4…………………………………………………………………………….…..………33
Figure 3.1…………………………………………………………………………….………..…62
Figure 3.2…………………………………………………………………………...……………64
Figure 3.3……………………………………………………………………………...…………65
Figure 3.4………………………………………………………………………………...………67
Figure 3.5…………………………………………………………………………………...……73
Figure 4.1…………………………………………………………………………………….…136
Figure 4.2…………………………………………………………………………………….…139
Figure 4.3………………………………………………………………………………….……142
Figure 4.4…………………………………………………………………………………….…144
Figure 4.5………………………………………………………………………………….……147
Figure 4.6………………………………………………………………………………….……159
Figure 4.7………………………………………………………………………………….……160
Figure S.1………………………………………………………………………………….……181

xi

Figure S.2………………………………………………………………………………….……183
Figure S.3………………………………………………………………………………….……185
Figure S.4………………………………………………………………………………….……187
Figure S.5………………………………………………………………………………….……189
Figure S.6………………………………………………………………………………….……191
Figure S.7………………………………………………………………………………….……193
Figure S.8………………………………………………………………………………….……195
Figure S.9………………………………………………………………………………….……197
Figure S.10……………………………………………………………………………….…..…199
Figure S.11……………………………………………………………………………….…..…201
Figure S.12…………………………………………………………………………………...…203
Figure S.13………………………………………………………………………….………..…205
Figure S.14…………………………………………………………………………………...…207

xii

Abstract
In my dissertation, I address three interconnected questions to better understand how social and
ecological factors interact to give rise to moral values across cultures. First, I provide an
Introduction to these questions (Chapter 1). Then, I examine how pathogen prevalence, as an
important socio-ecological factor, is related to different moral values across cultures, regions,
and individuals (Chapter 2). In Chapter 3, I focus on one particular culture, Iran, to develop a
culture-based model of morality in this understudied population. Then, based on lessons learned
from these two studies and the existing literature, I extend a psychometrically valid and reliable
measure of moral values, Moral Foundations Questionnaire-2 (MFQ-2) across cultures in
Chapter 4. Throughout this dissertation, I rely on Moral Foundations Theory to operationalize
moral values, using a pluralistic view on our judgments of right and wrong.

1

Chapter 1: Introduction
Do our judgments between right and wrong, or good and bad behavior, depend on
culture? The study of morality in psychological science has been, in large part, limited to
Western, Educated, Industrialized, Rich, and Democratic (WEIRD; Henrich, Heine, &
Norenzayan, 2010) populations (see Skitka & Conway, 2019). Most existing studies examining
moral values in non-WEIRD populations have typically failed to develop cultural models of
moral cognition or to incorporate socio-ecological factors in understanding the variations in
moral values and norms from one population to the next. In my dissertation, I intend to focus on
three interconnected questions to better understand how social and ecological factors interact to
give rise to moral values across cultures. First, in Chapter 2, I examine how pathogen prevalence,
as an important socio-ecological factor, is related to different moral values across cultures,
regions, and individuals. Second, in Chapter 3, I focus on one particular culture, Iran, to develop
a culture-based model of morality in this understudied population. Third, based on lessons
learned from these two studies and the existing literature, I extend a psychometrically valid and
reliable measure of moral values across cultures in Chapter 4. Throughout this dissertation, I rely
on Moral Foundations Theory (MFT; Haidt & Joseph, 2004; Graham et al., 2013) to
operationalize moral values, using a pluralistic view on our judgments of right and wrong.
Chapter 2 extends an evolutionarily informed view of the interaction of evolved
psychological mechanisms and moral concerns. Indeed, infectious diseases have been an
impending threat to the survival of individuals and groups throughout our evolutionary history.
As a result, humans have developed psychological pathogen-avoidance mechanisms and groups
have developed societal norms that respond to the presence of disease-causing microorganisms
in the environment. In Chapter 2, I demonstrate that morality plays a central role in the

2

psychological architecture evolved to avoid pathogens. I present a collection of studies which
together provide an integrated understanding of the socio-ecological and psychological impacts
of pathogens on human morality. Specifically, I argue that particular adaptive moral systems are
developed and maintained in response to the threat of chronic and acute pathogen occurrence in
the environment. We draw on computational linguistics, social psychology, and evolutionary
anthropology to establish connections between pathogens and moral codes in multiple languages,
experimentally induced situations, individual differences, U.S. counties, 67 nations, and the
historical periods over the last century.
Chapter 3 zooms in on one of the least studied cultures in psychological science: Iran. As
mentioned above, most moral psychology research has been conducted in WEIRD societies. As
such, moral judgment, as a psychological phenomenon, might be known to researchers only by
its WEIRD manifestations. Here, I started with evaluating MFT using the Moral Foundations
Questionnaire (MFQ) and follow up by building a bottom-up model of moral values in Iran, a
non-WEIRD, Muslim-majority, cultural setting. In six studies, I examine the structural validity
of the Persian translation of the MFQ, compare moral foundations between Iran and the U.S.,
conduct qualitative interviews regarding moral values, expand the nomological network of
“Qeirat” as a culture-specific set of moral values, and investigate the pragmatic validity of
“Qeirat” in Iranian culture. The findings suggest an additional moral foundation in Iran, above
and beyond the five foundations identified by MFT. Specifically, qualitative studies highlighted
the role of “Qeirat” values in Iranian culture, which are comprised of guarding and
protectiveness of female kin, romantic partners, broader family, and country. Significant cultural
differences in moral values are argued in this work to follow from the psychological systems
that, when brought to interact with particular socio-ecological environments, produce different

3

moral structures. This evolutionarily informed, cross-cultural, mixed-methods research sheds
light on moral concerns and their cultural, demographic, and individual-difference correlates in
Iran.
Chapter 3 advances MFT theoretically and develops the next Moral Foundations
Questionnaire (MFQ-2). I revisit MFT and psychometrically develop MFQ-2 in 25 cultures and
7 languages. In Chapter 3, I propose to broaden the moral domain to six foundations: Care,
Equality, Proportionality, Loyalty, Authority, and Purity. Three studies were conducted to
develop and validate the MFQ-2, a major revision of the MFQ. Study 1 (five cultures) specified a
refined top-down approach for measurement of moral foundations, and developed a large item
pool, progressively reducing it to contain 50 items. Study 2 (19 cultures) used a variety of
methods to provide evidence that MFQ-2 fares well in terms of reliability and validity across
cultural contexts. We also examined cultural, religious, ideological, and sex differences using the
new measure. Study 3 (three cultures) provided evidence for convergent validity of MFQ-2
scores, expanded the nomological network of moral foundations, and demonstrated the
predictive power of the measure compared with the original MFQ. Collectively, these studies
demonstrate that MFQ-2 has desirable psychometric properties, while still retaining the original
measure’s conceptual focus, brevity, and ease of understanding. These studies provide evidence
for the usefulness of MFQ-2 in simultaneously increasing the scope and sharpening the
theoretical resolution of MFT.

4

Chapter 2: Pathogens Are Linked to Human Moral Systems across Time and Space
Humans currently suffer, and have suffered over their evolutionary history, from
infectious diseases. Some of these infectious agents have found their way into human
populations relatively recently (e.g., human immunodeficiency virus), but many others are of
considerable antiquity, suggesting that disease-causing pathogens have posed debilitating threats
to human populations through their evolutionary past (Schaller & Murray, 2011). The presence
of strong biological anti-pathogenic mechanisms in humans, as well as other primate
populations, is a testament to the power of pathogens and their role in our evolution. The most
well-known mechanism to mitigate pathogenic threats is the body’s own immune system, whose
overall function is to prevent or limit infections after the disease-causing agents have breached
the body.
Although the human immune system is vital, its activation is not without costs. For
example, activation of the immune system in response to detection of pathogenic microbes is
metabolically taxing, robbing individuals of caloric resources that can otherwise be devoted to
other evolutionarily important tasks such as searching for desirable mates, caring for one’s kin,
and building social ties (Schmid-Hempel, 2003). Thus, it is advantageous for the body to prevent
its immune system from being chronically burdened by fighting pathogenic microbes.
Immunological defenses in response to pathogens occur only after pathogenic agents have
compromised the body’s immune surveillance system (i.e., reactive defense). Because of the
costs associated with generating an adaptive immune response (versus innate), there would have
been evolutionarily unique benefits associated with proactive defense: the behavioral prevention
of infection by avoiding infectious diseases (Schaller, 2006; Schaller & Park, 2011).

5

The behavioral immune system consists of a suite of psychological mechanisms that
detect cues of infectious pathogens in the environment, trigger disease-relevant psychological
responses, and facilitate behavioral avoidance of pathogens (Schaller & Park, 2011). If the so-
called behavioral immune system evolved to facilitate specific behavioral responses (e.g.,
avoidance), it may be considered to be a psychologically unique motivational system (Neuberg,
Kenrick, & Schaller, 2011). Distinct motivational systems are typically associated with distinct
affective experiences (Damasio, 2019; Damasio & Carvalho, 2013). The affective experience
associated with the behavioral immune system has been found to be disgust. There is a
substantial body of research implicating the specific role of disgust as a key component of the
behavioral immune system (Curtis, De Barra, & Aunger, 2011). The functional relationship
between pathogenic agents and disgust manifests in many ways. Specifically, sensory cues that
clearly connote proximity to infectious pathogens are especially likely to evoke disgust.
Furthermore, behaviors that violate normative expectations in social domains that are associated
with disease transmission (e.g., food preparation, physical proximity, personal hygiene, sexual
interaction) can also elicit disgust (Rozin, Haidt, & McCauley, 2008).
Disgust shares some characteristics with other negative affective experiences (e.g., fear,
anger, hate), but has a unique influence on information processing systems, which contribute to
keeping individuals away from communicable diseases. Disgust can be elicited by objects, such
as open wounds, unclean spaces, and dirt, but also by specific forms of social behavior. These
behaviors can include unusual sexual acts, eating unusual food, and actions that violate moral
codes of conduct (Haidt, McCauley, & Rozin, 1994; Tybur, Lieberman, Kurzban, & DeScioli,
2013; Wagemans, Brandt, & Zeelenberg, 2018). Consistently, some religions attach moral
impurity to particular sexual acts (e.g., extramarital relationships), certain food choices (e.g.,

6

eating pork), and touching unclean objects such as urine or feces. Moral psychologists have
argued that individuals who are more prone to experiencing pathogen-induced disgust may
condemn moral violations more strongly, and may be more likely to value practices associated
with moral purity such as temperance, chastity, piety, and cleanliness (Haidt & Joseph, 2004).
In addition to the mentioned individual-level effects of pathogens on cognitive,
emotional, and behavioral tendencies, infectious diseases have been found to shape important
society-level outcomes. Pathogen prevalence, as an important ecological factor, can influence
cultural and moral systems at the community level (e.g., Bastian et al., 2019; Jackson, Gelfand,
De, & Fox, 2019). For example, cross-cultural research suggests that in geographical regions
characterized by historically higher pathogen prevalence, people are less individualistic, exhibit
lower levels of openness to new things, and more strongly endorse values that emphasize
traditionalism, group loyalty, obedience, and respect for authority (Murray, Schaller, & Suedfeld,
2013; Tybur et al., 2016; Van Leeuwen, Park, Koenig, & Graham, 2012). In addition, higher
historical pathogen prevalence has been linked to higher society-level conformity pressures (i.e.,
cultural tightness), higher political conservatism, and linguistic heterogeneity (Schaller &
Murray, 2011). There is accumulated evidence suggesting that societies adopt varying rituals and
moral values to facilitate coping with chronic prevalence of pathogens (Fincher & Thornhill,
2012), or large-scale temporary increases in the occurrence of a disease such as the 1918
influenza pandemic (Alexander, 2019). Historians of medicine have proposed, using qualitative
data, that epidemics are typically followed by changes in social and moral norms, for example
increased religiosity, presumably to attenuate the effects of high rates of communicable disease
in the environment (Snowden, 2019).

7

In this work, drawing principally on psychology, anthropology, and behavioral ecology,
we propose that at the individual level, pathogenic cues are related to moral judgments across
individuals and experimental situations. At the society level, we predict that pathogen prevalence
is linked to human moral systems across regions, cultures, languages, and historical periods.
Specifically, we rely on Moral Foundations Theory (MFT; Graham et al., 2013; Haidt & Joseph,
2004) to operationalize human moral systems from a cultural and evolutionary perspective. MFT
aims to achieve a parsimonious basis for explaining the links between anthropological and
evolutionary accounts of moral intuitions. This framework suggests that moral intuitions derive
from innate psychological mechanisms that coevolved with cultural institutions. Care, Fairness,
Loyalty, Authority, and Purity — according to MFT (Graham et al., 2013) — are moral systems
that possess evolutionarily adaptive underpinnings present in individuals across cultures, with
each moral system producing automatic “gut-level” reactions of like or dislike when certain
phenomena are perceived in the social world, which in turn guide moral judgments of right and
wrong. Moral foundations are phenotypically plastic, i.e., although these five foundations have
adaptive fitness benefits and are present across human populations, they are environmentally
sensitive such that each foundation can be increased or decreased based on the ecologically
recurrent problems they can solve. In addition, moral foundations are context-dependent, i.e.,
although each individual values these foundations to some extent, contextual cues can subtly
change people’s gut-level judgments in these domains. For example, while conservatives
typically value Care and Fairness less than liberals (Graham et al., 2013), their score increases on
these foundations when analytic thought is activated using immediate contextual cues (Yilmaz &
Saribay, 2017).

8

Care is rooted in the instinct to protect and nurture offspring and weak individuals in
one’s group. Fairness is oriented toward concerns of justice, proportionality, equality, and
reciprocity. Care and Fairness are collectively referred to as “individualizing” foundations as
they are concerned with the rights of individuals. Loyalty is geared toward concerns of unity,
solidarity, togetherness, family, and tribe. Authority values function to defend authority and
social order within a hierarchical structure. Finally, Purity corresponds to physical and spiritual
cleanliness, decency, and dignity, valuing sacredness, and suppressing carnal desires. Loyalty,
Authority, and Purity are collectively referred to as “binding” foundations as they support and
enable group ties while discouraging selfish behavior for the good of the group.
Here, our overarching theory is that pathogen threat and prevalence are linked to
individual-level and population-level endorsements of moral foundations, particularly Purity.
Specifically, we argue that Purity is a particular adaptive moral system developed and
maintained in response to the threat of chronic and acute pathogen occurrence in the
environment. We expect that the connection between pathogens and moral Purity should be
manifest in contextual judgments and individual differences at individual-level. Since Purity
concerns are related to physical proximity-seeking behavior within groups, less unrestricted
sexual activity with outgroup members, and higher cleanliness in interpersonal interactions, we
hypothesize higher endorsements of Purity values in societies that suffer (or have historically
suffered) from high levels of pathogen prevalence.
In Studies 2.1 and 2.2, we use epidemiological data to examine how pathogen prevalence
is linked to moral systems across geographical regions. We hypothesize that regions with higher
prevalence of infectious diseases employ community-level moral concerns, particularly Purity, to
cope with (and hopefully reduce) the presence of communicable diseases. For example, an

9

outbreak of an infectious disease in a region can cause residents to lessen social contact, exclude
unknown outgroup members (especially those who are perceived to be from the same origin as
the disease) through fear of contagion (Kim, Sherman, & Updegraff, 2016), care more for
immunocompromised persons, and reduce unsafe sexual activity such as short-term sexual
encounters. Complementing Studies 2.1 and 2.2, in Study 2.3 we test the hypothesis that if
pathogens are accompanied by a particular type of moral system, they should also be
linguistically concomitant. In Studies 2.4 and 2.5, we investigate our individual-level hypotheses.
Specifically, we test whether individuals who generally are more avoiding of pathogenic stimuli
value particular moral foundations, and whether pathogenic stimuli are able to activate pathogen-
neutralizing moral systems related to physical distancing with unknown individuals, unsafe
sexual activities, and cleanliness. Lastly, in Study 2.6, we explore the relationship between rates
of infectious diseases and moral language as manifested in the 20
th
century books.
Study 2.1
In this study, we examine the covariation of moral values and pathogen prevalence in
geographical regions. The prevalence of infectious diseases specific to each socio-ecology
influences moral values in that region. Thus, it is crucial to examine how geographic variation in
pathogen prevalence is associated with regional differences in moral values. To test this
prediction, in Study 2.1 we investigate the relationship between pathogen prevalence at county
level in the U.S. and moral values held in those counties, controlling for county-level political
ideology and population.

10

Methods
Pathogen Prevalence. We use county-level pathogen estimates based on infectious
diseases data made available by the Centers for Disease Control and Prevention (CDC). The
CDC’s website (cdc.gov) has information on various infectious (and noninfectious) diseases in
the U.S. and provides county-level occurrences of pathogen-related mortality. We used archival
data
1
from 1999 to 2016 and compiled a data set with county-level mortality rate (per 100,000
residents) due to infectious diseases in 2,834 unique counties (M = 23.8, SD = 9.0) across 50
states and District of Columbia (DC). Scaled county-level estimates of pathogen mortality were
calculated. The population of each county was also acquired from the CDC.
Moral Values. We used the estimates of county-level distribution of moral values
provided by Hoover et al. (2019). These authors relied on data from www.YourMorals.org,
which is an online platform for collecting data on moral values using the Moral Foundations
Questionnaire (MFQ; Graham et al., 2011). MFQ is a self-report measure of moral foundations
with 30 items measuring the five moral foundations. The data collection process elapsed from
2012 to 2018 (N = 106,465). Although this is a relatively large sample, it cannot be used to
directly estimate county-level moral values as it does not use probability sampling at the county
level (Hoover & Dehghani, 2020), and many counties have very few respondents. To address
these estimation issues, Hoover et al. (2019) used Multilevel Regression and Synthetic
Poststratification (MrsP; Leemann & Wasserfallen, 2017), a model-based approach for sub-
national estimation that extends Multilevel Regression and Poststratification (MrP; Park,
Gelman, & Bafumi, 2004). Specifically, Hoover et al. (2019) used county-level estimates based
on the inclusion of a more diverse set of demographic variables. These estimates also account for
two levels of regional clustering, the county level and the region level, and include the

11

proportion of Democratic votes in the 2016 presidential election as a county-level factor. Finally,
the multilevel model also includes a hierarchical auto-regressive prior (Riebler, Sørbye,
Simpson, & Rue, 2016) that, under the presence of spatial auto-correlation, induces local spatial
smoothing between proximate counties (Hanretty, Lauderdale, & Vivyan, 2016; Hoover &
Dehghani, 2020; Selb & Munzert, 2011). We used estimates of Care, Fairness, Loyalty,
Authority, and Purity for each county. County-level moral values estimated by Hoover et al.
(2019) can be viewed at mapyourmorals.usc.edu
Political Ideology. Since political ideology is related to both pathogen avoidance (Tybur
et al., 2016) and moral values (Graham, Haidt, & Nosek, 2009), we controlled for the role of
political ideology in our statistical models. We collected data from presidential elections in the
U.S. (MIT Election Data and Science Lab, 2018) and calculated a county-level conservatism
index by subtracting the voters for the Democratic party from the Republican party and dividing
that by the total voters (including for the Green party) in that county. We then averaged these
county-level conservatism estimates for the same periods as with our pathogen data (2000, 2004,
2008, 2012, and 2016 elections). Our temporally-aggregated county-level conservatism estimates
were found to be reliable across 2000-2016 presidential elections (Cronbach’s α = .97, 95%CI
[.96, .97]).
Analytic Procedure. We ran five hierarchical linear models with random intercepts to
predict each moral concern by county-level mortality due to infectious diseases grouped in states,
while controlling for county-level political conservatism and population (in millions). We
excluded counties for which we did not have complete data and standardized mortality rate
before analysis. The final sample included 2,815 counties (out of 3,007 U.S. counties) clustered

12

in 49 states. Restricted maximum likelihood (REML) estimation was used in our hierarchical
models, accounting for the fact that fixed components are estimated when estimating variances.
Results
We found a positive fixed effect of pathogen prevalence in predicting Loyalty (B = 0.017,
SE = 0.001, p < .001), Authority (B = 0.036, SE = 0.001, p < .001), and Purity (B = 0.058, SE =
0.002, p < .001) controlling for political ideology and population. All models are presented in
Table 2.1. County-level pathogen prevalence did not significantly predict Care (B = −0.001, SE
= 0.001, p = .109) or Fairness (B = −0.0002, SE = 0.001, p = .731). The strongest effect of
pathogen mortality was observed for Purity. County-level purity values and pathogen mortality
rates are shown in Figure 2.1. Hence, we found evidence that binding values, especially Purity,
are related to contemporary rates of infectious diseases across U.S. counties after adjusting for
how conservative and populous these counties are.

13

Figure 2.1
Cross-county relationships between infectious disease mortality and moral values.

14

Table 2.1

Five multilevel models predicting moral values at the county level

Care Fairness Loyalty Authority Purity
(1) (2) (3) (4) (5)

Crude Rate (per 100,000)
−
0.001 0.0002 0.017
∗∗∗

0.036
∗∗∗

0.058
∗∗∗

−
(0.001) (0.001) (0.001) (0.001) (0.002)
Population (million) 0.0005
∗∗∗

0.001
∗∗∗

−
0.001
∗∗∗

−
0.001
∗∗∗

−
0.002
∗∗∗

(0.0001) (0.0001) (0.0001) (0.0002) (0.0003)
County-level Conservatism
−
0.089
∗∗∗

0.225
∗∗∗

0.252
∗∗∗

0.238
∗∗∗

0.509
∗∗∗

−
(0.002) (0.002) (0.004) (0.005) (0.008)
Constant 3.478
∗∗∗

3.408
∗∗∗

2.836
∗∗∗

3.059
∗∗∗

2.757
∗∗∗

(0.004) (0.005) (0.007) (0.014) (0.019)

Observations 2,815 2,815 2,815 2,815 2,815
Log Likelihood 6,179.247 6,204.850 4,979.899 4,119.078 2,917.189
Akaike Inf. Crit. -12,346.490 -12,397.700 -9,947.798 -8,226.157 -5,822.378
Bayesian Inf. Crit. -12,310.840 -12,362.040 -9,912.142 -8,190.500 -5,786.721

Note.
∗
p<0.05;
∗∗
p<0.01;
∗∗∗
p<0.001

15

Study 2.2
Pathogen prevalence can have regional influences on the development and persistence of
norms and beliefs (Thornhill & Fincher, 2014; Van Leeuwen et al., 2012). Our estimates of
pathogen prevalence in Study 2.1 relied on recent prevalence of infectious diseases in the U.S.,
making it crucial to replicate that cross-region effect across cultures. Previous cross-cultural
studies have also shown that moral systems differ across countries (Graham et al., 2011), and that
variation in pathogen prevalence is often associated with culture-level socio-psychological
outcomes (Murray & Schaller, 2010). Merging these two ideas, we postulate that cross-cultural
variation in pathogen prevalence is associated with moral systems while controlling for how
much those cultures are Western, Educated, Industrialized, Rich, Democratic (WEIRD)
(Henrich, Heine, & Norenzayan, 2010), homogeneous (Gelfand et al., 2011), or individualistic.
Study 2.2 was designed to evaluate the unique effect of historical ecological pathogen prevalence
on contemporary country-level moral values above and beyond cultures’ WEIRDness, tightness-
looseness, and individualism as potential confounding variables.
Methods
Pathogen Prevalence. We used the historical pathogen prevalence data compiled by
Murray and Schaller (2010), who used epidemiological atlases from the early 20
th
century to
gather prevalence data on nine infectious diseases (leishmanias, schistosomes, trypanosomes,
leprosy, malaria, typhus, filariae, dengue, and tuberculosis) in 230 geopolitical regions
worldwide. Since we have moral foundations data for 67 countries, we only used the historical
pathogen prevalence for the same countries. All 67 countries were contained within Murray and
Schaller’s (2010) data set.

16

Moral Values. We used nation-level data based on responses to the MFQ provided by
Atari, Lai, and Dehghani (2020) based on an online sample collected on yourmorals.org. The
data were collected from over 330,000 participants in 67 countries with at least 100 participants
per country. We collected estimates of five moral foundations (Care, Fairness, Loyalty,
Authority, and Purity). The median sample size in this data set was 439 participants per country.
Cultural Looseness. Cultural tightness (vs. looseness) taps into variance in norms,
values, and behavior. Tight cultures typically have many strong norms and a low tolerance of
deviant behaviors. Therefore, a tight culture facilitates homogeneity in traditions, norms, and
behaviors. Looser cultures score higher on well-being, freedom of choice, tolerance for sexual
deviations, and individualism, while being lower on traditionalism, population density, food
deprivation, natural disasters, and institutional repression (Gelfand et al., 2011). Cultural
looseness has been found to be associated with both moral values (Gelfand et al., 2011) and rate
of infectious diseases (Gelfand et al., 2020) across cultures, so we controlled for cultural
looseness in our analyses. We collected country-level indices of tightness-looseness from Uz
(2015), where smaller indices indicate tighter cultures. Out of 67 countries in our sample, 51 had
available data on looseness.
WEIRD cultural distances. A growing body of research suggests that populations
around the globe vary substantially along several important psychological dimensions, and that
people from societies characterized as WEIRD are particularly unusual (Henrich et al., 2010).
Recently, Muthukrishna et al. (2020) developed and validated country-level cultural distance
from the U.S., as a point of comparison. These authors’ “WEIRDness scores” are robust
indicators of cultural distance, grounded in evolutionary theory. Out of 67 countries in our
sample, 46 had data on WEIRD cultural distances.

17

Individualism. We controlled for individualism as it correlates with both pathogen
prevalence in different countries (Fincher, Thornhill, Murray, & Schaller, 2008) and moral
values (Hofstede, 2010). Country-level individualism indices were collected from Hofstede
(2010). Individualism, in this framework, is defined as a preference for a loosely-knit social
framework in which individuals are expected to take care of only themselves and their
immediate families, whereas collectivism represents a preference for a tightly-knit framework
in society in which individuals can expect their relatives or members of a particular ingroup to
look after them in exchange for unquestioning loyalty (Hofstede, 2010). Except for one
country (Afghanistan), we were able to secure individualism indices for all nations in our
sample.
Analytic Procedure. Since the number of cultures we had obtained data from was
small and countries are naturally geographically clustered, we used hierarchical linear models
to examine the predictive effect of country-level pathogen prevalence on country-level moral
values, while controlling for WEIRDness, cultural looseness, and individualism. We clustered
countries in six geographical regions: Africa, Asia, Europe, North America, Oceania, and
South America, yet Oceania was not included in the models as it only contains two countries
(Australia and New Zealand) and these countries’ looseness indices were missing. Our final
sample included 34 countries (Argentina, Bulgaria, Canada, Chile, Egypt, Finland, France,
Germany, Hungary, India, Indonesia, Iran, Italy, Japan, Mexico, Netherlands, Nigeria, Peru,
Philippines, Poland, Romania, Russia, Serbia, Singapore, Slovenia, South Africa, South
Korea, Spain, Sweden, Turkey, Ukraine, United Kingdom, United States, Vietnam) grouped
in five regions.

18

Results
The scatter plots of the relationships between pathogen prevalence and moral values
across 67 countries with complete data are shown in Figure 2.2. After controlling for
WEIRDness, cultural looseness, and individualism, multilevel models suggested that pathogen
prevalence, across 34 countries with complete data on all variables, was not associated with Care
(B = 0.04, SE = 0.06, p = .451) or Fairness (B = −0.01, SE = 0.06, p = .856). Loyalty was
negligibly positively associated with pathogen prevalence (B = 0.17, SE = 0.07, p = .021).
However, there was compelling evidence that country-level pathogen prevalence was associated
with Authority (B = 0.29, SD = 0.09, p = .003) and Purity (B = 0.27, SD = 0.09, p = .005).

19

Figure 2.2
Cross-country relationships between pathogen prevalence and moral values.

20

Endorsement of Purity and historical exposure to pathogens across 67 countries are
shown in Figure 2.3 (distributions of other moral values are presented in Supplementary
Materials). As shown, historical pathogen prevalence (i.e., aggregation of prevalence of
infectious diseases in the past century) is associated with contemporary Authority and Purity
values across countries (see Figure 2.2), above and beyond countries’ WEIRDness, looseness,
and individualism. Complete models are presented in Table 2.2.

21

Figure 2.3
Historical presence of infectious diseases and contemporary Purity values around the world.

22

Table 2.2
Five multilevel models predicting moral values across cultures

Dependent variable:

Care Fairness Loyalty Authority Purity
(1) (2) (3) (4) (5)

Pathogen Prevalence 0.042 −0.011 0.169
∗

0.288
∗∗∗

0.265
∗∗

(0.055) (0.060) (0.069) (0.086) (0.088)
Cultural Looseness −0.002 0.034 0.010 0.006 −0.080
(0.035) (0.038) (0.044) (0.053) (0.051)
Individualism 0.001 −0.0003 −0.004 −0.002 −0.004
(0.002) (0.002) (0.002) (0.002) (0.002)
WEIRDness Distance 0.084 0.648 1.085 0.518 0.693
(0.846) (0.913) (1.057) (1.268) (1.214)
Constant
3.437
∗∗∗

3.610
∗∗∗

2.464
∗∗∗

2.343
∗∗∗

1.914
∗∗∗

(0.115) (0.124) (0.143) (0.175) (0.178)

Observations 34 34 34 34 34
Log Likelihood 5.917 3.624 -0.559 -6.106 -5.713
Akaike Inf. Crit. 2.167 6.752 15.119 26.211 25.426
Bayesian Inf. Crit. 12.851 17.436 25.803 36.896 36.111

Note.
∗
p<0.05;
∗∗
p<0.01;
∗∗∗
p<0.001

23

Taking an evolutionary perspective, strict rules and social hierarchies dominated by
designated authorities or prestigious individuals have facilitated group coordination during
outbreaks. Of note, our results show no compelling evidence that Care, Fairness, or Loyalty can
be predicted by historical pathogen prevalence, especially when clustering is taken into account
and potential confounding variables are adjusted for. This cross-cultural study replicates the
Purity effect found in Study 2.1 (in U.S. counties), but also highlights the role of historical
pathogen prevalence on Authority values (e.g., obedience, respectfulness, hierarchy adherence)
across cultures. Purity and Authority values might be two equally important mechanisms to bring
order to cultures stressed by high prevalence of infectious diseases, one by strengthening social
hierarchies and order-restoring authorities, and one by reinforcing personal hygiene, inflicting
moral/social costs on promiscuity and uncleanliness.
Study 2.3
In this study, we examine the semantic association between pathogen-related words and
moral foundations using word embeddings — pre-trained models for distributed representation
of word meaning induced from patterns of word co-occurrences in a wide-coverage corpus of
text. In word-embedding models, each word is represented by a numeric vector such that the
geometry of the vectors captures semantic relations between the words. As such, if two concepts
are semantically close to one another, they should co-occur in similar linguistic contexts,
resulting in their embeddings being closer to one another in the semantic space (Collobert et al.,
2011; Garten et al., 2018). Since traces of psychological processes are manifested in language,
psychologically related concepts should co-occur in similar linguistic contexts (e.g., Garg,
Schiebinger, Jurafsky, & Zou, 2018). Therefore, we expect that pathogen-related concepts
should semantically be closer to the specific moral foundations which are psychologically more

24

related to them. Specifically, we use the word-embedding model of semantic representations to
examine the semantic similarity between pathogen-related words and words representing moral
foundations in English. In order to make sure that these associations are not idiosyncratic
features of English, we replicate this analysis in four other languages — Spanish, Farsi,
Japanese, and Hebrew — which represent a diverse collection of language families (Romance
Indo-European, Indo-Iranian, Japonic, and Northwest Semitic, respectively).
Methods
As discussed above, word embeddings are a popular natural language processing method
that represents each word by a vector, such that geometric relatedness — e.g., cosine similarity
between vectors — captures clusters of lexical semantic meaning and common usage patterns.
These similarities approximate both the intrinsic linguistic relationships among words, such as
the relationship among tenses for a given lemma, as well as their usage across large text corpora.
The “FastText” algorithm (Bojanowski, Grave, Joulin, & Mikolov, 2017) generates word
vectors using subword information in the same style as the popular “Word2vec” skipgram
algorithm (Mikolov, Sutskever, Chen, Corrado, & Dean, 2013), and has been used to generate
word embedding sets for a large host of languages (Grave, Bojanowski, Gupta, Joulin, &
Mikolov, 2018). We used 300-dimension FastText vectors trained on the “Common Crawl”
corpora
4
, for each language in our analyses.
We measure the moral loading of pathogen-related words by computing the cosine
similarity between each pair of moral- and pathogen-related words. Moral seed words were
taken from Garten et al. (2018), and a set of ten pathogen-related words was generated for this
work: “virus”, “flu”, “disease”, “infection”, “sickness”, “germ”, “contagion”, “illness”
(translated to Spanish, Farsi, Japanese, and Hebrew by native speakers who were blind to the

25

purpose of the study; see Supplementary Materials). The cosine similarity for all pairs of moral
and pathogen-related words are calculated to analyze the semantic similarity between moral
foundations and pathogens in natural language (see Garten et al., 2018). To examine the
similarities between moral foundations and pathogen-related words, a robust one-way Analysis
of Variance (ANOVA) was conducted with p-values adjusted for False Discovery Rates (FDR)
and pairwise comparisons based on Yuen’s trimmed means test. As a measure of effect size, we
relied on “explanatory measure of effect size” ξ which does not require equal variances and can
be generalized to multiple group settings (Wilcox & Tian, 2011).
Results
Similarity results revealed that Purity and Care words were closest to the pathogen-
related words in English (Purity and Care were not different from one another in their similarity
to pathogen-related words, p = .643). The robust one-way ANOVA showed that differences
between the five similarity distributions were significantly different (F (4, 127.41) = 14.23, ξ =
0.45, 95%CI = [0.34, 0.56], p < .001). Post-hoc pairwise tests indicated that Purity was
semantically closer to infectious diseases than Authority (p < .001), Loyalty (p < .001), and
Fairness (p = .011). Next, Care was shown to be more related to infectious diseases words than
were Fairness (p < .001), Loyalty (p < .001), and Authority (p < .001). Similarities of Fairness,
Loyalty, and Authority to pathogen words were not different from each other (ps > .643). In sum,
words related to infectious diseases appear substantially more often in similar contexts with
Purity and Care words compared with Fairness, Loyalty, and Authority. We also replicated these
patterns in Spanish (F (4, 109.57) = 2.85, ξ = 0.24, 95%CI = [0.09, 0.34], p = .027), Farsi (F (4,
123.22) = 2.58, ξ = 0.19, 95%CI = [0.05, 0.27], p = .040), Japanese (F (4, 112.03) = 6.18, ξ =
0.30, 95%CI = [0.17, 0.38], p < .001), and Hebrew (F (4, 93.37) = 7.63, ξ = 0.41, 95%CI =

26

[0.28, 0.51], p < .001), but the effect sizes were smaller in other languages (see Supplementary
Materials for details). These findings suggest that across five languages originating from
different cultures and language families, pathogen-related words co-occur with Purity and Care
more frequently than other moral concerns, highlighting that Purity and Care concerns are
associated with cognitive processes (e.g., passing moral judgments about suffering or
cleanliness) underlying representations of pathogen-related concepts.
Study 2.4
The previous studies are community-level analyses, investigating the relationship
between ecological pathogen prevalence and region-level moral concerns, in addition to
linguistic association of pathogen and morality-related words. Indeed, macro-level analyses of
communities cannot be generalized to individual-level processes as that risks committing the
ecological fallacy. Given that group and individual levels of analysis often do not yield
isomorphic results (Oyserman, Coon, & Kemmelmeier, 2002), deeper insights can be gained by
simultaneously analyzing the independent effect of group-level and individual-level processes. In
Study 2.4, we aim to examine how individual differences in moral concerns are related to
pathogen avoidance behaviors. This study further clarifies individual-level moral psychological
processes in avoiding infectious diseases. Specifically, we examine the relationship between
moral foundations and pathogen avoidance behaviors while controlling for political orientation
and religiosity in a cross-sectional design using a stratified sample from the United States.
Methods
Participants. We aimed to recruit a sample of 500 participants to detect small
correlational effects (ρ = .15) with high power (95%). We recruited a stratified national U.S.

27

sample from Qualtrics Panels, balanced with respect to age, gender, and political affiliation.
Participants (N = 513) were almost half (51.3%) female and predominantly (63%) White
American. In terms of age, 11.1% were in the 18-24 range, 18.1% were in the 25-34 range,
16.8% were in the 35-44 range, 18.9% were in the 45-54 range, 17.0% were in the 55-64 range,
and 18.1% were 65 years old or older. In terms of political affiliation, 263 participants (51.3%)
identified as a Democrat while the rest identified as a Republican.
Design and Materials. Our Institutional Review Board (IRB) approved this study (UP-
19-00395). To measure the constructs of interest, participants completed a set of measures
including the short Moral Foundations Questionnaire (MFQ-20; Graham et al., 2011)
(Cronbach’s αs = .62 - .73) and a 2-item measure of pathogen avoidance behavior (Cronbach’s α
= .59, 95%CI = [.52, .66]) along with their demographic details. We collected data in this study
in March 2019. These two questions, designed for this study, were “Imagine that a few months
ago you booked a flight to a beautiful country abroad for a vacation. One or two days before
your flight you find that there has been an outbreak of a dangerous infectious disease in that
country, but the local authorities have controlled the situation. How likely is it for you to go on
this vacation?” and “Some infectious diseases cannot be transmitted through skin contact unless
one’s skin has small bloody cuts. How likely is it for you to sit next to a person with such a
disease on the bus?”. Both questions were rated on a 7-point Likert-type scale ranging from 1
(Extremely unlikely) to 7 (Extremely likely). In this cross-sectional observational study, all items
and measures were counterbalanced. We used linear regression models to account for gender,
and political ideology, along with moral foundations to predict pathogen avoidance behaviors.

28

Results
In a linear model controlling for gender and political ideology, Purity was
significantly positively associated with pathogen avoidance behaviors (β = 0.25, SE = 0.09, p
= .006). Fairness was also shown to have a negative association with pathogen avoidance (β
= -0.34, SE = 0.13, p = .010). Other moral values, political ideology, and gender had non-
significant effects (ps > .07). These results highlight the individual-level processes between
pathogen avoidance behaviors and moral Purity. These patterns are consistent with our
cross-cultural and linguistic analyses (Studies 2 and 3), again implicating Purity concerns in
relation to perceptions of pathogens.
Study 2.5
When people feel more in immediate danger of infectious diseases they report lower
levels of sociability (i.e., decreased interest in frequent contact with others) and produce more
avoidant motor responses (Mortensen, Becker, Ackerman, Neuberg, & Kenrick, 2010). Research
also suggests that moral vigilance is increased when people are experimentally exposed to
salience of infectious disease threats (Murray, Kerry, & Gervais, 2019). In order to complement
our previous analyses using an MFT framework, in the current study we use an experimental
design and expose participants to visual cues of pathogens and assess their moral judgments
compared with a control group exposed to neutral stimuli.
Methods
Participants. Based on prior work and a power analysis to detect a small-to-moderate
effect size (r = .20) at p = .05 and 90% power, we aimed to collect 320 participants. We recruited
334 participants from Amazon Mechanical Turk. After removing the participants who failed an

29

attention check, a total of 316 participants remained in the sample (130 male, 182 female, 4
other). Most participants were White Americans (n = 247), followed by Black Americans (n =
41). The mean age was 32.6 years (SD = 11.3 years).
Design and Materials. Culpepper, Havlíček, Leongómez, and Roberts (2018) identified
main domains of pathogen-related disgust and generated a novel visual stimulus set of 20 images
depicting scenes of highly salient pathogen risk, along with a paired control set (20 images) that
are visually comparable but lack the pathogen cues. Our Institutional Review Board (IRB)
approved this experimental study (UP-18-00712). Participants were randomly assigned to the
experimental (vs. control) group, receiving experimental (vs. control) stimuli. Participants in the
experimental (vs. control) group were given three pathogen images (vs. non-pathogen
counterparts) and were asked to look at them for a few seconds and then answer questions about
these images. As a manipulation check, we asked participants to rate how “pleasant” each scene
was (1 = “Not at all pleasant”, 5 = “Extremely pleasant”). Right after being exposed to
pathogen (vs. non-pathogen) cues, participants completed the 30-item MFQ (Graham et al.,
2011) (Cronbach’s α = .62 - .78) along with their demographic details. Participants also
completed the Short-Form of the Positive and Negative Affect Schedule (PANAS; Thompson,
2007) after completing the MFQ to measure their momentary negative affect (Cronbach’s α =
.81) after being exposed to experimental (vs. control) stimuli. We controlled for negative affect
to account for the role of negative affect that might be evoked by pathogen stimuli in moral
judgments.
Results
First, we compared the “pleasantness” ratings to make sure that our experimental
manipulation worked. The experimental group (M = 0.18, SD = 0.41) rated the images

30

significantly less pleasing compared to the control group (M = 1.09, SD = 0.65), t = 14.63,
Welch-corrected df = 256.45, p < .001, Cohen’s d = 1.65). We ran regression analyses to predict
scores on moral foundations by condition (experimental vs. control), while statistically
controlling for gender, age, and negative affect. Results suggested that being exposed to
pathogen cues does not predict Care (B = 0.18, SE = 0.49, p = .723), Fairness (B = -0.30, SE =
0.48, p = .527), Loyalty (B = 0.25, SE = 0.65, p = .705), or Authority (B = 0.50, SE = 0.60, p =
.404). However, participants in the experimental (vs. control) condition scored an average of
1.54 (SE =0.74, p = 0.038) points higher on the Purity questionnaire items than participants in
the control condition. Therefore, in experimental settings, activating the behavioral immune
system by visual cues to pathogens can influence Purity judgments, but not moral judgments in
other domains. Domain-specific effects of pathogen cues highlight the adaptive benefits of Purity
in immediate pathogen-rich environments. In other words, Purity is a context-sensitive,
environmentally-plastic, pathogen-neutralizing suite of moral psychological mechanisms that
function to avoid pathogen contact.
Study 2.6
Our community-level analyses were observational, hence no strong claims can be made
with regard to the direction of the effect in the relationship between Purity and pathogen
prevalence. Also, our data were contemporary, masking historical precedence of these variables
over one another. For example, it is not clear whether particularly low levels of Purity (and other
moral values) precede higher levels of pathogen prevalence or, alternatively, high levels of
pathogen prevalence cause moral values to change in subsequent years. Here, we explore this
question by combining historical data on infectious diseases in the U.S. as well as historical
language data on moral values. Indeed, cultural and societal changes can be captured using

31

linguistic analysis of large historical corpora of books and texts produced in recent history
(Greenfield, 2013). Specifically, we rely on moral language used in published books in the 20
th

century (Michel et al., 2011), and historical prevalence of infectious diseases in the U.S. to
examine the temporal link between infectious diseases and severity of moral language, while
controlling for fluctuations in cultural looseness as a potential confounder.
Methods
Pathogen Prevalence. Data were adapted from Grossmann and Varnum (2015), who
reported 9 of the most frequent infectious diseases reported by the historical records of the CDC.
This data set included tuberculosis, syphilis, gonorrhea, malaria, typhoid and paratyphoid fever,
diphtheria, pertussis, measles, and poliomyelitis. The data included prevalence rates from 1912
through 2012.
Moral Language Usage. In order to collect yearly usage of moral language, we relied on
Google Ngram data, which is the largest available time-stamped corpus through 2008 (Michel et
al., 2011).The current data ranged from 1900 to 2008. The corpus consists of words and phrases
(i.e., n-grams) and their usage frequency over time. We relied on the Moral Foundations
Dictionary (MFD; Graham et al., 2009) and collected the frequency for each word in the
dictionary from 1900-2008.
Cultural Looseness. Cultural looseness was controlled for, as it correlates with both
moral values and existence of pathogenic threats (Gelfand et al., 2011). We used Jackson et al.’s
(2019) estimates of cultural looseness in recent history. These authors used Google Ngram data,
published between the years 1800 and 2000 and reported standardized frequency of loose (e.g.,
“Allow”, “Freedom”, “Autonomy”) and tight (e.g., “Restrain”, “Prevent”, “Adhere”) words.

32

Analytic Procedure. Prior to our analyses, we de-trended our time-series vectors by
regressing out monotonic effect of time, general moral language, and cultural looseness. Then we
subjected each time-series vector to augmented Dickey-Fuller root tests, to evaluate whether a
time-series vector has an underlying trend that renders it non-stationary. We next used
standardized vectors to examine the correlation between moral language and pathogen
prevalence, and used cross-correlations to test whether moral norms preceded decrease in the
prevalence of infectious diseases. Finally, we conducted additional tests of Granger causality to
assess the relationship between moral norms and pathogen prevalence (as well as the reverse
relationships) in the 20
th
century. Granger tests of causality are more conservative than cross-
correlations as they evaluate whether one time-series variable is predicting changes to another
time-series variable above and beyond values of the outcome (Seth, Barrett, & Barnett, 2015).
Results
Cross-correlations (the correlations between two variables at different time lags) are
visualized in Figure 2.4: negative lags (left side of the dashed red line) indicate that pathogen
prevalence precedes changes in moral language usage, while positive lags (right side of the
dashed red line) indicate that moral language usage precedes pathogens. As can be seen,
pathogen prevalence predicted immediate slight increases in Care and Loyalty. However, the
largest effect was observed for Purity. Cross-correlations between Purity language and pathogen
prevalence suggested that not only do they significantly co-occur in the same years, but pathogen
prevalence seems to drop immediately in subsequent years of high Purity norms in the United
States (see Figure 2.4), supporting the prediction that Purity functions to reduce communicable
diseases. We further investigated the direction of these relationships in Granger causality tests
(Table 2.3). We found strong evidence, consistent with cross-correlations, that Purity

33

immediately “Granger caused” pathogens to drop (1-year lag), and the effect held for the 5-year
lag too. Of note, high Care values seem to also precede higher rates of infectious diseases, but
the effect did not hold for the 5-year lag. In the reverse models in Table 2.3 (pathogens “Granger
causing” moral foundations), we did not find evidence suggesting that pathogen prevalence can
have immediate 1- or 5-year effects on moral foundations except for a small effect on Care
values with a 5-year lag.

Figure 2.4
Cross-correlations between pathogen prevalence and indicators of moral language. Correlations
outside the dashed blue horizontal lines are significant at α = .01. Negative lags (left side of the
dashed red vertical line) indicate that shifts in pathogen prevalence led to shifts in moral
language, whereas positive lags (right side of the dashed red vertical line) indicate that shifts in
pathogen prevalence followed shifts in moral language.

34

Table 2.3
Granger Causality Test Results with 1-year-lagged and 5-year-lagged Models

Granger Causality Model 1-year lag 5-year lag

Pathogen → Care F(1,86) = 0.14 F(5,78) = 4.06**
Pathogen → Fairness F(1,86) = 1.71 F(5,78) = 0.91
Pathogen → Loyalty F(1,86) = 0.31 F(5,78) = 2.84
Pathogen → Authority F(1,86) = 0.96 F(5,78) = 2.97
Pathogen → Purity F(1,86) = 2.36 F(5,78) = 0.49
Care → Pathogen F(1,86) = 8.91** F(5,78) = 1.86
Fairness → Pathogen F(1,86) = 0.04 F(5,78) = 1.71
Loyalty → Pathogen F(1,86) = 2.47 F(5,78) = 1.67
Authority → Pathogen F(1,86) = 0.79 F(5,78) = 1.11
Purity → Pathogen F(1,86) = 20.11*** F(5,78) = 3.44**

Note.
**
p < .01
***
p < .001

In this study, the results indicate that historical prevalence of pathogens highly co-occur
with moral values associated with group cohesion, specifically Purity values (indexed by more
frequent usage of Purity words), after controlling for other types of moral language and historical
estimates of cultural looseness. Cross-correlations and Granger causality tests revealed a pattern
where higher levels of Purity values co-occur with, and immediately causes, lowered levels of
pathogen prevalence in the United States in the 20
th
century. Therefore, it seems that

35

psychological mechanisms associated with Purity can act as a unique pathogen-neutralizing
mechanism (possibly through social and physical distancing), lowering subsequent levels of
infectious diseases. The finding that lower Purity values are historically related to more
infectious diseases can be explained by the fact that low Purity values bring about higher
unprotected sexual encounters, higher physical proximity-seeking, and lesser aversion toward
disgusting things, which in turn can exacerbate the contagion of infectious diseases such as
sexually transmitted infections.
General Discussion
A single pandemic caused by an infectious disease can kill tens of millions of people
worldwide and make hundreds of millions ill. In addition to biological systems to fight off
pathogens, humans have developed behavioral systems to avoid pathogenic agents before bodily
contact, or to stay away from highly contagious environments. Here, in a series of studies, we
demonstrate that across U.S. regions (Study 2.1), countries (Study 2.2), linguistic contexts
(Study 2.3), individuals (Study 2.4), and experimental conditions (Study 2.5), presence of
pathogens is consistently linked to moral Purity systems. Indeed, Purity values may function to
neutralize infectious diseases, possibly through lowered sexual contact, hygienic practices,
avoidance of unfamiliar foods and/or persons who may be perceived to be associated with the
origin of the disease. Finally, our historical analysis in Study 2.6 demonstrated that high moral
Purity norms in the 20
th
century in the U.S. is followed by a decrease in infectious diseases.
The unique relationship between pathogen salience and Purity highlights the
evolutionarily crucial role of moral Purity in survival and flourishing of the species across
historical periods and geographic regions in the face of recurring increase in the occurrence of
infectious diseases. At the county and national levels, Purity values might be more successfully

36

transmitted and sustained within pathogen-rich ecologies if such norms lead to reduced contact
with pathogenic agents (Murray & Schaller, 2016). Prior psychological work indicates that
pathogens can result in the cultural evolution of prophylactic norms and rituals (Tanaka, Kumm,
& Feldman, 2002; Tybur et al., 2016) embedded in Purity value systems, possibly through
lowering the likelihood of unsafe sex, avoidance of unhygienic food, and shunning pathogen-
rich places. In addition, Purity has been shown to have a particularly unique effect in promoting
within-coalition alliances, an effect referred to as “purity homophily” (Dehghani et al., 2016),
which in turn can provide ingroup coalitions in times of disease outbreaks especially in
pathogen-dense ecologies (Navarrete & Fessler, 2006). Indeed, in ancestral environments,
interaction with ingroup members may have posed less risk of disease transmission than
interaction with an outgroup member, since individuals possessed antibodies to many of the
pathogens present in their own community, in contrast to those circulating among people coming
from other regions.
At the individual level, people in high avoidance of infectious diseases might find Purity
rituals (e.g., refraining oneself from sexual intercourse with an unknown individual) appealing
for a number of reasons. First, “pure” sexual practices often expose individuals to substantially
fewer sexually transmitted infections (Bauch & McElreath, 2016). Second, Purity practices
facilitate safer and more traditional food preparation techniques which often include ingredients
with antimicrobial properties (Billing & Sherman, 1998). Third, hygiene-related Purity rituals
(e.g., burial rituals) can coordinate behaviors to limit pathogen transmission. We argue that each
of these lower-order mechanisms have intrapersonal (e.g., refraining oneself from contact with
pathogens) and interpersonal (e.g., penalizing others’ contact with pathogens) components.
While pathogenic threats can be detrimental to the survival of an individual or group, their

37

extremely low rates can have harmful effects on our health. The Hygiene Hypothesis, for
example, suggests that humans have evolved in environments where childhood exposure to
infectious agents was high. In industrialized settings, immunological diseases become common
due to the lack of this exposure (e.g., Apicella & Barrett, 2016; Yazdanbakhsh, Kremsner, &
Van Ree, 2002).
Care values, as implicated in our linguistic, temporal, and county-level analyses, might
be specific to the suffering and mortality accompanied by severe infectious disease (see Studies
1, 2, and 3). Particularly, it could be the case that pathogen outbreaks may lead to temporary,
subsequent increases in the tendency to care for afflicted individuals. In addition, individuals
typically talk about harms, sufferings, and deaths when speaking about infectious disease, which
explains our linguistic findings in Study 2.3. Authority values were also linked to country-level
pathogen prevalence in our cross-cultural analysis, almost as strongly as Purity values (Van
Leeuwen et al., 2012). This finding can be understood within cultural evolution frameworks
which combine familial transmission with selective learning from locally prestigious individuals.
In high-pathogen ecologies, individuals are safer as long as they are under family ties and abide
by powerful prestigious authorities (who potentially have more social power and/or more
knowledge), whereas such an effect may not hold for historically low-pathogen, WEIRD cultures
such as the U.S. or Canada. These results, along with little to no evidence regarding the
relationship between pathogen prevalence and ingroup loyalty, are wholly consistent with
previous findings in the cultural evolution literature (Hruschka & Henrich, 2013). Lack of strong
links between pathogen prevalence, Fairness, and Loyalty across time and space may also
suggest that Fairness and Loyalty are less pathogen-plastic, i.e., less dependent upon the
ecological factor of pathogen prevalence across time and place. Purity, on the other hand,

38

appears to be highly functionally flexible with respect to pathogens in the environment, i.e.,
sensitive to costs and benefits of pathogen avoidance in the environment. The results of these
studies broadly support the theory that purity-based moral systems have their functional roots in
the survival and reproduction in our species, and that Purity norms are culturally learned (Boyd,
Richerson, & Henrich, 2011).
While we highlight the specialized role of Purity in response to pathogenic threats, we
also warn against its exclusive nature in contemporary intergroup dynamics. That is, Purity
functions as a double-edged sword in fending off pathogenic threats such as COVID-19: on the
advantageous side, Purity values encourage social distancing through perceived severity of the
threat; and on the socially destructive side, they can lead to antipathic sentiment toward outgroup
members, especially those who are perceived to be somehow associated with the origin of the
outbreak (i.e., the Chinese in the case of COVID-19) (Rzymski & Nowicki, 2020). Purity values
can predispose people to negatively evaluate outgroup members who are perceived to be
potential carriers of pathogens, subsequently justifying prejudicial or exclusive behaviors (e.g.,
hate crimes against outgroup members) or policies (e.g., imposing travel bans or deportations).
By integrating insights from psychology, anthropology, and behavioral ecology paired
with methodologies from multiple disciplines, this research illuminates a unified way to better
understand and explain historical and contemporary variation in moral Purity (as well as other
moral foundations) in response to infectious diseases. This set of studies suggests that Purity
consistently functions as a mechanism to allay pathogenic threats across time and space at
different levels of analysis, further shedding light on evolutionary architecture and socio-
ecological plasticity of human moral systems.

39

Chapter 3: Foundations of Morality in Iran
The study of morality in psychological science has been, in large part, limited to Western,
Educated, Industrialized, Rich, and Democratic (WEIRD; Henrich, Heine, & Norenzayan, 2010)
populations (see Skitka & Conway, 2019). Most existing studies examining moral values in non-
WEIRD nations have typically failed to develop cultural models of moral cognition or to identify
culture-specific values. Rather, they have conventionally used a WEIRD model of morality
without adequate culturo-linguistic adaptation of such models. Graham et al. (2013) argued that
“as Shweder (1990) says, each culture is expert in some aspects of human flourishing, but not all.
Although we are working with researchers in other nations to explore the morality of other
cultures, much more work needs to be done to move beyond WEIRD research samples.” Given
the dearth of systematic morality research in non-WEIRD cultures, it is not yet known whether
WEIRD people are “moral outliers” in terms of the structure and content of their moral domain.
Iran, heir to an ancient civilization, is a particularly understudied cultural setting in moral
psychology. Here, we investigate Moral Foundations Theory (MFT; Graham et al., 2013; Haidt
& Joseph, 2004) as a widely-used, evolutionarily-informed, and cultural theory of morality
(Study 3.1), compare moral foundations and their underlying network structures between Iran
and the US (Study 3.2), conduct field interviews on beliefs about virtues and vices in Iran (Study
3.3), broaden the nomological network of a culture-specific moral system (“Qeirat”) in Iran
(Study 3.4), and demonstrate the pragmatic validity of “Qeirat” values for answering scientific
questions in morality in the Iranian context (Study 3.5).

40

Is Iran WEIRD?
A growing body of work indicates that populations around the world vary substantially
along important psychological dimensions, and that people from WEIRD societies are
particularly unusual (Gächter & Schulz, 2016; Henrich et al., 2010; Medin, Bennis, & Chandler,
2010; Rad, Martingano, & Ginges, 2018; Talhelm et al., 2015). For example, people from
WEIRD populations tend to be, on average, more individualistic, analytically-minded, and
impersonally prosocial while revealing less conformity, obedience, in-group loyalty, and
nepotism (Kanagawa, Cross, & Markus, 2001; Kitayama, Park, Sevincer, Karasawa, & Uskul,
2009). Indeed, our species is fundamentally cultural, and thus these cultural differences are
essentially psychological differences (Henrich, 2015). Some of these differences are now well
documented; however, efforts to explain this variation from evolutionary and historical
perspectives have just begun, with psychologists increasingly expressing more interest in non-
WEIRD samples (see Apicella & Barrett, 2016).
In our research, we contribute to this growing literature by studying Iran’s moral
psychology. However, simply calling Iran non-WEIRD is overly simplistic; a psychological
analysis of a non-WEIRD culture must include a nuanced, descriptive analysis of the studied
population. By continuing to refer to this research domain as the divide between WEIRD and
non-WEIRD, one could create a false dichotomy of populations (Muthukrishna et al., 2020),
hiding valuable nuance that influence empirical measurements and theoretical interpretations
(e.g., Doğruyol, Alper, & Yilmaz, 2019; Shaw, Cloos, Luong, Elbaz, & Flake, 2020). In this
section, we overview a socio-psychological and demographic profile of Iran, in order to situate
our studies and findings on the moral profile of the country. Given the aforementioned definition

41

of WEIRD psychology, Iran remains a “weird” non-WEIRD culture. For instance, Iran is
geographically and historically close to countries like Pakistan, Egypt, Yemen, Azerbaijan, or
Armenia, but is more educated and developed than these countries (United Nations Development
Programme, 2018). Using Muthukrishna et al’s WEIRD cultural distances, Iran’s distance from
the US is 0.15, comparable to Turkey and Armenia, and slightly higher than Japan’s distance to
the US (see Table 1). Moreover, analyses suggest that the people of Iran are more genetically
similar to people in Turkey than Saudi Arabia; however, culturally it is less clear which one Iran
is closer to (Bell, Richerson, & McElreath, 2009). In what follows, we provide a summary of
Iran on the dimensions of WEIRDness. It is important to note that WEIRDness is used as a
rhetorical tool to highlight that people from WEIRD populations are outliers on many
measurable psychological phenomena; it is not a theoretical construct.
Table 3.1
Iran’s GDP per capita, human development, and cultural distance from the USA and China
Country
GDP per capita
(2017)
Human
Development Index
(2018)
Cultural Distance
from the USA [95%
CI]
Cultural Distance
from China [95%
CI]
Iran $5,520.3
0.797 (world:
0.731)
0.150 [.145, .156] .125 [.122, .128]

Many Iranians identify as “Asian” (since Iran is located in the continent of Asia) rather
than “Middle Eastern” (which is a relatively vague term that originated not in the region itself,
but in Europe; Koppes, 1976). With respect to ethnicity, Iran has multiple ethnicities (e.g., Fars,
Kurd, Turk, Balooch), and Iranians living in the US typically (80%) consider themselves
“White” in American census surveys (Parvini & Simani, 2019). In terms of education, at national
level, 93% (close to typical WEIRD populations) of the population is literate (97% in young

42

adults) according to official reports (ISNA, 2019). Access to higher education has dramatically
increased over the last three decades and is currently highly accessible to people and is mostly
free, with women getting more degrees in higher education than do men (Shams, 2016). In fact,
Iran is among the world’s leaders in the percentage of women graduating in Science,
Technology, Engineering, and Mathematics (STEM) fields (see Richardson et al., 2020; Stoet &
Geary, 2018).
In terms of industrialization and economy, the country’s overall infrastructure, educational
system, legal system, and modern industries were substantially improved from 1925 to 1941.
During this time, Iran experienced a period of social change, economic development, and relative
political stability (Katouzian, 1981). Between 1964 and 1978, Iran’s gross national product grew
with the oil, gas, and construction industries expanding by almost 500%. The number of women
enrolling in higher education increased from 5,000 in 1967 to more than 74,000 in 1978. Before
the Islamic Revolution in 1979, the country was considered a growing power in the region,
however, after the revolution and concurrent with the US sanctions and the imposed Iran-Iraq
war (1980-1988), the economy underwent substantial fluctuations (Gheissari, 2009). Yet, over
the last three decades (1990-2020), Iran has been growing in terms of Gross Domestic Product
(GDP) per capita, women’s access to higher education, oil industry, clean energy, sustainable
clean water, public health, and agricultural products (International Monetary Fund, 2019) (see
Table 3.1).

43

There has been substantial debate over how “democratic” Iran is. Democracy is difficult to
measure, particularly in non-WEIRD countries. By one metric, known as Polity IV
1
, which relies
on several indicators to rate countries (from -10 for full dictatorship to 10 for full democracy),
Iran is rated as -7, the same as Cuba and China. Another ranking agency, known as V-Dem
2
,
ranks Iran as more democratic, with a score of 0.29 on a scale from 0 to 1, slightly less
democratic than Ukraine or Tunisia. In sum, although Iran’s population is highly educated and
industrialized, it is less Westernized, rich, and democratic than typical populations examined in
psychology journals (Thalmayer, Toscanelli, & Arnett, 2020).
Moral Foundations Theory
Moral Foundations Theory was developed in order to fill the need of a systematic theory of
morality, explaining its evolutionary origins, developmental aspects, and cultural variations.
Haidt and Joseph (2004) examined the existing literature in evolutionary psychology and
anthropology, looking for virtues and facets of moral regulation that were common across
cultures. Relying on the above-mentioned sources, these authors suggest five candidates for
being the basic, omnipresent psychological “foundations,” evolutionarily prepared psychological
predispositions upon which cultures construct their moral judgments and systems. Thus, MFT
can be viewed as a systematic attempt to specify the psychological mechanisms which allow for
intuitive bases of moral judgments as well as moral reasoning. Care, Fairness, Loyalty,
Authority, and Purity are theorized to have solved adaptive problems over humans’ evolutionary
history (at both individual and group levels; e.g., Koleva, Selterman, Iyer, Ditto, & Graham,

1
http://www.systemicpeace.org/
2
https://www.v-dem.net/en/

44

2014; Van Leeuwen, Park, Koenig, & Graham, 2012). These foundations are hypothesized to be
universally available, yet culturally variable (Haidt, 2007). Importantly, MFT theorists
maintained that they “do not believe these are the only foundations of morality. These are just
the five [they] began with— the five for which [they] think the current evidence is best”
(Graham et al., 2013, p. 67).
The first two foundations (Care and Fairness) are often referred to as individualizing
foundations while the other three (Loyalty, Authority, and Purity) are labeled binding
foundations (Graham & Haidt, 2010). The individualizing foundations correspond to the
traditional focus of moral psychology on ethics of individual rights (Haidt, 2007); however, the
binding foundations correspond to more collectivist virtues that were underrepresented in
WEIRD studies of morality before MFT was theorized (e.g., Yilmaz, Harma, Bahçekapili, &
Cesur, 2016). Therefore, MFT provides a unique opportunity for evolutionarily-informed, cross-
cultural research on moral values (Atari, Lai, & Dehghani, 2019).
The Moral Foundations Questionnaire (MFQ; Graham et al., 2011) is the primary self-
report measure of the degree to which individuals endorse each of the five dimensions introduced
by MFT. This measure has fared well in terms of validity and reliability in WEIRD populations
(Davies, Sibley, & Liu, 2014; Graham et al., 2011; Nilsson & Erlandsson, 2015; Métayer &
Pahlaven, 2014; van Leeuwen & Park, 2009), but its validity is less clear in non-WEIRD cultures
(e.g., Atari et al., 2019; Yilmaz et al., 2016). Recently, Doğruyol et al. (2019) provided evidence
that the five-factor model of moral foundations, operationalized by the short version of the MFQ,
is stable and invariant across WEIRD and non-WEIRD societies; however, these authors used
the problematic dichotomy of WEIRD vs. non-WEIRD rather than treating societies on a

45

continuum of WEIRDness (Muthukrishna et al., 2020). Research using the MFQ has provided
valuable information regarding between-group differences in moral values. Graham et al. (2011)
compared participants from Eastern cultures (South Asia, East Asia, and Southeast Asia) with
participants from Western cultures (United States, United Kingdom, Canada, and Western
Europe) and found small effect sizes with Easterners scoring higher on Loyalty and Purity, while
Westerners scoring negligibly higher on Care, Fairness, and Authority. Another well-developed
line of research has examined liberal-conservative differences in moral foundations. For
example, conservatives (vs. liberals) rely more strongly on binding moral foundations (Graham,
Haidt, & Nosek, 2009). In addition, MFQ factors have been found to be correlated with
normative personality traits, specifically the five-factor model of personality (i.e., Extraversion,
Agreeableness, Conscientiousness, Neuroticism, and Openness to Experience; see McCrae &
John, 1992). Specifically, higher scores on individualizing foundations have been found to be
associated with higher scores in Agreeableness, Neuroticism, and Openness to Experience.
Higher scores on the binding foundations, on the other hand, have been associated with higher
Extraversion, higher Conscientiousness, and lower Openness to Experience (Lewis & Bates,
2011).
Graham and Haidt (2010) posit that religion binds individuals into moral communities.
They claim that religion is intertwined with all five moral foundations, but most strongly
promotes “binding” foundations. Consistently, Loyalty, Authority, and Purity (i.e., the binding
foundations) are typically found to be associated with religiosity in different contexts. For
example, Johnson and colleagues (2016) found the binding foundations to be correlated with
different facets of religiosity (outreaching faith, religious commitment, authoritarian
representation of God, and Biblical literalism) in a sample of Christian Americans. More recent

46

correlational studies have successfully replicated this effect (see Yi & Tsang, 2020). Yalçındağ
et al. (2019) replicated these associations between religiosity and bindings values in Turkey and
also found a small negative association between religiosity and Fairness.
Unlike the US, moral psychology is only an emerging field in Iran, and the Persian
translation of MFQ has only been used in one empirical study (Seifi-Ghozlu, Hamidi, Sharifi, &
Khalili, 2015), to our knowledge. The existing morality research in Iran has focused attention on
bioethics (e.g., Larijani & Zahedi, 2008), moral development in children (e.g., Hooshyari,
Delavar, Minaee, & Eskandari, 2018), business ethics (e.g., Mujtaba, Tajaddini, & Chen, 2011),
“immoral” personality traits (e.g., Aghababaei, Mohammadtabar, & Saffarinia, 2014), sacred
values (Dehghani et al., 2010), and moral dilemmas (e.g., Sachdeva, Iliev, Ekhtiari, & Dehghani,
2015). To date, Iranian researchers have not developed culture-specific models of moral values.
Iranians’ Beliefs about Right and Wrong
Although empirical research is lacking with regard to moral values in Iran, it is worthwhile
to briefly review lay beliefs about right and wrong in Iranian culture. Especially as we take an
evolutionary approach to the interpretation of culture-based differences in moral values, it is
relevant that a region (or society) possesses a distinct moral history, and that current differences
are not just a product of recent politics, but can be the result of cultural learning across
generations (Schulz, Bahrami-Rad, Beauchamp, & Henrich, 2019). Historically, Iran is heir to
two ancient moral codes: Zoroaster’s (∼1000 BCE) tenet of good thoughts, good words, good
deeds (see Jackson, 1896) and Cyrus’s (∼539 BCE) cylinder considered the first human rights
document (see Daryaee, 2009). Even in the current-day Muslim-majority Iran, these two moral

47

precepts are highly respected. Both of these codes emphasize caring for other humans and justice
in the society.
In recent history, Iranian scholars have attempted to identify moral values in the
contemporary Muslim-majority Iran. Jamalzadeh (1966), over half a century ago and based on
mere observation, argued that Iranian virtues are intellect, hospitability, compassion, and
friendliness, whereas vices are believed to be lying, cheating, flattery, and greed. Delkhamosh
(2004) surveyed university students and identified traditionalism, respect for religious
authorities, and nationalist pride (“Ergh-e-Melli”) as important moral values in Iranians. Nejat,
Bagherian, Hatami, and Shokri (2015) conducted a qualitative study in Tehran to identify mental
representations of moral concerns and identified kindness, honesty, humility, sacrifice, unity,
social order, meritocracy, respectfulness, observing religious teachings, Ta’asob (a type of
“honor” which is focused on protection of a deeply held belief), patience, Hijab, liberty, and free
speech as virtues. Harassment, violence, lying, discrimination, infidelity, reneging, profanity,
defamation, subversion, oppression, adultery, and sexual harassment were also identified as vices
reported by Iranians. Religion scholars in Iran (e.g., Tabatabai, 1996), have identified five moral
“traits” — based on teachings from the Quran and Hadith — that are highly virtuous and
prophet-like: (1) “Qeirat”
3
(a type of “honor” which is focused on caring for spouse, family
members, and community), (2) forgiveness, (3) civility, (4) honesty, and (5) courage.
Qeirat is a complex culture-specific set of moral values and the word itself does not have a
straight English translation, but the closest translation is honor. Literally, Qeirat means
“protecting a loved/sacred thing or person against intrusion.” Tabatabai, a well-regarded Islamic

3
It can also be transliterated as “Gheirat” (e.g., Razavi et al., 2020; Srivastava, 2020)

48

scholar, considers Qeirat a motivating “instinct” that exists in all humans. Atari, Barbaro,
Shackelford, and Chegeni (2017) adapted the concept of Qeirat as a moral psychological code
which is conceptually very close to mate retention strategies — behaviors that decrease the
likelihood of a temporary or permanent relationship defection by manipulating perceptions of
alternative mates such that they perceive those options to be unattractive or unattainable (Buss,
1988). These authors examined Qeirat in a sample of Iranian men and women and reported
strong positive correlations between mate retention behaviors and Qeirat in Iran. Thereafter,
Atari (2018) argued that Qeirat is a culture-specific moral code and that there are (at least) ten
nouns synonymous to it in the Persian language (the transliterations are Keramat [benevolence,
greatness, courteousness], Sharaf [honor, prestige], Sherafat [honor, prestige], Janam [courage],
Ehteram [respect], Hormat [sacredness, respectfulness], Izzat [group-based honor], Aberoo
[one’s reputation that should be retained at all costs], Ta’asob [honor, usually males showing it
toward females], and Heyssiat [respectability, face]; Other relevant concepts include Haya
[modesty, not engaging in promiscuous relations, humility], Effat [female honor, chastity], Esmat
[innocence], and Namoos [female kin such as mother and sister, or one’s romantic partner who
are to be protected at all costs]). These words are used in slightly different contexts. For
example, “Izzat” might be more relevant in group-level interactions (Aslani et al., 2016),
“Namoos” refers to people (e.g., mate, family member) or entities (e.g., one’s country or
religion) which are connected to the self and should be protected against norm violations
(Razavi, Shaban-Azad, & Srivastava, 2020), and finally “Qeirat” might be more closely
associated with family values and mate retention strategies (Atari et al., 2017; Atari & Jamali,
2016).

49

Interestingly, psycho-lexical studies based on Persian (also known as Farsi) lexicon and
literature have identified a mix of religiosity and Qeirat as a dimension of basic personality
structure, while failing to find other personality dimensions typically found in Western languages
such as Agreeableness or Openness to Experience (Farahani, Farzad, & Fotoohi, 2004;
Herfehdoust, Falsafinejad, Delavar, Sohrabi, & Tamimdari, 2015). Srivastava (2020) suggests
that Qeirat has no direct translation in English and is “a distinctive set of thoughts, feelings, and
actions associated with violations of Namoos, a set of people and entities one feels protective
toward” and that any personality psychological examination of Iranians without consideration of
Qeirat “might miss something important in [Iranian] context”. This underscores the pivotal role
of these moral psychological words in Farsi in describing oneself and other people.
Overview of Present Studies
Iran is one of the largest Middle Eastern countries, with a population of approximately 85
million, the majority of whom (89%), according to official reports, identify as Shiite Muslims;
however, there are many reasons to believe that this figure might be an overestimation (Tezcur,
Azadarmaki, & Bahar, 2006). For example, results of the global Gallup Poll in 2009, which
asked “Is religion important in your daily life?”, indicated that 73% of Iranians responded “yes”
(and 26% responded “no”), very close to responses of Americans (69% responded “yes” and
31% responded “no”). Co-existence of various ethnicities, subcultures, and religions, as well as
the eventful history of this country paired with currently practiced Shiite faith among the
majority of people, render Iran an interesting cultural setting for moral psychological research. In
addition, empirical research on Iranians’ moral values and their similarities and differences with
those in WEIRD populations is currently lacking. The present research aimed to address the

50

dearth of research in a series of inter-connected studies: We examine the structural validity of the
MFT typology operationalized by the MFQ (Graham et al., 2011) in Iran (Study 3.1), compare
moral foundations and their underlying network structure between Iran and the US (Study 3.2),
conduct field interviews focusing on beliefs about virtues and vices in Iran (Study 3.3), conduct
interviews about the concept and manifestations of Qeirat (Study 3.4a), broaden the nomological
network of Qeirat values in a large sample (Study 3.4b), and examine how Qeirat can explain
Islamic religiosity, political conservatism, and mate retention behaviors above and beyond Care,
Fairness, Loyalty, Authority, Purity, and honor values (Study 3.5).
Study 3.1: Assessment of Moral Foundations in Iran
Study 3.1 was designed to provide a fluent translation for the MFQ in Persian and to
examine the factor structure of this measure using both exploratory and confirmatory factor-
analytic techniques. We collected a large data set from the general population in Tehran, Iran.
We also examined how scores on moral foundations were related to religiosity, political
attitudes, socio-economic status, and demographics. It is particularly important to examine the
factor structure of WEIRD models and measures in less WEIRD countries since parent structures
may not hold (see Laajaj et al., 2019). The present study aims to provide evidence for structural
validity of the MFT typology, operationalized by MFQ.
Methods
Participants. The sample consisted of 544 individuals (52.6% male, 37.5% female, 9.9%
preferred not to report). All participants were recruited from the general population from public
places in Tehran, Iran. Participants ranged in age from 15 to 75 (M = 32.5, SD = 9.8 years). In

51

terms of highest educational qualification, 31 participants (5.7%) reported some high school
education, 104 participants (19.1%) reported having a high school diploma, 121 participants
(22.2%) reported an associate’s degree, 191 participants (35.1%) reported a bachelor’s degree,
81 participants (14.9%) reported a postgraduate degree, and 8 participants (1.5%) reported a
doctoral degree. Eight participants did not report their highest educational qualification.
Measures. All participants completed the following self-report measures.
Moral Foundations Questionnaire (MFQ). Participants completed the 30-item MFQ
(Graham et al., 2011) which consists of two 15-item sections, namely Relevance and
Judgments. The first section measures the relevance individuals ascribe to each of the
foundations. Items on the Relevance section are rated along a 6-point Likert-type scale ranging
from 0 (Not at all relevant) to 5 (Extremely relevant). The Judgments section consists of
contextualized items that can gauge actual moral judgments related to the five moral
foundations. Items on the Judgments section are rated along a 6-point Likert-type scale ranging
from 0 (Strongly disagree) to 5 (Strongly agree). See Procedure for our translation process.
Political Orientation. Participants rated their affiliation with the rightist political party in
Iran (“Osoolgara”) as opposed to the leftist party (“Eslah-talab”) along a 7-point scale ranging
from 1 (Very Eslah-talab) to 7 (Very Osoolgara). Another item asked participants to rate their
political conservatism on a scale ranging from 1 (Very liberal) to 7 (Very conservative). We
averaged these two items in order to achieve a political orientation score where higher scores
indicate more conservative political orientation. Previous studies have used a similar method for
assessment of political orientation (Jost & Thompson, 2000).
Self-Rating of Religiosity (SRR). The SRR is a single-item measure of intrinsic religiosity
(Abdel-Khalek, 2007). Participants responded to the question, “What is your level of religiosity

52

in general?” — as translated and validated by Afhami et al. (2017) — on an 11-point scale
ranging from 0 (indicating no religiosity) to 10 (indicating high level of religiosity). Although
single-item measures are limiting in terms of breadth, single-item measures of religiosity have
demonstrated desirable psychometric properties in various samples and have been reliably used
in Iran (Afhami et al., 2017).
Procedure. All MFQ items were translated into Persian using the standard back-
translation technique (Brislin, 1970). Specifically, the first author translated the items into
Persian from the original English version, then two independent translators unaffiliated with this
study translated the items back into English. Small differences that emerged during the
translation process were resolved between the translators, resulting in the final Persian version of
the scale that was administered on the current sample. Research assistants approached potential
participants and invited them to take part in a psychological research about “values and beliefs.”
The order of presentation of all scales in the study was pre-randomized. All participants
completed a set of paper-and-pencil measures after giving oral informed consent to take part in
the study. Participants were not compensated, but were provided with debriefing information
upon completion of the questionnaire.
Data Analysis. Following the recommendations regarding sample size in factor analytic
studies (Tabachnick & Fidell, 2013), we aimed to recruit 300 participants for exploratory factor
analysis (participant to item ratio = 10). First, we randomly split the data into training (n = 300)
and test (n = 244) data sets for exploratory and confirmatory factor analyses. To examine the
factor structure of the MFQ, we computed principal-axis exploratory factor analysis (EFA) using
Kaiser normalization and Promax rotation on the training data using the Psych package (Revelle,
2017) in R (version 3.6.1). EFA is the appropriate method of data reduction when the aim is to

53

explore the possible underlying structure of a variable in the absence of a predetermined model.
We used parallel analysis (Hayton, Allen, & Scarpello, 2004) to determine the number of factors
to be retained in the EFA. Then, we fit several Confirmatory Factor Analyses (CFAs) models
using the lavaan package (Rosseel, 2012) on the test data (n = 244) to compare alternative
structures underlying the Persian MFQ. We fit the models with maximum likelihood estimation
with robust (Huber-White) standard errors. We assessed model fit using the robust chi-square
index (χ
2
), robust normed chi-square (χ
2
/df), robust root mean square error of approximation
(RMSEA), robust comparative fit index (CFI), robust Tucker-Lewis index (TLI), Akaike
information criterion (AIC), and sample-size adjusted Bayes information index (BIC). Normed
chi-square (χ
2
/df) values of < 3.00 indicate good fit (Hu & Bentler, 1999). The RMSEA index
and its 90% confidence interval (90%CI) provide a correction for model complexity where
values lower than .06 indicate very good fit and values between .06 and .10 indicate mediocre fit
(Hu & Bentler, 1999). The CFI and TLI values of .90 and higher indicate good fit for the model.
AIC and sample-size-adjusted BIC are usually used when comparing multiple models and lower
values represent better fit.
Results
Exploratory Factor Analysis. We conducted EFA on the training sample, n = 300,
which satisfies the commonly recommended participant-to-item ratio of 10:1 (Tabachnick &
Fidell, 2013). Bartlett’s test of sphericity, χ
2
(435) = 5121.676, p < .001, and the Kaiser-Meyer-
Olkin measure of sampling adequacy, KMO = .90, exceeded minimum criteria that should be
met before EFA can be conducted. The scree plot and results of parallel analysis suggested that
five factors should be extracted. These five factors explained 54% of the total variance. As can
be seen in Table 3.2, the extracted factors are not readily interpretable. For example, the fifth

54

factor includes items that originally belong to Fairness, Care, and Loyalty, thus any aggregate
score of this factor cannot be meaningfully interpreted.
Table 3.2
Exploratory Factor Analysis of the MFQ Items (n = 300; Study 3.1)
Item (Original
Factor)
PA1 PA2 PA3 PA4 PA5
justice (F) 0.80 -0.06 0.10 -0.14 0.03
animal (C) 0.78 -0.05 0.21 -0.02 -0.08
fairly (F) 0.76 0.05 -0.10 0.11 0.03
compassion (C) 0.62 0.12 -0.15 0.09 0.18
sexroles (A) 0.61 0.23 0.10 -0.09 -0.05
family (L) 0.60 0.17 0.08 0.02 -0.21
history (L) 0.60 0.13 -0.02 0.06 0.02
harmlessdg (P) 0.56 0.14 0.14 0.03 0.00
soldier (A) 0.07 0.68 -0.17 0.00 0.14
team (L) 0.24 0.64 -0.11 -0.09 0.24
chastity (P) 0.19 0.60 -0.06 0.01 0.06
unnatural (P) 0.19 0.57 0.14 -0.01 -0.21
rich (F) 0.07 0.53 -0.13 -0.04 0.00
traditions (A) -0.35 0.48 0.40 0.03 0.04
kill (C) 0.31 0.47 0.08 -0.15 -0.04
kidrespect (A) 0.35 0.38 -0.06 0.20 -0.11
cruel (C) 0.12 -0.11 0.85 -0.02 -0.07
disgusting (P) -0.05 0.02 0.73 0.15 -0.17
rights (F) 0.22 -0.17 0.72 -0.12 0.14
loyalty (L) 0.26 -0.19 0.52 -0.01 0.19
chaos (A) 0.08 -0.12 0.44 0.10 0.25
god (P) -0.03 0.27 0.29 0.11 0.11
treated (F) 0.16 -0.13 -0.08 0.82 0.07
emotionally (C) 0.21 -0.16 -0.02 0.74 0.03
respect (A) -0.19 0.12 0.02 0.73 -0.09
lovecountry (L) -0.01 0.00 0.03 0.72 -0.01
decency (P) -0.03 0.00 0.16 0.45 0.09
unfairly (F) -0.05 0.03 -0.01 -0.02 0.91

55

weak (C) -0.07 0.03 0.05 0.20 0.62
betray (L) -0.01 0.17 0.27 -0.07 0.52
Note. Loadings corresponding to each factor are presented in bold. C = Care; F =
Fairness; L = Loyalty; A = Authority; P = Purity.

Confirmatory Factor Analysis. We conducted five distinct CFAs on the test data. We
specifically tested the following models: (1) the 1-factor model with a single latent variable, (2)
the 2-factor model which integrates Care and Fairness as the first factor while integrating the
Loyalty, Authority, and Purity factors as the second factor, (3) the 3-factor model which consists
of a Care-Fairness factor, a Loyalty-Authority factor, and a Purity factor, (4) the 5-factor model
originally proposed by Graham et al. (2011), (5) the model derived from our EFA in this study.
All fit indices are presented in Table 3.3. Comparing alternative structures suggests that MFQ
items are best reduced to five factors as found in the EFA. Of note, the original five-factor model
of Graham et al. (2011) shows the best fit indices of the first four models (see Table 3.3). We did
not apply modification to these models based on the modification indices available through co-
varying error terms in the structural model. Although the EFA-based model provided better fit
indices in CFAs, we chose to continue our analyses using the original five-factor model proposed
by Graham et al. (2011) as the EFA-based model seems difficult to interpret.

56

Table 3.3
Goodness-of-Fit Indices of Structural Models Underlying MFQ (n = 244; Study 3.1)
Model χ2Robust χ
2
/df CFIRobust TLIRobust AIC BICCorrected RMSEA
Robust
90%CI
1-factor model 1173.61 2.90 .738 .718 19442.63 19453.91 .112 [.105,
.120]
2-factor model 1167.87 2.89 .739 .719 19437.59 19449.06 .112 [.104,
.119]
3-factor model 1166.79 2.90 .739 .718 19437.97 19449.81 .112 [.105,
.120]
5-factor model 1101.20 2.79 .760 .736 19353.81 19366.96 .108 [.101,
.116]
EFA model 744.04 1.88 .883 .871 18840.86 18854.01 .076 [.067,
.084]
Note. CFI = comparative fit index; TLI = Tucker-Lewis index; AIC = Akaike information criterion; BIC
= Bayes information criterion; RMSEA = root mean square error of approximation; CI = confidence
interval.
Concurrent Validity. The Spearman partial correlation coefficients between moral
foundations and political conservatism, religiosity, educational qualification, socio-economic
status, and age are summarized in Table 3.4. As can be seen, after controlling for all other
variables, Care was negatively correlated with political conservatism (partial ρ = -.11, p = .021).
Authority was positively correlated with political conservatism (partial ρ = .11, p = .019).
Religiosity was positively associated with higher Purity concerns (partial ρ = .09, p = .042).
More educated individuals reported slightly higher levels of Loyalty (partial ρ = .10, p = .027).
Notably, SES was not related to any moral foundations. In terms of differences between women
and men, sex differences were non-significant for Care (t = 1.76, Welch-corrected df = 425.04, p
= .079, Cohen’s d = 0.16), Fairness (t = 1.93, Welch-corrected df = 416.59, p = .055, Cohen’s d =
0.18), Loyalty (t = 1.71, Welch-corrected df = 437.35, p = .088, Cohen’s d = 0.16), Authority (t
= 0.06, Welch-corrected df = 457.55, p = .950, Cohen’s d = 0.01), and Purity (t = 1.06, Welch-
corrected df = 439.19, p = .289, Cohen’s d = 0.10).

57

Table 3.4
Partial Spearman Correlation Coefficients among Variables (Study 3.1)

Care Fairness Loyalty Authority Purity
Conservatism -0.11* 0.03 -0.07 0.11* 0.06
Religiosity -0.07 -0.04 -0.01 0.04 0.09*
Education -0.04 0.03 0.10* -0.06 -0.01
SES 0.05 -0.01 -0.02 0.03 0.00
Age 0.03 -0.03 -0.03 0.03 0.06
Note. *p < .05
Internal Consistency. The internal consistency coefficients for the five foundations were:
Care: .78, 95%CI [.75, .81]; Fairness: .73, 95%CI [.69, .76]; Loyalty: .76, 95%CI [.73, .79];
Authority: .74, 95%CI [.70, .77]; Purity: .78, 95%CI [.75, .81]. All internal consistency
coefficients fell within the widely recommended threshold of .70 (Tavakol & Dennick, 2011).
Discussion
In this study, we provided a new translation of the MFQ in Persian. We found in our
exploratory factor analysis, that the Persian MFQ’s structure is different from the original five-
factor model found by Graham et al. (2011). Confirmatory factor analyses suggested that neither
of the models reached satisfactory fit indices recommended in the literature, so the above-
mentioned findings indicate that the factor structure of the Persian MFQ is problematic and fails
to replicate the structures typically found in WEIRD populations (Davies, Sibley, & Liu, 2014;
Graham et al., 2011; Nilsson & Erlandsson, 2015; Métayer & Pahlaven, 2014; van Leeuwen &
Park, 2009). Of note, a recent study using another Persian translation of the MFQ (Nejat &
Hatami, 2019) showed that the factor structure of the Persian MFQ was not consistent with the
original five-factor model.

58

The associations between MFQ subscales and related constructs were mixed, but in large
part the Persian MFQ does not fully replicate the previously established relationships between
political ideology, religiosity and moral foundations (Graham et al., 2009), further suggesting
that Persian MFQ should be used with great caution until more suitable models of moral values
or alternative measures of MFT are developed. Notably, reasonable foundation-level internal
consistency coefficients provide some evidence that foundation-level scores are not just noise;
there is some meaningful association between theoretically related items, however, their sum
scores should be treated with caution.
Study 3.2: Moral Foundations in Iran and the U.S.
Development of the MFQ was originally based on factor analytic approaches (Graham et
al., 2011), and network psychometric methods have not been used in this literature. It has been
recently suggested that the traditional reliance on latent variable models (as in Study 3.1) may be
inadequate, because the bulk of the structural covariation in psychometric scales can be a
consequence of interactions between the items in psychometric tests (Costantini et al., 2015).
Here, we compared moral values – defined by the MFT – between two samples from Iran and the
US, and compare the psychometric network of the moral domain, as measured by the MFQ, in
these cultures. In Study 3.2, we compared the structure and content of moral foundations
between Iran and the US, a highly WEIRD society. We went beyond classical test theory and
factor analytic approaches, used in Study 3.1, by examining the network of items and
foundations, and how they influence each other. In other words, we ask which moral foundations
are “central” in moral domain networks in Iran vs. the US. We investigate which specific moral

59

values are most central by modeling the moral domain as a network of interconnected
foundations.
Methods
Participants. We recruited 247 participants (127 women, 120 men) in public places in
Tehran. Participants completed a set of paper-and-pencil questionnaires including MFQ (internal
consistency coefficients: Care: .60; Fairness: .54; Loyalty: .65; Authority: .70; Purity: .79). We
also recruited 300 participants (152 women, 148 men) who completed the MFQ on a crowd-
sourcing website in the US (yourmorals.org) (internal consistency coefficients: Care: .69;
Fairness: .63; Loyalty: .69; Authority: .76; Purity: .86). Both data sets (overall N = 547) were
collected in 2017. Iranian participants were not compensated. American participants were given
feedback about their moral values and political ideology.
Measures. All respondents in the Iranian sample completed the Persian translation of the
30-item MFQ developed in Study 3.1. The Iranian participants completed other individual
difference measures reported elsewhere. The American sample completed the English version of
the 30-item MFQ, along with other individual difference measures reported elsewhere. All items
in both samples are rated along a 6-point Likert-type scale coded from 0 to 5.
Procedure. To recruit Iranian participants, research assistants approached potential
participants in public areas and invited them to take part in psychological survey research. The
order of presentation of all measures was randomized. All participants completed a set of paper-
and-pencil measures after giving oral consent to participate in the study. All Iranian participants
were provided with debriefing information upon completion of the measures. American
participants were recruited through a crowd-sourcing website and completed several measures

60

related to moral and political psychology. All participants provided consent that their data would
be used for research purposes.
Data Analysis. Since missing data should be handled prior to psychometric network
analysis, we first conducted multiple imputation using the mice package (Buuren & Groothuis-
Oudshoorn, 2010) in R version 3.4.2 (R Core Team, 2017). We used the predictive mean
matching (PMM) algorithm and 50 iterations for managing missing values in both Iranian (1.5%)
and American (0.5%) data sets. After managing missing values in both data sets, they were
subjected to network analysis which treats the MFQ as a network of mutually interacting items.
Here, items correspond to nodes and their interrelation serves as edges using a Gaussian
Graphical Model (GGM) estimation. In the estimated networks, an edge indicates a nonzero
partial correlation between two nodes, while controlling for all other nodes in the network. In
other words, two connected items show a level of covariation that cannot be explained by their
relations to other items. In order to control for spurious connections that may arise due to
multiple testing, and to control for the computational size of the problem, we applied the
adaptive LASSO algorithm which assigns different penalty weights for different coefficients
with cross-validation (Zou, 2006).
We used parcor (Krämer, Schäfer, & Boulesteix, 2009) and qgraph (Epskamp, Cramer,
Waldorp, Schmittmann, & Borsboom, 2012) R packages to estimate matrix of partial correlations
based on LASSO regularization and to analyze and visualize weighted networks, respectively.
Notably, the adaptive LASSO yields a more parsimonious graph as it pushes a large number of
weighted edges to be exactly zero, so only the most important relations are shown in the final
network. In network models, each node also plays a different role and has different
characteristics. To analyze the place and function of items within the network, we used the

61

strength centrality index. A highly central node in a given network is one from which it is easy to
travel to other nodes in the network. In addition, centrality in networks identifies the most
important nodes in a network. Such items in a psychometric scale may correspond to
characteristics that have a particularly important function related to the construct being
measured. For example, the MFQ item “chastity is an important and valuable virtue” may be
related to different moral phenomena in Iran vs. the US.
Two independent networks were built using data from Iran and the US. We checked for
differences in these network structures by performing a permutation test using the
NetworkComparisonTestR package (van Borkulo, Epskamp, & Millner, 2016). Difference has
been defined as the deviation in absolute weighted sum scores of the connections. We randomly
regrouped participants from the Iranian sample and the American sample repeatedly (1000 times)
and calculated the differences between these subsamples (van Borkulo et al., 2016). The
resulting distribution is used to test the observed difference of the original subsamples. Both
weighted networks (i.e., using edges’ weights) and unweighted network structures (i.e., the
presence of an edge between two nodes) were tested. Notably, the latter is tested to examine if
the basic structure of the samples are similar, whereas the first investigates whether the strength
of individual connections in the network structures are similar.
Results
Cross-Cultural Differences. We ran a measurement invariance analysis to examine
whether respondents from Iran and the US interpret the MFQ items in a conceptually similar
way. We used multi-group CFA with maximum likelihood estimation using robust (Huber-
White) standard errors and a scaled test statistic that is (asymptotically) equal to the Yuan-
Bentler test statistic. As can be seen in Table 3.5, Iranians and Americans do not interpret MFQ

62

items in nearly similar ways, and all fit indices are substantially lower than recommended
thresholds for invariance testing (Chen, 2008). Therefore, raw scale means cannot be
meaningfully compared. In particular, item intercepts are substantially different across these
cultures, causing the scalar model to fare much worse than the metric model in Table 3.5.
Distributions of moral foundations in both cultures is visualized in Figure 3.1. After setting both
loadings and intercepts to be equal across these groups, the scalar model estimated factor scores
for both countries. The Iranian sample’s estimates of Fairness (B = 0.35, SE = 0.09, Z = 2.70, p =
.007), Loyalty (B = 1.42, SE = 0.09, Z = 15.69, p < .001), Authority (B = 0.79, SE = 0.09, Z =
8.75, p < .001), and Purity (B = 1.35, SE = 0.09, Z = 15.78, p < .001) were higher than those of
their American counterpart; however, estimated factor score of Care (B = 0.18, SE = 0.25, Z =
0.71, p = .476) was not significantly different across groups.
Table 3.5
Results of Measurement Invariance Models (Study 3.2)
Model χ2Robust df ∆df CFI ∆CFI RMSEA ∆RMSEA p BIC
1. Configural 2637.30 790 -- 0.713 -- 0.082 -- -- 52221
2. Metric 2951.80 815 25 0.670 0.043 0.087 0.005 < .001 52378
3. Scalar 4252.20 840 25 0.441 0.229 0.111 0.024 < .001 53521
Figure 3.1.
Violin plots of cultural differences in moral foundations (Study 3.2).

63

Item Network Analysis. Figures 2 and 3 graphically present networks of the MFQ items
for the Iranian and American samples, respectively. As mentioned, edges in these networks
represent partial correlations between items while statistically controlling for all other items.
Positive correlations are represented as green edges and negative correlations are shown in red.
The width of the edges corresponds to the magnitude of the partial correlation in the network
matrix. Out of 435 possible edges, the Iranian network had 74 non-zero edges (17.0%) with 66
positive edges and 8 negative ones. The American network had 93 nonzero edges (21.4%) with
81 positive edges and 12 negative edges. In both networks, items that belong to the same
foundation (the same color in the Figures) tend to be more closely connected; however, there are
some strong edges between items from different foundations (e.g., the “loyalty” item to the
“unfairly” item in the Iranian network). In contrast, some items are theoretically related, but are
not linked in these networks because their relationship can be explained by their relations to
other items. For example, the items “team” and “history” are both included in the Loyalty
foundation, but are not linked in the American nor the Iranian network. This highlights how
network analysis can advance our understanding of how morally relevant phenomena are situated
within the moral domain network while controlling for all other measured phenomena.

64

Figure 3.2.
Item-level network of moral foundations in Iran (Study 3.2).
Note. Positive correlations are represented as green edges and negative correlations are shown
in red.

The item-level networks from the two samples were compared based on 1000 iterations to
create a reference distribution. We assessed the difference (Diff) between two networks based on
several invariance measures (i.e., network structure invariance, global strength invariance, edge
invariance). Network structures were estimated with l1-regularized partial correlations for
weighted networks. The structure of the observed networks was statistically significantly
different (Diff = 0.406, p < .001). The global strength of the Iranian item-level network (Global

65

Strength = 12.52) and that of the American network (Global Strength = 12.24) were not different
(Diff = 0.286, p = .772).
Figure 3.3.
Item-level network of moral foundations in the US (Study 3.2).
Note. Positive correlations are represented as green edges and negative correlations are shown in
red.

66

Foundation Network Analysis. Figures 4 and 5 visualize the foundation-level networks
for the Iranian and American samples. There are five nodes in these networks and the maximum
number of edges is 10 in both. The Iranian network had 7 edges (70.0%; 7 positive and 0
negative). The American network had 4 edges (40.0%; 4 positive and 0 negative). In both
networks, Care and Fairness are closely linked, while Loyalty, Authority, and Purity are linked
together. Interestingly, the network of moral foundations in Iran is more intertwined between
individualizing and binding foundation, whereas that of the American sample is two discrete
subgraphs including one individualizing and one binding. This shows that individualizing and
binding values are more closely intertwined with one another in Iran compared to the American
culture where individualizing and binding values are relatively independent networks of moral
values. Loyalty in particular, seems to be the most central moral foundation in Iran which might
be indicative of the value of close family ties, in-group coalitions, and love of the country. In the
US, on the other hand, authority is the most central moral foundation.

67

Figure 3.4.
Foundation-level network of moral foundations in Iran (Study 3.2).

68

We also compared the foundation-level network structures between Iran and the US using
1000 iterations. Comparing weighted networks, the network structures were significantly
different (Diff = 0.27, p = .003). In terms of global strength, the Iranian network’s global strength
(2.23) was significantly higher than that of the American sample (1.69), p < .001. Therefore,
based on foundation level of analysis, the structure and strength of moral networks are different
in Iran and the US.
Discussion
Study 3.2 provided evidence for a new approach for modeling moral foundations using a cross-
cultural sample from two sites. We modeled the Persian MFQ and the English MFQ networks at
item- and foundation-levels. In all networks, emerging edges that survive the Lasso penalization
are likely to have causal relations (Borsboom & Cramer, 2013). One of the MFT-consistent
findings of our network analyses is that Iran’s network showed a more densely interconnected
network. It may be concluded that endorsing different foundations are more closely intertwined
in Iran, with Loyalty being the most central foundation possibly due to stronger social ties, larger
kin networks, higher ingroup values in Iran. In the US, in contrast, individualizing and binding
values are completely segregated as shown in our network analysis, a finding that has been
concealed in correlational research in the literature. Findings of Study 3.2 help explain why
previous research on WEIRD samples found a relatively clean two-factor solution in exploratory
factor analysis, while we did not find this in the Iranian sample in Study 3.1. These findings
further call for contextualized studies of moral values in Iran using an engaged and diverse set of

69

methodologies, rather than top-down examination of the moral domain using WEIRD models of
moral cognition, presumably developed by WEIRD researchers (see Medin, 2017).
Study 3.3: Iranians’ Beliefs about Virtues and Vices
Studies 1 and 2 pointed to the fact that MFQ might not adequately capture the moral
domain in Iran, and that more nuanced exploration of moral values in Iran is warranted. Here, in
Study 3.3, we conduct an interview-based qualitative study, asking a sample of Iranians what
virtues and vices matter to them most and who their “moral heroes” are. Studies 3.1 and 3.2 used
MFT as a theoretical framework for studying moral concerns in Iran and comparing these moral
concerns to those of a US sample. While important, these results are bound to theoretical
assumptions of MFT (see Graham et al., 2013), and are unable to identify culture-specific (or
language-specific) emergent moral values, or how frequently particular moral concerns are
implicated in Iran. In fact, MFT predicts and encourages cultural studies of moral values, which
may not have been operationalized in the MFQ (Graham et al, 2013; Iyer, Koleva, Graham,
Ditto, & Haidt, 2012). Study 3.3 aims to close this gap and potentially uncover culture-specific
moral values in Iran.
Methods
We interviewed 43 individuals (15 women, 27 men, 1 non-binary) in Tehran and asked
them about their beliefs about moral virtues and vices. The mean age of the sample was 35.10
years (min = 23, max = 60, SD = 12.03). Participants were approached in public places (public
libraries, universities, mosques, and shopping malls) and were asked whether they would be
interested in taking part in a study about moral values. After agreeing to participate, a researcher

70

asked five questions and wrote down participants’ answers. No recording device was used. These
questions were: (1) what do you believe to be the most important moral virtues?; (2) what do you
believe to be the most important moral vices?; (3) name two of your “moral heroes”; (4) try to
think of “moral” acts you personally did and explain what they were generally about; (5) try to
think of “immoral” acts you personally did and explain what they were generally about. In cases
where the questions were not fully comprehended by participants, the researcher elucidated the
question, making sure that the question was fully understood. The interviews were not timed, but
they typically took about 10 to 20 minutes to complete. Participants were assured about the
confidentiality and anonymity of their data and that their responses would only be used for
research purposes. No compensation was made to participants.
Results
We content-analyzed the moral and immoral acts and broke them into themes (Teddlie &
Tashakkori, 2009). For example, “giving money to somebody who really needed some money”
was categorized as “prosociality/helping” and “cheated on my boyfriend” was categorized as
“infidelity/cheating.” These themes were then combined with the virtues and vices extracted in
questions 1 and 2. The first ten moral acts/virtues and immoral acts/vices are presented in Table
3.6. Participants’ answers to the “moral hero” question were sparsely distributed. Eighteen
participants (40.9%) regarded one of their friends as their moral hero, followed by their mothers
(29.5%), colleagues (20.5%), teachers (15.9%), and fathers (13.6%).

71

Table 3.6
Virtues and Vices Identified in Study 3.3
Virtues (Number of occurrences) Vices (number of occurrences)
Honesty/Truthfulness (33) Lying/Distrustfulness (33)
Prosociality/Helping (29) Insincerity/Hypocrisy (23)
Affability/Kindness (13) Cheating/Infidelity (18)
Loyalty/Fidelity (11) Aggression/Violence (14)
Humility/Modesty (9) Self-centeredness/Hubris (9)
Responsibility/Authority (9) Theft/Unlawfulness (8)
Sanctity/‘Aberoo’ (8) Impoliteness/Rudeness (8)
Respectfulness/Politeness (8) Harm/Manipulation (6)
Fairness/Justice (8) Degradation/Promiscuity (6)
Forgiveness/Compassion (8) Greed/Avarice (4)

Discussion
Haidt and Joseph (2008) suggested that “the first step in mapping the moral domain of any
culture, […] should therefore be to list and count the norms that get the most attention.” Study
3.3 was motivated by this suggestion and follows up Studies 1 and 2 where we found evidence
that the MFQ did not capture moral values well in Iranian culture. Study 3.3’s qualitative
approach can give us insights on what norms and values Iranians care about in a bottom-up
fashion. Although some of these values exist in the MFT typology, some are missing. Moral
values such as “honesty” and “politeness” have been identified in the literature (e.g., Landy &
Bartels, 2018), but we also found novel values that are specific to the Iranian culture.
Specifically, “Aberoo” and related terms (e.g., “Haya”) about the noxiousness of promiscuity —
all under the tent of Qeirat — refer to honor, dignity, and reputation. Some of these values can be

72

considered more masculine (e.g., “Ta’asob”), some can be considered more feminine (e.g.,
“Haya”), and finally some of them are less gendered (e.g., “Aberoo”). However, since previous
work has not examined these constructs in Iranian culture (or in other Persian-speaking nations
such as Afghanistan), their linguistic features and psychological implications are not well
understood (for an exception see Razavi et al., 2020). Atari (2018) collectively referred to these
values as “Qeirat” values, highly preferred in mate selection preferences in Iran, and strongly
correlated with benefit-provisioning and cost-inflicting mate retention strategies. Razavi and
colleagues (2020) examined the emotional aspects of Qeirat and found that violations of Qeirat
values (e.g., Namoos) will evoke strong emotional reactions in Iranian culture. Still, the structure
and utility of Qeirat values remains unclear.
It is evident that participants in the present sample mostly reported people in their close
social circles as their moral hero. Moral heroes have a strong moral character, which includes
compelling moral reasoning, personal development, and a heart-warming life story (Frimer,
2018; McAdams & Guo, 2015). When asked to name moral heroes, Americans usually list
Martin Luther King, Jr., Nelson Mandela, Mahatma Gandhi, and John F. Kennedy (Frimer &
Sinclair, 2016), but Iranians in the present sample mostly listed people they closely knew (e.g.,
friends, family members) with few people listing figures like Mahatma Gandhi. It is worth noting
that one of the limitations of Study 3.3 is lack of parallel studies in WEIRD societies (see
Graham, Meindl, Koleva, Iyer, & Johnson, 2015). Hence, while we can rely on previous work on
moral heroes in the US and other nations (e.g., Frimer & Sinclair, 2016), no direct comparison
can be made.

73

Figure 3.5.
Foundation-level network of moral foundations in the US (Study 3.2)

Study 3.4: Extending the Nomological Network of Qeirat Values
As found in Study 3.3, Qeirat (and its related concepts) seems to be a salient cultural-moral
system in Iran which has not been widely researched in the past. In Study 3.4a, we conducted
qualitative interviews to identify this concept’s definition, distinctiveness, and manifestations in
Iranian culture. In Study 3.4b, we developed a valid and reliable measure of Qeirat values, and
expanded the nomological network of this concept in an Iranian sample. Specifically, we factor
analyzed the 24 Qeirat values indicators found in Study 3.4a, and then examined how moral
foundations, the Big Five personality dimensions, religiosity, and happiness are related to scores

74

on this newly developed scale. Our main goal in Study 3.4b was to extend the nomological
network (series of connected theoretical concepts and observable properties that give the
constructs particular meaning) of this newly-uncovered culture-specific moral system.
Methods
Participants. In Study 3.4a, a community sample of 205 participants (54.1% women) was
selected from Tehran. Potential participants were approached by a researcher in public places in
Tehran and were invited to complete a set of questionnaires (some reported elsewhere; Chegeni,
Pirkalani, & Dehshiri, 2018), including an open-ended question about the concept of Qeirat.
Participants ranged in age from 19 to 61 (M = 31.9, SD = 7.9).
In Study 3.4b, following the recommendations regarding sample size in factor analytic
studies, we aimed to recruit around 500 participants for exploratory and confirmatory factor
analysis (Tabachnick & Fidell, 2013). Participants were 519 individuals recruited in university
settings in Tehran. Eighteen participants left at least one of the target measures blank and were
thus excluded. Therefore, the final sample consisted of 501 participants (54.7% women). The
participants ranged in age from 18 to 45 (M = 25.3, SD = 5.0). Around half of the sample
(46.7%) were undergraduate students and the remainder were master’s (45.1%) and doctoral
students (6.8%). Seven participants (1.4%) chose not to disclose their educational qualification.
Measures. All participants completed the following self-report measures.
Qeirat Values Scale (QVS). We content-analyzed the open-ended questions in Study 3.4a
and re-phrased responses into declarative sentences appropriate for survey studies. Overall, we
identified 24 sentences, shown in Table 3.7. In Study 3.4b, participants completed the 24-item
Qeirat Values Scale (QVS) developed based on responses in Study 43.a. Items were rated along a

75

7-point Likert-type scale ranging from 1 (Strongly disagree) to 7 (Strongly agree). The factor
structure and internal consistency coefficients of the scale are presented in Results.
Moral Foundations Questionnaire (MFQ). Participants completed the Persian version
of the 30-item MFQ (Graham et al., 2011) reported in Study 3.1. The internal consistency
coefficients were .65, .69, .68, .74, and .81 for Care, Fairness, Loyalty, Authority, and Purity,
respectively.
Political Orientation. Participants rated their political affiliation as in Study 3.1. The
internal consistency coefficient of this two-item measure was relatively high in the current
sample (Cronbach’s α = .68).
Self-Rating of Happiness. We used a single-item measure of happiness (i.e., Do you feel
happy in general?), developed by Abdel-Khalek (2006), rated on an 11-point scale ranging from
0 to 10. The Persian versions of single-item measures of happiness have previously been used in
Iran (Farzianpour et al., 2011).
Iranian National Religiosity Scale (INRS). Participants completed a 36-item measure of
Islamic religiosity developed and validated by Khodayarifard et al. (2010). These authors
developed a larger item pool based on Quranic teachings and Hadith, and reduced the number of
items using factor analysis in national samples from Iran. This measure is the only locally
developed measure of Islamic religiosity in Iran. The developers of the scale have suggested that
a single score be computed by averaging all 36 items (Khodayarifard et al., 2018). Cronbach’s α
in the present sample was .96.
Ten-Item Personality Inventory (TIPI). The TIPI (Gosling, Rentfrow, & Swann Jr, 2003)
is an ultra-short measure of the Big Five dimensions of personality. Each of the Big Five

76

dimensions (i.e., Extraversion, Agreeableness, Conscientiousness, Emotional Stability, and
Openness to Experience) are measured with two items. All items are rated along a 7-point Likert
type scale ranging from 1 (Strongly disagree) to 7 (Strongly agree). The Persian version of the
TIPI has been translated and used by Atari, Barbaro, Sela, Shackelford, and Chegeni (2017).
Procedure. Potential participants were invited to take part in a psychological study in
university settings. Participation was on a voluntary basis and participants were not
compensated. All participants who agreed to take part were provided with a set of paper-and-
pencil measures right after they provided oral consent. All instruments were presented in a pre-
randomized order. All participants were verbally debriefed once all questionnaires had been
returned.
Data Analysis. Our analytic method in this study resembled that of Study 3.1. We
partitioned Study 3.4b’s data into training and test data sets for EFA and CFA, respectively. We
used Pearson correlations to examine the associations between study variables. Finally, we ran
five linear regression models to predict Qeirat values overall scores as well as the QVS
subscales.
Results
Study 3.4a. As mentioned in our Methods, we content-analyzed the open-ended questions
and organized responses into declarative sentences in order to be used in Study 3.4b. Our
qualitative method resembled that of Study 3.3. After merging similar responses into themes, we
identified 24 distinguishable sentences. These items can be seen in Table 3.7.

77

Study 3.4b: Factor Analysis of Qeirat Values. Similar to our analytic procedure in Study
3.1, we ran an EFA on a subsample of n = 300 and a CFA on the remaining data (n = 201).
Results of the EFA are presented in Table 3.7 (Bartlett’s χ
2
(276) = 5353.39, p < .001, KMO =
.94). We labeled the four factors as “Commitment to Family”, “Protecting Namoos”, “Love of
Country”, and “Supporting Family/Mate.” Results of the CFA on the test subsample suggested
that the four-factor model (χ
2
Robust = 543.82, χ
2
/df = 2.21, CFIRobust = .852, TLIRobust = .834, AIC =
13731.05, BICCorrected = 13735.62, RMSEARobust = .093, 90%CI = [.083, .104]) fit the data
substantially better than the uni-dimensional model (χ
2
Robust = 894.70, χ
2
/df = 3.55, CFIRobust =
.656, TLIRobust = .623, AIC = 14296.63, BICCorrected = 14300.69, RMSEARobust = .140, 90%CI =
[.131, .151]), supporting the multidimensionality of this scale. The correlation coefficients
between these four factors are presented in Table 3.8.
Table 3.7
Exploratory Factor Analysis of Qeirat Items (n = 300; Study 3.4b)
Variable PA1 PA2 PA3 PA4
A man should love his family. 0.83 0.09 0.24 0.16
A man should feel committed toward his family. 0.82 0.10 0.30 0.15
A man should protect his family against any dangers. 0.76 0.23 0.16 0.23
A man should be committed to his romantic partner in every possible way. 0.69 0.18 0.16 0.28
A husband is responsible for protecting his wife. 0.60 0.09 0.38 0.36
A man should ensure the happiness and peace of his family. 0.59 0.33 0.23 0.36
A man should not let anything bad happen to his family. 0.57 0.37 0.15 0.32
A man should treat others’ Namoos like his own. 0.43 0.30 0.35 0.18
A man should be vigilant about his romantic partner’s social behavior. 0.09 0.82 0.15 0.14
A man should be vigilant about his romantic partner’s interactions with other men. 0.08 0.77 0.13 0.10
A wife should dress in a way her husband approves. 0.10 0.64 0.15 0.05
A real man has Qeirat. 0.26 0.64 0.28 0.34
A man should guard against other men’s interest in his romantic partner. 0.33 0.49 0.13 0.34

78

A woman should dress modestly. 0.14 0.46 0.36 0.10
A man should react if somebody shows sexual interest in his Namoos. 0.19 0.44 0.23 0.12
Everyone should love their own country. 0.21 0.30 0.82 0.17
People should do their best to make their country proud. 0.27 0.24 0.76 0.15
Everyone should defend their country, if called upon. 0.21 0.32 0.75 0.16
Everyone should be happy when a compatriot wins in an international competition. 0.31 0.19 0.58 0.37
Everyone should be happy when their country advances in science and technology. 0.35 0.15 0.56 0.40
A man should be responsive to his family’s needs. 0.45 0.22 0.24 0.75
A real man always wants to make his romantic partner happy. 0.40 0.15 0.28 0.68
A man should be willing to endanger his own life to protect his family. 0.30 0.33 0.20 0.45
A man should be supportive of and loyal to his family. 0.29 0.26 0.28 0.40
Cronbach’s α 0.90 0.85 0.91 0.80
Note. Loadings corresponding to each factor are presented in bold. These are translated items of
the QVS from Persian.

79

Table 3.8
Zero-level Correlation Coefficients between Variables (Study 3.4b)
Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
1. Qeirat:
Commitment to
Family
1

2. Qeirat:
Protecting
Namoos
0.52 1

3. Qeirat: Love of
Country
0.64 0.57 1

4. Qeirat:
Supporting
Family/Mate
0.72 0.56 0.65 1

5. Overall Qeirat
Values
0.83 0.82 0.86 0.85 1

6. Islamic
Religiosity
0.31 0.55 0.45 0.29 0.5 1

7. Care 0.36 0.16 0.37 0.33 0.35 0.18 1

8. Fairness 0.34 0.19 0.32 0.34 0.34 0.13 0.73 1

9. Loyalty 0.51 0.43 0.61 0.52 0.62 0.39 0.58 0.57 1

10. Authority 0.39 0.55 0.51 0.39 0.56 0.54 0.40 0.31 0.64 1

11. Purity 0.49 0.60 0.56 0.46 0.64 0.61 0.49 0.48 0.68 0.74 1

12. Political
Conservatism
0.14 0.38 0.17 0.12 0.26 0.38 -
0.08
-
0.11
0.08 0.31 0.25 1

13. Extraversion -
0.04
0.06 0.09 0.01 0.04 0.13 0.05 0.04 0.10 0.14 0.09 0.00 1

14. Agreeableness 0.22 0.11 0.20 0.14 0.20 0.18 0.16 0.16 0.17 0.19 0.20 0.10 -
0.22
1

15.
Conscientiousness
0.21 0.17 0.24 0.13 0.23 0.23 0.18 0.17 0.22 0.27 0.34 0.02 0.02 0.26 1

16.Emotional
Stability
-
0.07
-
0.05
0.00 -
0.14
-
0.07
0.07 -
0.09
-
0.10
-
0.06
-
0.05
-
0.07
0.02 -
0.05
0.07 0.08 1

17. Openness to
Experience
0.05 -
0.22
0.00 0.06 -
0.05
-
0.13
0.20 0.22 0.06 -
0.15
-
0.08
-
0.19
0.06 0.00 0.15 0.07 1
18. Happiness 0.03 0.00 0.14 -
0.07
0.03 0.22 0.11 0.04 0.10 0.07 0.11 -
0.02
0.29 0.01 0.14 0.27 0.11

Study 3.4b: Demographic Correlates of Qeirat Values. We examined sex differences in
QVS subscales as well as the overall scores. After correcting for multiple comparisons, there was
no sex difference in “Commitment to Family” (t = 1.28, Welch-corrected df = 476.35, p = .202,
Cohen’s d = 0.12), “Love of Country” (t = 2.19, Welch-corrected df = 430.25, p = .029, Cohen’s

80

d = 0.20), “Supporting Family/Mate” (t = -0.57, Welch-corrected df = 476.24, p = .566, Cohen’s
d = -0.05), or the overall Qeirat values (t = -0.37, Welch-corrected df = 460.92, p = .713,
Cohen’s d = -0.03), but men scored significantly higher than women on “Protecting Namoos” (t
= -3.66, Welch-corrected df = 481.11, p < .001, Cohen’s d = -0.33). Age, education, and socio-
economic status were not significantly related to Qeirat values (ps > .127).
Study 3.4b: Individual Difference Correlates of Qeirat Values. The correlation
coefficients between Qeirat values, moral foundations, Big Five personality traits, Islamic
religiosity, political conservatism, and happiness are presented in Table 3.8. We also ran five
linear regression models to predict the four factors of Qeirat values and the overall Qeirat values
score with moral foundations, Big Five traits, Islamic religiosity, political conservatism, and
happiness as independent variables (see Table 3.9). As can be seen, overall Qeirat values scores
were predicted by Loyalty (B = 0.42, SE = 0.06, t = 7.00, p < .001), Purity (B = 0.22, SE = 0.06, t
= 3.86, p < .001), Islamic religiosity (B = 0.15, SE = 0.04, t = 3.55, p < .001), and political
conservatism (B = 0.05, SE = 0.02, t = 2.29, p = .023). Thus, Loyalty and Purity play an
important role in Qeirat values along with Islamic religiosity and political conservatism.

81

Table 3.9
Linear Models Predicting Qeirat Values (Study 3.4b)

Dependent variable:
Commitment
to Family
Protecting
Namoos
Love of
Country
Supporting
Family/Mate
Overall Qeirat
Values
Care .07 (.07) −.24
∗∗

(.09) .07 (.08) .01 (.07) −.02 (.06)
Fairness −.02 (.08) .10 (.10) −.12 (.09) .02 (.08) −.01 (.07)
Loyalty .37
∗∗∗

(.07) .19
∗

(.09) .68
∗∗∗

(.08) .45
∗∗∗

(.07) .42
∗∗∗

(.06)
Authority −.07 (.06) .16
∗

(.08) .03 (.08) −.01 (.07) .03 (.05)
Purity .20
∗∗

(.07) .40
∗∗∗

(.08) .13 (.08) .16
∗

(.07) .22
∗∗∗

(.06)
Islamic Religiosity .04 (.05) .30
∗∗∗

(.06) .20
∗∗∗

(.06) .07 (.05) .15
∗∗∗

(.04)
Political Conservatism .04 (.02) .12
∗∗∗

(.03) .02 (.03) .03 (.03) .05
∗

(.02)
Extraversion −.04 (.02) .01 (.03) .01 (.03) −.01 (.03) −.01 (.02)
Agreeableness .07
∗

(.04) −.04 (.05) .07 (.04) .04 (.04) .04 (.03)
Conscientiousness .03 (.03) .002 (.04) .05 (.04) −.02 (.04) .02 (.03)
Emotional Stability −.03 (.02) −.01 (.03) .01 (.03) −.05
∗

(.03) −.02 (.02)
Openness to Experience .04 (.03) −.11
∗∗

(.04) .004 (.04) .07 (.03) −.0005 (.03)
Happiness −.01 (.02) −.04
∗

(.02) .02 (.02) −.05
∗∗

(.02) −.02 (.01)
Constant 3.73
∗∗∗

(.33) 2.39
∗∗∗

(.42) 1.53
∗∗∗

(.41) 3.77
∗∗∗

(.36) 2.86
∗∗∗

(.29)
R
2
.33 .48 .45 .33 .51
Adjusted R
2
.31 .47 .44 .31 .49
Note. *p < 0.05 **p < 0.01 ***p < 0.001
Discussion
Study 3.4 used a mixed-methods methodology to combine interview-based qualitative
study of Qeirat values (Study 3.4a) and quantitative analyses of its relationship with related
constructs (Study 3.4b) including moral foundations, religiosity, political conservatism,
personality dimensions, and subjective happiness. Based on our qualitative analysis, we
developed and validated a 24-item self-report measure of Qeirat values, which demonstrated a 4-
factor structure (Commitment to Family, Protecting Namoos, Love of Country, and Supporting

82

Family/Mate). Importantly, these results demystified a few lay conceptions of the concept of
Qeirat in Iranian culture. For example, Qeirat values are believed to be more masculine and
valued by men; however, we only found a male-favoring sex difference in one facet of Qeirat
values, not others. This facet was “Protecting Namoos” which consists of (over)protection of
female kin in different domains, but mostly in terms of sexual activities (see Tizro, 2013). In
addition, the overall Qeirat values scores were not significantly different between women and
men. In terms of the relationship between moral foundations and Qeirat values, Loyalty and
Purity were shown to be reliable predictors of Qeirat values, even after controlling for other
moral foundations, personality dimensions, political ideology, and subjective happiness. At the
broadest level, these findings provide preliminary evidence for the nomological network of
Qeirat values, and demonstrate the multidimensional nature of Qeirat values in Iran. The current
findings also highlight that while Qeirat is comparable with honor values and “culture of honor”
researched in the US (Cohen, Nisbett, Bowdle, & Schwarz, 1996), it consists of culturally
relevant practices (e.g., protecting Namoos) that are not conceptualized to be a part of honor
cultures. We examine the relationship between honor and Qeirat in Study 3.5.
Study 3.5: Pragmatic validity of Qeirat as a novel moral foundation
Qeirat values, according to our theoretical delineation in Study 3.4, are conceptually
proximate to, but not identical with, the concept of honor. Western conceptualizations of honor
cultures based on anthropological and cross-cultural research suggest that honor values include
maintenance of good reputation through good family standing, social interdependence, and
maintenance of gender-specific codes of behavior (Oyserman, 2017; Peristiany, 1965) which
have been linked to violence and homicide (Grosjean, 2014). In Study 3.5, we have three

83

principal predictions to demonstrate the pragmatic validity of Qeirat as a potential moral
foundation: (1) Qeirat values are related to, but distinct from honor values; (2) Qeirat values
explain additional variance in outcomes after accounting for other moral foundations and honor
values; (3) Qeirat values are linked to mate guarding behaviors in the Iranian culture.
Methods
Participants. We aimed to recruit 100 Iranian participants using advertisements on social
media. A post-hoc power analysis suggested that this sample size can detect a correlation
coefficient of 0.3, at p = 0.05, with 87% power. Accordingly, any correlation below 0.3 should
be treated with caution in this sample. Our final sample included 105 participants (63% female).
The average age was 33.9 years (SD = 9.9 years) and most participants had at least some college
education (88.6%). About half of the participants were in marital relationships (46.7%).
Participants reported their subjective socio-economic status on a five-point Likert-type scale,
where most participants identified as middle class (60%).
Measures. All participants completed the following self-report measures.
Qeirat Values Scale (QVS). All participants completed the 24-item QVS developed and
validated in Study 3.4. Items were rated along a 7-point Likert-type scale ranging from 1
(Strongly disagree) to 7 (Strongly agree). Internal consistency coefficients for Commitment to
Family (α = .88), Protecting Namoos (α = .89), Love of Country (α = .91), and Supporting
Family/Mate (α = .78) were satisfactory. The overall Qeirat values score was also internally
consistent (α = .94).
Moral Foundations Questionnaire (MFQ). Participants completed the Persian version
of the 30-item MFQ (Graham et al., 2011) validated in Study 3.1. The internal consistency

84

coefficients were .43, .47, .63, .79, and .83 for Care, Fairness, Loyalty, Authority, and Purity,
respectively.
Political Orientation. Participants rated their political affiliation as using a single item
ranging from 1 (Politically liberal) to 7 (Politically conservative). The average rating was 3.35
(Md = 4, SD = 1.32).
Iranian National Religiosity Scale (INRS). Participants completed the 36-item INRS
(Khodayarifard et al., 2010). All items were rated on a 6-point Likert-type scale ranging from1
(Never) to 6 (Always). The INRS scores were highly internally consistent in the current sample
(Cronbach’s α = .96).
Mate Retention Inventory-Short Form (MRI-SF). All participants completed the MRI-SF
(Buss, Shackelford, & McKibbin, 2008) which consists of 38 items assessing 19 mate retention
tactics. Items were rated in terms of frequency on a 4-point Likert-type scale ranging from 1
(Never) to 4 (Often) how often they performed each behavior within the past year. Atari et al.
(2017) translated the instrument into Persian and reported satisfactory psychometric properties of
the Persian translation of the MRI-SF in a sample of Iranian adults. These authors found two
superordinate components underlying mate retention behaviors in Iran: benefit-provisioning
(behaviors that reduce the likelihood of partner infidelity by increasing relationship satisfaction)
and cost-inflicting (behaviors that reduce the likelihood of partner infidelity by lowering the
partner’s self-esteem). Composite scores of benefit-provisioning (Cronbach’s α = 0.92) and cost-
inflicting (Cronbach’s α = 0.88) components were calculated by averaging the relevant items.
Honor Values Scale (HVS). Participants completed the 18-item Honor Values Scale
(HVS; Novin & Oyserman, 2016) which has been developed based on expansion of IJzerman,
van Dijk and Gallucci (2007). Items were rated on a scale from 1 (Strongly disagree) to 7

85

(Strongly agree). Since this scale has not been validated in Iran, we translated the measure into
Persian using the standard back-translation technique (Brislin, 1970). Items were averaged and
total scores were shown to be internally consistent (Cronbach’s α = 0.84).
Procedure. Potential adult participants who were in a committed relationship were invited
to take part in a psychological study on social media. Participation was on a voluntary basis and
participants were not compensated. All participants who agreed to take part were provided with
all the measures and were debriefed and thanked once they finished the survey.
Data Analysis. We used Pearson correlation to examine the association between Qeirat
values and honor values where perfect or very high correlations (> .85) indicate that the two
constructs are redundant (see Campbell & Fiske, 1959). Linear regressions were used to
investigate whether Qeirat values predict outcomes (Islamic religiosity, political conservatism,
benefit-provisioning mate retention, and cost-inflicting mate retention) after accounting for Care,
Fairness, Loyalty, Authority, Purity, and honor values. All data analyses were conducted using
the Psych package (Revelle, 2017) in R (version 3.6.1).
Results
Qeirat values are distinct from honor values. The correlation between overall Qeirat
values and honor values was .46, 95%CI [.29, .60], p < .001. This moderate correlation (shared
variance = 21%) suggests that the two constructs are not redundant and are capturing
distinguishable constructs. Subscale-level correlations between facets of Qeirat showed moderate
correlations (.27 < r < .43) as well (Table 3.10), indicating that none of these subscales are
capturing the same construct as honor values.

86

Table 3.10
The Correlation Coefficients between Study Variables (Study 3.5)
Note. *p < .05 **p < .01 ***p < .001. BPMR = benefit-provisioning mate retention; CIMR =
cost-inflicting mate retention.
Qeirat values predict outcomes beyond moral foundations. We ran two regression
models to predict Islamic religiosity and political conservatism using demographic variables,
moral foundations, honor values, and Qeirat values. As can be seen in Table 3.11, Qeirat values
significantly predicted Islamic religiosity (B = 0.33, SE = 0.10, p = .001) and political
conservatism (B= 0.49, SE = 0.18, p = .006) after accounting for all other variables. Honor
Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14
1. Care 1
2. Fairness .37*** 1
3. Loyalty .22* .15 1
4. Authority .24* .17 .68*** 1
5. Purity .31** .25** .69*** .75*** 1
6. BPMR -.10 -.06 .16 .16 .15 1
7. CIMR -.12 .06 .17 .18 .16 .53*** 1
8. honor
values
.20* .30** .41*** .44*** .52*** .22* .20* 1
9. religiosity .14 .15 .53*** .63*** .73*** .21* .17 .48*** 1
10.
conservatism
.18 .22* .33*** .48*** .51*** .08 .02 .24* .45*** 1
11. Qeirat:
commitment
to family
.20* .23* .47*** .56*** .64*** .20* .21* .43*** .54*** .40*** 1
12. Qeirat:
Protecting
Namoos
.07 .11 .59*** .66*** .69*** .35*** .39*** .41*** .68*** .48*** .67*** 1
13. Qeirat:
Love of
country
.13 .12 .68*** .64*** .67*** .21* .14 .41*** .64*** .50*** .46*** .63*** 1
14. Qeirat:
supporting
family/mate
.09 .27** .49*** .47*** .51*** .23* .28** .27** .49*** .35*** .68*** .64*** .52*** 1
15. Overall
Qeirat
.14 .20* .68*** .71*** .76*** .31** .31** .46*** .72*** .53*** .80*** .90*** .80*** .83***

87

values, on the other hand, were not found to significantly predict Islamic religiosity (B = 0.20,
SE = 0.16, p = .220) or political conservatism (B = -0.27, SE = 0.28, p = .336).
Qeirat values are linked to mate guarding. We ran two regression models to predict
benefit-provisioning and cost-inflicting mate retention behaviors using demographic variables,
moral foundations, honor values, and Qeirat values (see Table 3.11). Qeirat values significantly
predicted benefit-provisioning (B = 0.21, SE = 0.09, p = .024) and cost-inflicting (B = 0.13, SE =
0.07, p = .048) mate retention behaviors after controlling for demographic variables, moral
foundations, and honor values.
Table 3.11
Regression Results Predicting Religiosity, Conservatism, and Mate Retention Behaviors (Study
3.5)
Predictor Dependent variable
Religiosity Conservatism BPMR CIMR
Sex (0=m,1=f) -0.09 (.14) -0.03 (.24) -0.36** (.13) -0.18 (.09)
Age 0.00 (.01) 0.01 (.02) -0.02** (.01) -0.01 (.01)
Education 0.01 (.06) -0.01 (.10) 0.12* (.05) 0.03 (.04)
SES -0.02 (.10) 0.32 (.17) 0.09 (.09) 0.14* (.06)
Marital status 0.07 (.16) 0.18 (.28) 0.23 (.15) 0.08 (.11)
Care -0.07 (.13) 0.14 (.22) -0.02 (.12) -0.07 (.09)
Fairness -0.07 (.13) 0.26 (.23) -0.06 (.12) 0.07 (.09)

88

Loyalty -0.15 (.12) -0.38 (.21) -0.03 (.11) -0.02 (.08)
Authority 0.10 (.10) 0.19 (.17) -0.01 (.09) -0.00 (.06)
Purity 0.35** (.11) 0.27 (.18) -0.07 (.10) -0.04 (.07)
Honor values 0.20 (.16) -0.27 (.28) 0.35* (.15) 0.19 (.11)
Qeirat values 0.33** (.10) 0.49** (.18) 0.21* (.09) 0.13* (.07)
Constant 0.71 (.99) -1.89 (1.71) 1.39 (.91) 0.43 (.65)
R
2
.62 .38 .29 .24
Adjusted R
2
.58 .30 .19 .14
Note. *p< .05 **p < .01 ***p < .001. BPMR = benefit-provisioning mate retention; CIMR =
cost-inflicting mate retention.

Discussion
Study 3.5 had three specific aims to further examine the pragmatic validity of Qeirat as a
potential moral foundation (at least in Iran), that is, to show that Qeirat is a scientifically useful
construct for both answering existing questions about morally relevant phenomena and allowing
researchers to formulate new questions (see Graham et al., 2011). First, we demonstrated that
Qeirat is a distinguishable construct from honor as conceptualized and measured in the literature.
However, it is possible that Qeirat and honor have similar evolutionary architectures which are
universally available, serving to protect family members, land, and reputation especially in the
absence of powerful central authorities, but are triggered by different socio-ecological needs. For
example, mate retention aspects of Qeirat/honor may be more manifest in male-biased sex ratios

89

or in contexts where mate poaching is prevalent. Second, we show that Qeirat values are useful
as they predict important outcomes (here, religiosity and conservatism) above and beyond
existing moral foundations and even honor values. It is worth noting that individualizing
foundations of Care and Fairness had particularly low reliabilities in this sample, which may
explain why the MFQ had low predictive power in this study. These low reliability coefficients
in the present sample further suggest that the MFQ does not perform well in Iranian samples.
Third, drawing upon Atari et al. (2017), we postulated that Qeirat values are linked to mate
retention behaviors. Hence, Qeirat values facilitate a broad menu of behaviors ranging from acts
of kindness and resource provisioning, to vigilance, manipulation, and violence to retain and
protect committed exclusive relationships in Iranian culture. This is preliminary evidence that
Qeirat (and honor) may have evolved to solve the evolutionarily recurrent problem of mate
guarding (see Shackelford, 2005).
General Discussion
Psychological theories, whether empirically-constrained or driven by folk conceptions, are
generalizable across human populations only when their supporting data come from diverse
human populations. Morality is a universal human phenomenon, present in our literature, history,
society, relationships, and identity. But it is apparent, from the present studies and prior work on
cross-cultural differences in moral values, that the moral domain’s content and structure may
substantially differ from one culture to another. MFT’s operationalization, while a useful
framework for understanding ubiquitous moral phenomena across cultural contexts, is
nevertheless presently unable to assess the moral domain comprehensively in some cultures,
particularly non-WEIRD ones (see Purzycki et al., 2018). This does not undermine the theory
itself (see Graham et al., 2013); rather, it points to an important direction in psychology, the

90

holistic and contextualized translation of theories and measures — in terms of both culture and
language — for truly cross-cultural evolutionary psychology of morality (Apicella & Barrett,
2016). MFT theorists have explicitly welcomed new foundations to be added to their framework
as methods and theory co-evolve in moral psychology. Specifically, with regard to addition of
new “foundations”, Graham et al. (2013, p.58) laid out a number of foundationhood criteria and
rhetorically posited that they “do not know how many moral foundations there really are. There
may be 74, or perhaps 122, or 27, or maybe only 5, but certainly more than one.”
In the broadest sense, the current program of research provides two insights: First, the
MFQ, as the most used operationalization of MFT, does not fare well in Iran. This finding does
not undermine the theory itself as future research can aim to develop measures that capture the
constructs introduced by MFT using culturally fluent items. However, even if future research
does so and succeeds in capturing the five moral foundations, there might still be culturally
important moral systems overlooked. Second, we used a descriptive set of studies to uncover a
new system of moral meanings called Qeirat. While we found similarities between the concept of
Qeirat and the culture of honor in the US, they seem to be tapping into different constructs.
Discovering Qeirat as a new moral system that cannot be mapped onto existing moral
foundations and that is linked to evolutionary mechanisms not identified before can motivate
research to further explore the foundationhood of Qeirat.
In our studies, we aimed to measure how well the operationalization of MFT fares in Iran.
Our psychometric studies showed that the Persian translation of MFQ does not replicate its
parent structure (Graham et al., 2011), nor are MFQ scores quite meaningfully comparable
between Iran and the US. In addition, our network psychometric analyses revealed that the
network of moral foundations is substantially different between Iran and the US, with the Iranian

91

network showing a more intimate relationship between individualizing and binding moral
foundations. These findings call for at least two future directions. First, it is invaluable to
develop better operationalizations of MFT based on these findings and recent cross-cultural
ethnographic work (Lang et al., 2019; Purzycki et al., 2018). Such theoretical revisions were both
predicted and encouraged by MFT (Graham et al., 2013; Iyer et al., 2012). Second, these results
show that mere translation of models based on WEIRD psychology is not sufficient to wholly
understand human psychology, particularly the moral domain (see Medin, 2017). Our qualitative
studies (in an “informed curiosity” manner; see Rozin, 2001) uncovered moral concepts which
were previously incognito to (Western) psychologists (for another example, see Berniūnas,
2020). Future moral psychological work can enjoy truly cross-cultural exploratory research to
uncover moral systems currently unknown to scholars of morality.
We have demonstrated, through a mixed-methods design, that Iranians possess a set of
moral values that is different from the US (and possibly other WEIRD populations), which may
not be explained by the MFT (but most closely related to the conception of Loyalty in MFT). But
why do Iranians have a significantly different structure of moral values than Americans, other
WEIRD populations, or even other non-WEIRD populations (see Doğruyol et al., 2019)? It is
evident that Loyalty is an important, organizing concern in Iranian morality, and that Qeirat
underlies the Iranian mind in many ways, not just morality (e.g., basic personality models,
mating psychology). One explanation takes the evolutionary perspective in these differences:
morality exists as a set of culturally transmitted information in the society, functioning for the
preservation and promotion of the self and of group. When the ecology — specifically, threats to
the self- and group-level thriving (e.g., pathogen prevalence, population density, biased sex ratio,
resource scarcity) — changes, the network of moral concerns can thus adapt (see Haidt &

92

Joseph, 2004; Oishi & Graham, 2010; Purzycki et al., 2018; Tomasello, 2018; van Leeuwen,
Park, Koenig, & Graham, 2012). Here, we argue that Qeirat values are particularly adaptive in
Iran’s ecology due to historically high prevalence of pathogens, slightly male-biased sex ratio,
and scarcity of environmental resources, especially in southern regions. Qeirat values maintain
intensive kinship networks which can function to keep resources in the group. Ecological
conditions that facilitate Qeirat values give rise to higher intrasexual rivalry for access to mates,
sensitivity to sexual norm violation, and vigilance to guard current sexual partners, especially
female partners (Kandrik, Jones, & DeBruine, 2015; Karimi-Malekabadi & Esmaeilinasab,
2019). Evolution of Qeirat values both supports and enables tight kinship networks and group
coalitions in which the risk of contact with pathogens is minimized, mate poaching is heavily
penalized, resources are retained within the group, and societal norms are maintained.
Iran is relatively higher in the historical prevalence of pathogens (Murray & Schaller,
2010) and is slightly more male-biased in terms of population sex ratio (e.g., Chao, Gerland,
Cook, & Alkema, 2019). Historical presence of pathogens can induce higher ingroup loyalty
(Van Leeuwen et al., 2012) and male-biased sex ratio is associated with family values and
religiosity (Karimi-Malekabadi & Esmaeilinasab, 2019). Qeirat values are different from, but
comparable with the Southern culture of honor in the US, where Southerners are believed to be
readier to react (aggressively) to insults (Cohen et al., 1996). Importantly, Qeirat values are
oriented toward one’s mate, family, and country while the focus of Southern culture of honor is
one’s own dignity (e.g., Guerra, Giner-Sorolla, & Vasiljevic, 2013). The distinction between
Qeirat and honor was evident in Study 3.5 in which we found moderate correlations between
facets of Qeirat values and honor values. Yet, it is possible that Qeirat and honor have similar
underlying evolutionary functions, but are differentially formed by different socio-ecological

93

triggers (i.e., lack of resources, herding vs. farming) that are conducive to stronger family ties,
pro-group behaviors, higher frequency of mate retention behaviors (Nowak, Gelfand, Borkowski,
Cohen, & Hernandez, 2016; Shackelford, 2005).
Limitations of this project are worth noting. First, we did not collect data from different
subcultures of Iran in terms of ethnicity (e.g., Kurd, Turk, Balooch). It is recommended for future
research to examine moral foundations and Qeirat values across different regions, subcultures,
religions, and languages in Iran. This is particularly important because there is substantial
variance in socioecological factors across Iran’s geographic regions. Additionally, it might be a
good next step to examine Qeirat values in other, potentially WEIRD, populations. Second, we
did not examine all criteria for foundationhood of Qeirat, such as emotional consequences of
violation of Qeirat values (Razavi et al., 2020) or third-person normative judgments relevant to
Qeirat. Emotional reactions to moral foundations are well researched (Atari, Mostafazadeh
Davani, & Dehghani, 2020), but emotional correlates of Qeirat values are yet to be investigated.
Third, our findings are correlational and cross-sectional in nature. Future research is encouraged
to use experimental and longitudinal designs to better uncover sources of Qeirat values and their
development in the lifespan.
The first implication of this research is that true cultural and evolutionary research in moral
psychology cannot merely gather data from non-WEIRD samples using WEIRD models. Rather,
using models and measurement methods designed in WEIRD societies is itself a flawed practice,
and requires a reinvention of the theoretical framework (and its operationalizations) for a new
culture and/or language, potentially by researchers who can actively engage with the target
culture and/or language (Medin, 2017; Medin, Ojalehto, Marin, & Bang, 2017). The second
implication of the present project is that Qeirat values represent a measurable and

94

psychologically meaningful culture-specific set of kinship-intensifying moral norms that
function to bind individuals to their mates, families, and countries in ways previously unknown
to Western psychologists. Future investigations on Qeirat contribute to the cumulative theoretical
frameworks in understanding the role of culture and social norms in moral values (Razavi et al.,
2020; see Muthukrishna & Henrich, 2019).

95

Chapter 4: Theoretical Refinement of MFT and Moral Foundations Questionnaire-2
Moral Foundations Theory (MFT; Graham et al., 2013; Haidt & Joseph, 2004) was
designed to explain both the variations and ubiquitous aspects of moral judgments. Specifically,
MFT proposed five universally available, but contextually variable, moral concerns: Care,
Fairness, Loyalty, Authority, and Purity
4
. Measurement of “moral foundations” can be
particularly difficult given how much people vary in the extent to which they endorse them.
Graham and colleagues (2009, 2011) developed the Moral Foundations Questionnaire (MFQ) to
address the need to have a valid and reliable measure of MFT. This self-report measure, and its
adaptations, have been used in hundreds of empirical studies in different social and behavioral
fields, and across various cultures. However, recent theoretical critiques of MFT and
psychometric examinations of MFQ in diverse samples call for theoretical refinement and
psychometric improvement of the questionnaire. Here, we describe the development of the Moral
Foundations Questionnaire-2 (MFQ-2), based on an updated theoretical view on the number of
foundations and their content. We develop MFQ-2 using a new item pool administered across
countries in their local languages. We present the structural validation of the MFQ-2, its relation
to political ideology and religiosity, as well as an examination of cross-cultural similarities and
differences.
Moral Foundations Theory
How can a moral psychological theory fully account for the content and structure of
morality when people disagree so much, and so viciously, on moral issues? MFT was designed

4
These foundations have come with other names too. Haidt and Graham (2007) referred to them as Harm/Care,
Fairness/ Reciprocity, Ingroup/Loyalty, Authority/Respect, and Purity/Sanctity.

96

to fill the need for a systematic theory of morality, explaining its evolutionary origins, lifespan
development, and cultural variations. Haidt and Joseph (2004) reviewed evolutionary
psychology, cultural psychology, and anthropology to find areas of moral regulation that were
common, but not necessarily “universal,” across cultures. For instance, virtues of purity and
practices regulating sex and food (e.g., Douglas, 1966) were abundantly cited in the evolutionary
literature on disgust sensitivity (Rozin et al., 2000). The results of this cross-disciplinary review
produced five top candidates for being the psychological “foundations” upon which cultures
construct their moralities. These five foundations were consistent with, and expanded upon,
several taxonomies of moral concerns, including Fiske’s (1992) framework of social relations;
Shweder et al.’s (1997) account of the “three ethics” of autonomy, community, and divinity that
are found in different cultures; and Hogan et al.’s (1978) evolution-based socioanalytic theory of
moral development. Indeed, even if all moral systems are social constructions, they are
constructed by people whose minds are not blank slates (Marcus, 2004; Pinker, 2003). As such,
MFT allows for intuitive or emotional bases of moral judgments as well as more deliberate
reasoning processes (see Greene et al., 2001; Haidt, 2001).
MFT rests on four falsifiable tenets about human morality: (a) nativism, (b) cultural
learning, (c) intuitionism, and (d) pluralism. Nativism is the idea that there is a “first draft” of the
moral mind, meaning that human morality is somewhat organized in advance of experience. In
this regard, MFT borrows Marcus’s (2004) metaphor that the mind resembles a book: “Nature
provides a first draft, which experience then revises...‘Built-in’ does not mean unmalleable; it
means ‘organized in advance of experience’” (pp. 34-40). Cultural learning is the idea that this
“first draft” gets edited during development within a particular cultural context. Culture
facilitates smooth social coordination and clarifies group boundaries (Oyserman, 2011): people

97

learn moral values from others horizontally (i.e., learning from same-generation peers), obliquely
(i.e., learning from unrelated elders), and vertically (i.e., learning from one’s biological parents)
(see Mesoudi, 2019). MFT is a theory about the universal first draft of the moral mind and about
how that draft gets revised in variable ways across cultures. Intuitionism is the idea that intuitions
come first before strategic reasoning takes place. In metaphorical terms, moral evaluations
happen in the “gut” rather than the “head”. Moral intuitions tend to fall into related but
distinguishable categories. MFT was designed to say exactly what those categories are and in
what ways these automatic moral intuitions vary across cultures. Finally, pluralism is the idea
that there were many recurrent challenges in our evolutionary history to coordinate social life,
therefore multiple moral sources of moral intuition exist (Graham et al., 2013).
The first two foundations — Care and Fairness — generally correspond to Shweder and
colleagues’ (1997) ethics of autonomy. Loyalty and Authority generally correspond to the ethics
of community; and Purity generally corresponds to the ethics of divinity. The first two
foundations center around protection of individuals and are commonly referred to as
“individualizing” foundations while Loyalty, Authority, and Purity are referred to as “binding”
foundations because they are focused on preservation of group harmony and binding individuals
into larger groups and institutions. Moreover, MFT provides a theoretical framework to describe,
explain, and predict differences in moral concerns across individuals and groups. Pragmatically,
MFT provides a basis to understand how moral values motivate individual-level behaviors as
well as how groups build institutions in different cultural settings. Theory-driven scale validation
is a means of testing hypotheses and theory-driven predictions (Flake, & Fried, 2020; Hogan &
Nicholson, 1988) and hence MFQ was developed in 2011 to provide a self-report measure of to
what extent individuals endorse these five moral foundations.

98

Moral Foundations Questionnaire
The first effort to develop a theory-driven measure of MFT was the development and
validation of MFQ (Graham et al., 2011), which remains the primary self-report measure of the
degree to which individuals endorse each of the five dimensions introduced by MFT. Graham
and colleagues (2011) used both exploratory and confirmatory factor analyses to develop the 30-
item MFQ. In this measure, two “individualizing” foundations of Care and Fairness focus on the
welfare of individuals and the responsibility to respect others’ rights, comparable with the
traditional accounts of morality: Care (whether someone is hurt or harmed) and Fairness
(whether someone cheats or is deprived of their rights). There are also three “binding” subscales:
Loyalty (whether one’s ingroup is betrayed), Authority (whether respected people, customs, and
traditions are properly respected by others), and Purity (whether the intrinsic purity of an object
is degraded). MFQ has two parts: in part one (called “Relevance”) participants are explicitly
asked to evaluate the moral relevance of several foundation-related concerns (e.g., “Whether or
not some people were treated differently from others” for Fairness). In part two (called
“Judgements”) participants are asked specific and contextualized moral-judgment statements on
which they can agree or disagree (e.g., “If I were a soldier and disagreed with my commanding
officer’s orders, I would obey anyway because that is my duty” for Authority).
Prior work has established some evidence for construct validity of the MFQ. For example
Graham et al. (2011) documented that Care scores were positively correlated with empathy,
generosity, and pacifism; the Fairness scores were positively associated with valuing social
justice and negatively correlated with social dominance; the Loyalty scores were positively
correlated with concerns over national security; the Authority scores were positively correlated
with respect for tradition and right-wing authoritarianism; and finally, the Purity scores were

99

positively correlated with valuing self-discipline, religious attendance, disgust sensitivity, and
unfavorable attitudes toward casual sexual encounters. Scores on the MFQ have also been shown
to predict emotional reactions to various moral transgressions (Atari et al., 2020a), political
ideology (Kivikangas et al., 2021; also see Hatemi et al., 2019), religiosity (Yi & Tsang, 2020),
vaccine hesitancy (Amin et al., 2017), patterns of language use (Kennedy et al., 2021), and
attitudes toward public policy (Clifford & Jerit, 2013), multiple sources of intuition (Hoover et
al., 2021), and charitable giving (Nilsson et al., 2020).
MFQ has been used in a wide variety of settings to examine group differences and
cultural practices. Haidt and Graham (2007) first applied the theory to understand the “culture
war” between political liberals and conservatives in the United States. Drawing on Shweder and
colleagues as well as several political theorists (e.g., Burke, 1790/2003; Mill, 1859/2003; Sowell,
2002), being a liberal was hypothesized to embrace a morality in which the individual is the
locus of moral value. Liberals in the United States have indeed been found to score slightly
higher than their conservative counterparts on Care and Fairness. These findings has since been
replicated multiple times (Klein et al., 2018). On the other hand, conservatives tend to score
higher on Loyalty, Authority, and Purity than do liberals. In a similar vein, Koleva et al. (2012)
sought to understand the psychological underpinnings of specific “culture war” attitudes in 20
issues (e.g., abortion, immigration, same-sex marriage) using an MFT perspective. These authors
found that MFQ scores predicted attitudes about culture war issues after adjusting for ideology,
age, gender, religious attendance, and interest in politics. Most notably, people’s scores on Care
predicted their moral disapproval of death penalty and animal testing, while people’s Purity
scores predicted their disapproval of abortion, same-sex relations, and gambling above and
beyond political conservatism and other demographic characteristics.

100

Between-Group Differences
Cultural Differences. About three decades ago, Schweder and Haidt (1993) called for
culturally informed theories of moral cognition. Scholars have argued that moral appraisals differ
substantially across individuals, cultures, and historical periods. For example, Shweder showed
that in India, among Brahmans, it is “immoral” for a son to eat meat or cut his hair during the 10
days that follow the death of his father. However, this practice may be considered amoral in
Western cultures. Hence, moral pluralism, best operationalized by MFT, provides an
unprecedented opportunity for descriptive study of cultural differences in moral values. Graham
et al. (2011) did not provide a comprehensive picture of cultural variation in moral foundations,
but they did compare participants from Eastern cultures (South Asia, East Asia, and Southeast
Asia) with those from Western cultures (United States, United Kingdom, Canada, and Western
Europe). Eastern participants showed stronger concerns about Loyalty and Purity compared with
their Western counterparts, and they were only very slightly more concerned about Care,
Fairness, and Authority. According to this research, larger cultural differences in Loyalty and
Purity stood to reason in light of established cultural differences in collectivism (Triandis, 1995)
and the role of spiritual-physical purity concerns in daily life and religious practice, particularly
in South Asia (Shweder et al., 1997). The small effect sizes for all the East-West differences
suggest that group-differences within cultures (e.g., by gender or political ideology) could
exceed the East-West variations given so much attention in cross-cultural research.
Only recently has it become possible to examine morality beyond typical WEIRD
(Western, Educated, Industrialized, Rich, and Democratic; Henrich et al., 2010) samples in
psychological research. For example, Atari et al. (2020b) evaluated the MFQ in the Iranian
culture, an understudied non-WEIRD cultural setting, and followed up by building a bottom-up

101

model of moral values. These authors also compared moral foundations between Iran and the
U.S. finding that Iranians’ scores on MFQ cannot be reliably compared with their American
counterparts as the two cultures differ in the pattern of responding to questionnaire items, but
after accounting for these pattern differences, Iranians still scored higher on Loyalty, Authority,
and Purity. These authors also found a culture-specific moral construct, “Qeirat,” which does not
have a straight English translation, but is semantically close to “honor” and consists of guarding
and protectiveness of female kin, romantic partners, broader family, and country. Endorsement
of Qeirat was highly correlated with Loyalty, Authority, Purity, and Islamic religiosity. Another
study in Mongolia found that the term “moral” brings to mind exclusively WEIRD associations,
and the culture-specific “yos surtakhuun” brings to mind specifically Mongolian associations.
Berniūnas (2020) argues that it is plausible to think that Mongolians’ strong emphasis on
authority and respect, “khündlekh,” points to the socially shared cultural model of “yos
surtakhuun” behavior.
Notably, in the last few years a plethora of cultural psychological tools and opportunities
have become available for psychologists which were not available when MFQ was developed
about a decade ago. Most notably, Muthukrishna et al. (2020) developed and validated a tool and
a quantitative method for measuring the psychological and cultural distance between societies,
hence creating a distance scale with any population as the point of comparison, sometimes
referred to as the WEIRDness cultural distance. Hence, it is now possible to examine how
distance from WEIRD societies (typically exemplified by the U.S.) is associated with moral
foundations in different cultures. In addition, collecting stratified and representative samples
online across cultures has become readily available (Litman et al., 2017) which is particularly

102

important given the selection biases associated with crowdsourcing websites such as
YourMorals.org using which MFQ was initially validated (Kivikangas et al., 2021).
Ideological Differences. The titular findings of Graham and colleagues (Graham et al.,
2009), “Liberals and conservatives rely on different sets of moral foundations,” have become
increasingly important, as MFT has been widely utilized in the explanation of political
differences (e.g., Federico, Weber, Ergun, & Hunt, 2013; Hatemi et al., 2019; Koleva et al.,
2012; McAdams et al., 2008; van der Linden & Panagopoulos, 2019). The rapidly growing body
of work that has applied MFT in the context of political research has been based on the analyses
and interpretations by Graham and others according to which the original findings are consistent
and generalizable across different countries and cultures (Graham et al., 2011). In the large-scale
“Many Labs 2” study (Klein et al., 2018), which carried out large-scale replications for several
earlier findings including the association between moral foundations and political orientation, the
general patterns were similar; however, the average effect size was found to be smaller and the
effects showed significant heterogeneity. The link between moral foundations and political
ideology is typically represented in research as negative correlations between conservatism and
Care and Fairness, and positive correlations between conservatism and Loyalty, Authority, and
Purity.
In a systematic review and meta-analysis, Kivikangas and colleagues (2021) found that
while these ideological differences are mostly stable, they are smaller or less predictable outside
politically interested White American samples. Ideological differences also depend on how the
respondents are recruited, from which country and what demographics, hence it is important to
have diverse samples when a claim is made about ideological differences. Kivikangas and
colleagues (2021) found that the widely used, large-scale YourMorals sample (Graham et al.,

103

2011) has considerably inflated effect sizes. This difference was not explained by most
demographics (education, gender, and age) but is likely attributable to self-selection and
confounding political interest, which may also influence popular MTurk sampling (see Hauser &
Schwarz, 2016). Yet, moral foundations-political ideology associations were diminished
considerably when examining Black, and to a smaller extent, Hispanic respondents instead of
White respondents in the United States, or when examining samples from countries with
different political history and socio-economic structures. These variations in findings have
implications for designing studies and their sampling, most notably that studies on moral values
that want to make general claims about ideology should have diverse samples in terms of cultural
background and demographics.
Sex Differences. MFT provides a useful theoretical framework to examine how different
cultures show differing patterns of sex differentiation in each moral foundation. Graham et al.
(2011) found in an international sample that women scored higher on Care, Fairness, and Purity,
while men scored higher on Loyalty and Authority. These authors argued that women’s higher
Care and Purity concerns may be rooted in their higher emotional empathy and disgust
sensitivity. However, Graham and colleagues (2011) did not account for culturally variable sex
differences, did not consider the non-independence of cultures, did not examine country-level
correlates of sex differences, and did not compute multivariate estimates of sex differences. Atari
et al. (2020c) conducted two large-scale studies to exactly address these shortcomings. These
authors collected individual-level MFQ data and compiled a set of country-level data from
independent international organizations. Using both frequentist and Bayesian multilevel models
and considering non-independence of cultures, the results suggested that women scored higher
than men on Care, Fairness, and Purity across cultures. The sex differences in these moral

104

judgements seem to be replicable and robust across cultures. On the other hand, sex differences
in Loyalty and Authority were quite variable across cultures and negligible in size. Finally,
multivariate sex differences in moral foundations were larger in more individualist and gender-
equal cultures such as Sweden and France.
Religious Differences. There is little doubt that Gods command their followers to obey
moral directives and treat other individuals compassionately and fairly (Bloom, 2012; McKay &
Whitehouse, 2015) and many religious individuals think that the morally right action is the one
that God commands or requires. Therefore, it is not a surprise that many moral injunctions can be
seen in holy texts. The Care foundation, for example, is evident in the Hebrew Bible’s injunction
against murder (Exodus 20:13), in the ancient Hindu praise of the person who hurts nobody and
is compassionate toward all beings (Bhagavad-Gita 16:2), and in the Quran’s commandment to
be kind to “orphans, to the needy, to neighbors near and far, to travelers in need” (4:36). Many of
the religious commandments to treat others compassionately and fairly are limited to the
treatment of other individuals within the religious community; for example, the Hebrew Bible’s
“love your neighbor as yourself” (Leviticus 19:18) was intended to apply only to other Israelites
(Anderson, 1998). The Quran commands, “Do not take the Jews and Christians as allies: they are
allies only to each other. Anyone who takes them as an ally becomes one of them—God does not
guide such wrongdoers” (5:51). So, group boundaries and maintenance of group harmony is
central to understanding moral regulations in religious texts and religiously motivated behavior.
Theorizing by Graham and Haidt (2010) suggests that there may be important differences in the
content of religious and non-religious individuals’ moral foundations and that these differential
values may explain some important differences in moral judgment and behavior. In particular,

105

they proposed that, unlike their nonreligious counterparts, religious individuals cherish virtues
related to group loyalty, respect for authority, and bodily/spiritual purity.
Notably, conservatism is often linked with religiosity (e.g., Graham et al., 2009; Piazza &
Sousa, 2014), and religious individuals in the U.S. often endorse conservative ideologies (Jost et
al., 2014). It is understandable, therefore, that religious individuals prioritize the moral
foundations that also tend to be important to conservatives (i.e., Loyalty, Authority, and Purity;
Graham & Haidt, 2010). This claim is partially supported in that one important outcome of
religion is binding people into cooperative groups (Norenzayan, 2013). Other research has
examined the association between different facets of religiosity (see Atran & Norenzayan, 2004)
and moral foundations. Individual differences on religious dimensions may include out-reaching
vs. exclusive faith, high vs. low commitment to a religious group, belief in an authoritarian vs. a
benevolent God, and literal vs. metaphorical or anti-literal interpretations of the scriptures.
Controlling for conservatism and religious commitment, Johnson et al. (2016) found among
Christian Americans that Fairness was associated with outreaching faith; Care was positively
correlated with outreaching faith and negatively correlated with belief in an authoritarian God;
Authority was associated with literalism; and Purity was associated with literalism and
authoritarian God representations.
Gaps in Theory and Measurement
As reviewed above, MFT has been shown to be a highly generative theoretical
framework in multiple fields. However, there remain several limitations. Over the past decade, a
large body of empirical findings regarding the nature of moral foundations have amassed;
however, no comprehensive theoretical synthesis has been made to push the theory forward and
refine its claims based on new data and findings from diverse sources. Second, while an

106

increasing number of studies have found poor psychometric characteristics for MFQ, no
revisions have been made to the measure, hindering research on moral foundations. Here, we
focus on these two inter-related issues.
Revisiting the Theory. Each moral foundation has an evolutionary model, developmental
path, and cultural relevance. Since MFT was first conceptualized, psychologists have given
particular theoretical consideration to the Fairness foundation (e.g., Haidt, 2012). For some
individuals, especially liberals, Fairness means Equality, wherein everyone receives the same
quantity of resources. In an Equality-as-Fairness view, everyone ought to have equal rights and
any transgression from equal outcomes is morally disapproved. For other individuals, especially
conservatives, Fairness may mean Proportionality,
5
wherein people receive outcomes and
resources only relative to the amount of effort and input invested (see Jost et al., 2009).
Individuals who highly endorse the Proportionality-as-Fairness view tend to be concerned when
people receive benefits they don’t deserve, for example, when someone earns money without
having worked hard for it (Haidt, 2012; Rai & Fiske, 2011). One’s merit can be perceived along
a number of dimensions, such as effort, talent, ability, persistence, time, sacrifice, or
productivity.
Scholars from multiple disciplines have rightly criticized MFT for having failed to
include moral concerns for Equality and ignoring systemic inequalities (Janoff-Bulman &
Carnes, 2013b). In addition, MFT has yet to take into account people’s altruistic willingness to

5
Our use of the term “proportionality” is interchangeable with “equity” consistent with prior work (e.g., Deutsch,
1975; Haidt & Joseph, 2011; Rai & Fiske, 2011). We do not use the term “equity” because it has recently changed
its semantic connotation to mean “equality” in population-level literatures and social justice movements where
“equity” means ensuring all groups have what they need irrespective of their starting point. We use the term
“proportionality” throughout this paper to avoid confusion.

107

address existing societal inequalities — even at the expense of one’s own group within the same
society (Janoff-Bulman & Carnes, 2013a, 2013b). Many redistributive policies, such as
“Medicaid” in the U.S., receive considerable support among affluent liberal voters despite the
fact that these policies are paid for by the rich to benefit the poor. More recently, Skurka et al.
(2020) argued for the inclusion of Proportionality as a potential foundation since Proportionality
is conceptually distinct and empirically distinguishable from the original foundations as
measured in the MFQ, including Fairness. These authors found Proportionality to be highly
relevant to moral judgments regardless of political ideology, unlike the original five foundations.
So, in a sense, Proportionality may be considered less of politicized moral concern.
Studying “equity” has a long history in social psychology; however, since the term
“equity” has recently changed its meaning to ensuring that all groups have what they need
irrespective of their starting point, to avoid confusion, we use the word “proportionality”
synonymous with equity in this paper. From the late 1940s through the 1980s, social
psychological research on justice was dominated by a homo economicus guiding metaphor of
human nature (Skitka, 2009). Here, the core argument is that social life may be understood as
representing a series of negotiated exchanges, and that people use subjective cost-benefit
computations to guide their social interactions. The homo economicus perspective assumes that
people’s goals and concerns are ‘‘self-interested’’ or ‘‘selfish’’. That said, this perspective posits
that pure self-interest is avoided by people’s acceptance of the need for fairness in material and
social exchange to maximize their interests in the long run (e.g., Walster et al., 1976). Equity
theory (e.g., Adams, 1963, 1965; Homans, 1961) is probably the most well-known theory
coming out of this school of thought, arguing that rewards and punishments should be distributed
in accordance with recipients’ inputs or contribution. Adams’s (1965) work led him to conclude

108

that “when [a person] finds that his outcomes and inputs are not in balance in relation to those of
others, feelings of inequity result” (p. 280) and that “there can be little doubt that inequity results
in dissatisfaction, in an unpleasant emotional state, be it anger or guilt” (p. 283).
While there is evidence that the basic tenets of equity theory are widely shared across
different cultures (Konow, 2000), the salience of fairness considerations may vary depending on
the social norms that govern the particular decision-making context. Contextual manipulations of
the choice environment in dictator games can substantially alter behavior by altering subjects’
sense of what is normatively appropriate (e.g., Krupka & Weber, 2013).
Modern evolutionary models of Fairness often rely on preference for a proportionate ratio
of rewards to input in humans and non-human primates (e.g., Brosnan & de Waal, 2014; Fehr et
al., 2008). For example, Bräuer and Hanus (2012) define fairness in terms of “an interest in the
ideal of equity” and an evolved sense of inequity-aversion. Similarly, in their evolutionary
account of the origins of fairness, Debove and colleagues (2017) portray equity as the core of
human fairness evaluations. These authors argued that equity can be understood in the context of
the cooperative environment in which humans evolved. Debove et al. (2017) modeled a
population of individuals who cooperate to produce and divide resources, and select their
cooperative partners based on how they are willing to divide the resources. Agent-based
simulations, an analytical model, and extended simulations provided converging evidence that
equity is the best-fitting evolutionary strategy for such cooperative environments: agents
maximize their fitness by dividing benefits in proportion to their own and their partners’ relative
contributions. Hence, the need to be chosen as a cooperative partner creates a selection pressure
strong enough to explain the evolution of equity preferences among humans (see Axelrod &
Hamilton, 1981). Equity is not, however, limited to human fairness. For instance, it appears that

109

brown capuchin monkeys (Brosnan & de Waal, 2003) and long-tailed macaques (Massen et al.,
2012) may take effort into consideration when judging the fairness of rewards, lending credence
to the theory that equity evolved as a means of fostering cooperation between conspecifics.
Yet because the comparative psychological work is limited to research designs that probe
animals’ equity considerations (e.g., Proctor et al., 2013), it would be premature to conclude that
our recent ancestors lacked concern for equality. Consistent with this plausible distinction
between proportionality and equality, Brosnan (2013) suggests that both Equality and
Proportionality are amenable, in principle, to study in non-human primates, but that
interpretational challenges make it difficult to distinguish between these flavors of fairness in
non-human animals. In addition, it can be argued that Equality and Proportionality culturally
evolve (and diverge) much faster than can be traced in genetic evolution (see Chudek & Henrich,
2011).
Although merit-based reasoning is culturally widespread (Almås et al., 2010; Konow,
2000, Liénard et al., 2013; Schäfer et al., 2015; Zhang, 2020), people from collectivist and non-
Western cultures appear more likely to prefer equal distributions than do people from
individualistic cultures. In one study, Chinese adults liked allocators who divided rewards
equally more than those who divided rewards according to merit and viewed equal allocations as
fairer than proportional ones (Leung & Bond, 1984). Schäfer et al. (2015) showed that while
children from a modern Western society distributed the spoils of a joint enterprise precisely in
proportion to productivity, children from a gerontocratic pastoralist society in Africa did not take
merit into account at all. In addition, children from a partially hunter-gatherer, egalitarian
African culture distributed the spoils more equally than did the other two cultures, with merit
playing only a limited role. In other work, Huppert and colleagues (2019) had 4- to 11-year-olds

110

from 13 countries play a distributive game. Children were asked to allocate candy to fictional
recipients who varied in wealth (amount of candy resources), merit (effort on homework), or
elicited empathy (had a broken leg). Children from individualist societies favored equitable
distributions when the recipient was high in merit or high in need, whereas children from
collectivist cultures preferred equal distributions regardless of recipients’ merit.
The prior cited work on evolution of proportional and merit-based fairness in human
sociality might make one think that proportionality is the only “mode” of fairness in humans;
however, human social organization has consisted of essentially egalitarian, small-scale societies
for the majority of human history (Bowles, Smith, & Borgerhoff Mulder, 2010; Flannery &
Marcus, 2012). The patterning of status inequality over human evolution has been described as a
U-curve (Knauft et al., 1991), in which (i) chimp-like dominance hierarchies gave way to (ii)
egalitarianism within hunter-gatherer societies for much of the Pleistocene (2.5 million years ago
to 10 thousand years ago), followed by (iii) recent growth of inequality with the rise of large-
scale, agricultural societies. Thus, humans were largely egalitarian for most of our species’
existence, an inference based in part on ethnography of hunter-gatherer societies. This, however,
does not mean human fairness is solely based on “equality.” Rather, the ecological and
demographic conditions common to small-scale societies favored the suppression of steep,
dominance-based hierarchy (Boehm, 1999) and incentivized relatively shallow, prestige-based
hierarchy (von Rueden et al., 2019). Hence, shifts in ecological and demographic conditions, and
cultural-evolutionary forces, may weaken one flavor of fairness or give prominence to another
(see Von Rueden, 2020; Haynie et al., 2021). Institutions are made using both of these norms,
with some enhancing social equality and some ensuring that people are directly rewarded based
on their input.

111

The distinction between Equality and Proportionality represents a fundamental gap in
justice motivations. However, just as the intuitionist perspective on moral decision making
(Haidt, 2001) has shown that while there are multiple moral concerns that people universally
have access to, there are not an infinite number of justice flavors. There are many ways to define
justice; however, there are a few specific ways of defining justice that capture most of the
variance that we see in the world and that can be of pragmatic utility. The original
conceptualization of MFT does not differentiate between Equality and Proportionality, hence it
may miss out a substantial amount of variance in people’s justice motives.
Here, we make the case that MFT (and moral psychology, more broadly) benefits from
breaking Fairness into Equality and Proportionality. We note that an individual’s low scores on
Proportionality would not necessarily mean they are concerned with Equality; hence the two
constructs are not different ends of a single spectrum. Practically, people can take merit into
account in their decision making while actively caring about reducing inequality in the society
(as seen in some economically conservative, socially liberal individuals in the U.S.). We
speculate evolutionary processes, developmental paths, and cultural relevance of these flavors of
Fairness, generating a plethora of novel testable ideas, some of which we test in the present
article. This theoretical refinement of MFT’s Fairness foundations is intended as a corrective to
the often-tacit assumption in moral psychology that Fairness boils down to one single
conceptualization of (re)distributing resources in the context of social living. In short, we argue
that Fairness beliefs are diverse and heavily contingent on the socio-ecological contexts and
political systems in which people are chronically embedded. Our theoretically justified
differentiation between Equality and Proportionality, as well as developing valid measures for
both, opens the door to an array of interesting questions within the framework of MFT. For

112

example, this refinement raises questions of what ideological, economic, ecological, cultural, or
even historical factors give rise to different flavors of fairness, namely, Equality and
Proportionality. The relationship between these two constructs can be in itself an interesting
question as well. For example, based on recent work in formal computational modeling of
ecological niche, it can be the case that availability of diverse socioecological niches to
individuals within societies (i.e., more complex societies) can cause Equality and Proportionality
to be more “orthogonal” (i.e., more distinguishable constructs rather than one being a special
case of the other, or reflecting different aspects of a more basic psychological construct)
(Smaldino et al., 2019). In addition, cultural tightness and lower individualism may account for
higher covariance between these flavors of Fairness (see Gurven, 2018).
Revisiting the Measurement. In developing MFQ, Graham et al. (2011) conducted
confirmatory factor analyses (CFAs) based on the English version of MFQ in order to determine
whether the five-factor model of MFT fits data better than alternative models and showed that
the five-factor model fits the data better than the two-factor (individualizing vs. binding) and
single-factor models. Furthermore, independent scale validation studies in different cultures have
replicated this initial finding (e.g., Davies et al., 2014; Nejat & Hatami, 2019; Nilsson &
Erlandsson, 2015; Yalçındağ et al., 2019; Yilmaz et al., 2016). However, in all these studies, fit
indices of the 5-factor model were substantially below the conventional thresholds. A recent
cross-cultural study using the short form of MFQ in 27 different cultures also showed
measurement non-invariance across cultures (Iurino & Saucier, 2019). In other words, there is
some evidence suggesting that the five-factor model proposed by the theory is not cross-
culturally valid, and subscale scores may not be meaningfully compared across cultures because
patterns of responding are different from one culture to another. Internal consistency of MFQ

113

subscales (or foundations) also fail to reach conventional thresholds of 0.70, especially in more
diverse or representative samples. In most non-WEIRD populations, the individualizing
foundations (Care and Fairness) fare especially poorly, and Loyalty and Authority scores
typically fail to achieve adequate internal consistency (Nejat & Hatami, 2019).
In addition to the shortcomings with its factor structure, the wording of MFQ items have
been also rightly criticized. Many of MFQ items are explicitly political, especially biased toward
American political issues. For example, liberals may draw on the binding foundations, but in a
very distinct way than conservatives do (Janoff-Bulman & Carnes, 2013b). Conservatives may
feel disgust in reaction to violations of sexual purity, while liberals may feel disgust in reaction
to violations to the purity of the environment. According to Janoff-Bulman and Carnes (2013b),
conservatives prefer proscriptive virtues (i.e., involving inhibition), but liberals are more likely to
prefer prescriptive virtues (i.e., involving activation). Hence, the MFQ may not adequately
capture liberal ways of thinking about binding values. Finally, some MFQ items contain words
that preclude some populations. The Purity subscale of MFQ, in particular, conflates moral
purity with religiosity as one item explicitly contains the word “God” (i.e., “whether or not
someone acted in a way God would approve of”). According to classical measurement theory,
each item is assumed to provide an equivalent indicator of the latent construct (Chen, 2008), but
this particular item elicits a different pattern of scores for believers and nonbelievers (Davis et
al., 2017). Specifically, a conservative believer might strongly endorse all Purity items, whereas
a conservative nonbeliever might strongly endorse all Purity items except for the one explicitly
referring to God. A liberal believer, on the other hand, might strongly endorse the “God” item
and weakly endorse other items designed to capture Purity, whereas a liberal nonbeliever might
disagree with all items (Davis et al., 2017). This item is also problematic when MFQ is used in

114

widely secular societies (e.g., China) or where God could mean many different things in different
subpopulations (e.g., India).
Recently, Doğruyol et al. (2019) provided evidence that the five-factor model of moral
foundations, operationalized by the short version of the MFQ (20 items), is stable and invariant
across WEIRD and non-WEIRD societies; however, these authors used the problematic
dichotomy of WEIRD vs. non-WEIRD rather than treating societies on a continuum of
WEIRDness (Muthukrishna et al., 2020). Atari et al. (2020b) reported non-invariance of MFQ
scores between a non-WEIRD society and the U.S., as well as some difficulty in translating some
items into local languages (e.g., the item “I would call some acts wrong on the grounds that they
are unnatural”), arguing that MFQ scores may not be meaningfully comparable across different
cultures. In addition, Atari and colleagues (2020b) used network psychometric methods and
found that regardless of mean endorsement of moral foundations, the network of items and
foundations are substantially different between the two countries, with Iran having a denser
interconnected network of moral foundations, compared with the more segregated network of
moral concerns in the U.S., wherein Care-Fairness and Loyalty-Authority-Purity are two
disconnected “islands.” A recent study in the U.K. also failed to replicate the five-factor model
originally proposed by Graham and colleagues (2011), and suggested that “compassion” and
“traditionalism” may account for the structure of MFQ in the U.K.
Overview of the Present Research
After a decade from the development of its gold-standard measure, MFT has substantially
expanded the range of moral concerns under investigation in moral psychology, by encouraging
researchers to look beyond individual harm and fairness. A large body of research has examined
different theoretical claims by MFT, producing new insights into moral foundations on which

115

cultures build institutions and large-scale cooperation. This body of work has also highlighted
the assumptions and tools requiring refinement and further development.
In this work, we have five major goals. First, we refine MFT’s view on Fairness by
breaking it into two constructs, namely, Equality and Proportionality. Second, we generate a
completely new item pool and develop the Moral Foundations Questionnaire-2 (MFQ-2) across
cultures using local languages using highly generalizable samples. Third, we examine the
structural validity and measurement invariance of MFQ-2 across cultures. Fourth, we examine
group differences (cultural, ideological, sex, and religious differences) using the novel MFQ-2,
conceptually replicating prior work that has established these differences. Fifth, we establish
external validity of the MFQ-2 by examining associations between other scales meant to capture
similar and discriminant constructs.
Our measurement philosophy follows recommendations by Flake et al. (2017) in
following three phases of measure development: substantive (phase 1: literature review,
construct conceptualization, item pool development); structural (phase 2: item analysis, factor
analysis, reliability, measurement invariance); and external (phase 3: convergent and
discriminant validity, group differences). Our studies come in three phases and five studies
which we summarize in Table 4.1.

116

Table 4.1.
Description of Studies, Aims, and Samples
Phase Study Description
Sample
size
Cultures
1 4.1a
Literature review, panel discussion, item
pool development
840 2
1 4.1b Panel discussion, item pool reduction 971 3
1 4.1c Panel discussion, item pool reduction 1,549 3
2 4.2
Factor analysis, reliability, measurement
invariance, group differences
3,902 19
3 4.3 Convergent and discriminant validity 1,410 3

Study 4.1a
Study 4.1a was conducted to define the top-down structure we intend for MFQ-2 (Care,
Equality, Proportionality, Loyalty, Authority, and Purity) and to develop a preliminary MFQ-2
item pool that could be used to operationalize this theory-driven model. Ideally, this item pool
should be broad and balanced, with each foundation represented by several candidate items. Our
conceptual definitions of the six foundations we aim to measure are shown in Table 4.2. In all
studies, we have data from at least two cultures in order to avoid focusing narrowly on one
particular “default” culture. To avoid the “home-field disadvantage” (Medin et al., 2010), we
also made sure that our team has a diverse set of cultural backgrounds and views to make sure
that our item pool was not Eurocentric or biased toward a particular ideology. Here, we describe

117

the process of generating the item pool, initial analyses, and reducing the item pool for use in the
next studies.
Table 4.2.
Conceptual Definitions of Six Moral Foundations
Foundation Definition
Care Intuitions about avoiding emotional and physical damage to another
individual.
Equality Intuitions about equal treatment and equal outcome for individuals.
Proportionality Intuitions about individuals getting rewarded in proportion to their merit.
Loyalty Intuitions about cooperating with ingroups and competing with outgroups.
Authority Intuitions about deference toward people with status and influence in social
groups.
Purity Intuitions about avoiding bodily and spiritual contamination.

Methods
Participants and Procedure. We aimed to recruit 1,000 participants from the U.S. and
India using TurkPrime (Litman et al., 2017). After removing participants who failed any of our
four attention checks, 840 participants remained for statistical analyses (India: n = 346; U.S.: n =
494). All participants first completed the item pool (see Measures), then they completed MFQ,
and finally reported their demographic details. The present sample ranged in age from 18 to 77
years old (M = 34.24, SD = 11.02), and included an approximately equal number of men and
women (55.83% male). Most of our American sample identified as White (71.3%).

118

Measures. The measures used in the present study are described below.
Moral Foundations Questionnaire-2 Item Pool. We reviewed the extant MFT literature
as well as criticisms regarding some of the items in MFQ. We aimed to develop an initial item
pool with over 100 items all in a declarative form, similar to the “Judgments” part of the MFQ.
Since the “Relevance” items have been shown to reduce internal consistencies and they have
caused confusion among some researchers (e.g., by only using Relevance items rather than using
both Relevance and Judgments, a practice that should be avoided), we decided to jettison the
Relevance format. All authors met 7 times to finalize an item pool of 116 items. While there was
some disagreement regarding some items, all authors agreed that these 116 items are acceptable
candidates to measure 6 foundations (Care: 15 items; Equality: 18 items; Proportionality: 25
items; Loyalty: 19 items; Authority: 20 items; and Purity: 19 items). While we no longer have a
“Fairness” subscale in MFQ-2, some items did not clearly belong to either Equality or
Proportionality in the first round of data collection (e.g., “When the government makes laws, the
number one principle should be ensuring that everyone is treated fairly”). Therefore, we left
these items as they were to explore how they relate to new Equality and Proportionality items.
The response option was provided from 1 (Does not describe me at all) to 5 (Describes me
extremely well) based on our qualitative examination of different response options and
consultation with survey researchers (see Krosnick & Fabrigar, 1997). In this study, we also
provided the option for all participants to give feedback if any of the items were not
comprehensible, did not read well, or were otherwise unclear.
Moral Foundations Questionnaire. Participants completed the old 30-item MFQ
(Graham et al., 2011) which consists of two 15-item parts, namely Relevance and Judgments.
The first part measures the five foundations using the relevance individuals ascribe to each of the

119

foundations. Items on the Relevance section are rated along a 6-point Likert-type scale ranging
from 0 (Not at all relevant) to 5 (Extremely relevant). The Judgments section consists of
contextualized items that can gauge actual moral judgments related to the five moral foundations.
Items on the Judgments section are rated along a 6-point Likert-type scale ranging from 0
(Strongly disagree) to 5 (Strongly agree). The internal consistency coefficients were .66, .64, .79,
.78, and .86 for Care, Fairness, Loyalty, Authority, and Purity, respectively.
Political Ideology. All participants rated their political affiliation with the Republican
party or the Democratic party along a 7-point scale ranging from 1 (Strong Democrat) to 7
(Strong Republican). Another item was political conservatism rated on a scale ranging from 1
(Very Liberal) to 7 (Very Conservative). We averaged these two items to create a political-
orientation score, on which higher scores indicated more conservative political orientation. A
similar method was used in previous work for assessment of political ideology (Jost &
Thompson, 2000). The internal consistency of these two items was high in the American sample
(α = .90). In the Indian subsample, we only use the conservatism item to quantify political
ideology.
Religiosity. All participants self-reported their religious affiliation as well as religiosity
using the single item “On a scale from 0-10, how religious do you consider yourself?”
Participants indicated the level of religiosity using a slider ranging from 0 to 10. The Indian
subsample was significantly more religious (M = 6.73, SD = 2.69) than the American subsample
(M = 3.99, SD = 3.59), t = 12.66, Welch-corrected df = 834.08, p < .001.

120

Results and Discussion
We first examine descriptive statistics for each item. Some items had floor or ceiling
effects, indicated by high skewness. These items were considered for discarding. We also
performed a number of different factor analyses and reliability analyses to see which items hold
together well while keeping the breadth of each foundation. Specifically, we conducted
exploratory factor analyses (EFAs) for items belonging to each foundation. For Care, we
dropped three items as they did not reach adequate the 0.4 item-factor loading criterion, hence 12
items were selected to be used in Study 4.1b. Based on similar criteria and qualitative analysis of
items and feedback from participants and all authors, 14, 18, 16, 15, and 15 items were selected
to be administered in Study 4.1b for Equality, Proportionality, Loyalty, Authority, and Purity,
respectively. Cultural differences in each item as well as correlations between these items and
MFQ subscales are presented in Supplementary Materials. Therefore, we reduced our initial,
crude item pool of 116 items to a sharper and more focused set of 90 items for further data
collection and analysis in Study 4.1b.
Study 4.1b
Study 4.1b was conducted to refine the 90-item pool from Study 4.1a into a more fine-
grained MFQ-2 pool. To do this, we administered the 90 items to a large and diverse sample of
adults from three cultures, namely, India, the U.S., and Iran. We specifically chose Iran because
MFQ’s structure was particularly inconsistent with the structure typically observed in Western
cultures (Atari et al., 2020b) and because Iran is approximately culturally equidistant from both
WEIRD cultures (e.g., the U.S.) and developed Eastern countries (e.g., China) (Muthukrishna et
al., 2020). We used these data to select the next set of MFQ-2 items and conduct a preliminary
examination of the MFQ-2’s basic measurement properties and structure.

121

Participants and Procedure. We aimed to recruit 1,000 participants from the U.S. and
India using TurkPrime. We translated all 90 items into Farsi using the standard back translation
technique (Brislin, 1970) and aimed to recruit Iranian participants by advertising the survey link
on social media platforms. We did not have any a priori expectations for the number of
participants to be recruited from Iran since the survey was relatively long and we could not
compensate participants. After removing participants who failed any of our four attention checks,
971 participants remained for statistical analyses (India: n = 380; U.S.: n = 491; Iran: n = 100).
All participants first completed the 90-item pool, then they completed MFQ, and finally reported
their demographic details. The present sample ranged in age from 18 to 77 years old (M = 34.81,
SD = 16.92), and included an approximately equal number of men and women (53.26% male).
Most Americans identified as White (69.2%).
Measures. All participants completed the 90-item pool of MFQ-2 finalized in Study 4.1a
(Care: 12 items; Equality: 14 items; Proportionality: 18 items; Loyalty: 16 items; Authority: 15
items; Purity: 15 items). They then self-reported their political ideology, religiosity, and
demographic details. For political ideology, we used the two-item measure in Study 4.1a (α =
.88), used a single-item measure of conservatism in India and Iran (with slight wording
modification in Farsi for cultural fluency).
Results and Discussion
As in Study 4.1a, we examined all items’ descriptive statistics, checking potential ceiling
or floor effects in any of the cultures we had data from. After item analysis, we conducted
foundation-level EFAs across cultures (see Supplementary Materials). After item analysis and
EFAs, 19 items were discarded overall, leaving 71 items for administration in Study 4.1c.
Cultural differences in each item as well as correlations between these items and MFQ subscales

122

are presented in Supplementary Materials. In this study, we reduced our item pool to 71 items for
further analysis in Study 4.1c while making sure that items hold together well and correlate with
relevant MFQ foundations reasonably.
Study 4.1c
In Study 4.1c, we prepare the final item pool for our main cross-cultural data collection
effort in Study 4.2. We administer the 71-item pool in three cultures in order to further reduce
the number of items. Here, we chose three cultures with the highest feasible distance in order to
maximize the diversity of our samples. Based on Muthukrishna et al.’s (2020) WEIRD cultural
distance, we carefully chose the U.S., Ecuador, and China. Ecuador is culturally distant from
both the U.S. and China, is a Spanish-speaking country with relatively high diversity in people’s
languages and subcultures and remains one of the most understudied cultures in moral
psychology. Because the geography of Ecuador is so diverse, the lifestyles, principal work, and
economic structure of its population are also diverse. There are fishermen along the coasts,
cattlemen in the southern highlands, farmers on central highland slopes, and oil workers in the
Amazon. In addition, here we address one of the important limitations of our samples in Studies
4.1a and 4.1b, that is, relying on convenience sampling. Here, we recruit stratified national
samples mirroring national demographics in terms of gender, education, age, and political
ideology (as well as race/ethnicity in the U.S.). In addition, in this study we used psychometric
network methods to diversify our methodological toolbox while choosing the best-performing
items (Christensen et al., 2020).
Participants and Procedure. We aimed to recruit 1,500 participants from the U.S.,
Ecuador, and China using Qualtrics panels. We translated all items into Spanish and Mandarin
using the standard back translation technique (Brislin, 1970). Two independent bilingual

123

researchers double-checked the final items for cultural fluency. A third-party translation
company certified both translations. Participants who failed any of the three attention checks
were terminated, in order to achieve stratified samples. Overall, 1,549 participants remained for
statistical analyses (U.S.: n = 515; Ecuador: n = 517; China: n = 517). All participants first
completed the 71-item pool, then they completed MFQ, and finally reported their demographic
details. Participants also completed a few items at the end of the survey, related to another
project. The present sample ranged in age from 18 to 87 years old (M = 40.92, SD = 16.02), and
included an approximately equal number of men and women (49.90% male). Most Americans
identified as White (73.8%).
Measures. All participants completed the 71-item pool of MFQ-2 (Care: 10 items;
Equality: 10 items; Proportionality: 13 items; Loyalty: 13 items; Authority: 12 items; Purity: 13
items). As in Studies 4.1a and 4.1b, they then completed MFQ (α coefficients ranged between
.62 [Fairness] and .78 [Purity]), political ideology, religiosity, and demographic details. For
political ideology, we used the two-item measure in Studies 4.1a and 4.1b (α = .72), and used a
single-item measure of conservatism in Ecuador and China.
Results and Discussion
As in Studies 4.1a and 4.1b, we examined all 71 items’ descriptive statistics. We also
conducted foundation-level EFAs. Since our aim was for MFQ-2 to have 6 items per foundation
(similar to MFQ), we aimed to select 7 to 9 items for each foundation. Our aim in this study was
to combine item analysis, factor analysis, psychometric network analysis, and qualitative
examination of the breadth for each foundation’s items to avoid redundancy. We wanted the final
measure to adequately represent each foundation’s considerable bandwidth — rather than

124

narrowing the range of moral concerns assessed — in order to maintain the MFQ-2’s descriptive
and predictive breadth.
For Care, we dropped two items based on our qualitative examination of remaining items,
leaving 8 items for Study 4.2. For Equality, two items were discarded for having an item-factor
loading problem in at least one culture, and one item was discarded for similarity to another item,
leaving 7 items for Study 4.2. For Proportionality, two items were discarded based on
psychometric network analysis (centrality issue) and three items were dropped to increase item
diversity, leaving 8 items for inclusion in Study 4.2. For Loyalty, we dropped three items based
on EFA results, and discarded two items to reduce content redundancy. In addition, we added
one new item to Loyalty items (“It is more important to be a good team member than to express
oneself”) to test whether it can hang with other items in a desirable way, leaving a total of 9
items for inclusion in Study 4.2. For Authority, we discarded one item based on EFA results and
discarded two items due to content redundancy with other existing items, leaving 9 candidate
items for Study 4.2. Finally, for Purity, we discarded three items based on qualitative
examination of items’ content and discarded one item due to centrality issues in the psychometric
network analysis, leaving 9 items for administering in Study 4.2. Hence, in this study, we
selected 50 candidate items for translation and use in Study 4.2 across cultures, aiming for the
final MFQ-2 to have 36 balanced items. In short, we reduced our item pool to 50 items for
translation and cross-cultural administration in Study 4.2.
Study 4.2
We conducted Study 4.2 to finalize MFQ-2 based on the item pool constructed in Study
4.1. Our goal here was to (a) finalize the MFQ-2 items based on cross-cultural data; (b) establish
measurement invariance across groups and examine group differences in MFQ-2 scores; and (c)

125

examine the predictive power of MFQ-2 across cultures. To do this, we administered the final
50-item pool from Study 4.1 to a diverse sample of adults from 19 new countries, none of which
were sampled in Study 4.1. After examination of the MFQ-2’s basic measurement properties and
multidimensional structure, we combine data from Studies 4.1 and 4.2 to examine group
differences in each of the 6 scores MFQ-2 yields.
Participants and Procedure. We aimed to recruit stratified samples from diverse
cultural backgrounds. Based on Muthukrishna et al.’s (2020) cultural distance metric, we made a
list of candidate cultures. We then cross-referenced that with the feasibility of stratified data
collection administered by Qualtrics panels targeting 200 participants per culture. Overall, 19
cultures were selected. We collected nationally stratified samples from these 19 cultures (N =
3,902). Details about these samples are provided in Table 4.3. All measures were translated into
local languages using a third-party translating company. Then, independent bilingual researchers
checked the translations and made sure of the fluency of all items. Discrepancies and
modifications were addressed between the translation companies, independent researchers, and
the authors. All participants completed the 50-item pool and a few demographic questions.
Participants who failed any of the three attention checks were terminated from continuing the
survey.

126

Table 4.3.
Description of Samples Across 19 Cultures in Study 4.2
Culture n %female
Age: M
(SD)
Language
Sample’s majority
religion (%)
WEIRD cultural
distance [95%CI]
Argentina 205 48.8% 42.5 (15.0) Spanish Christianity (62.0%) .071 [.069, .075]
Belgium 205 49.8% 45.1 (17.0) French Christianity (47.8%) NA
Chile 205 49.3% 42.4 (16.2) Spanish Christianity (58.5%) .078 [.075, .081]
Columbia 205 48.8% 41.0 (15.0) Spanish Christianity (64.4%) .102 [.099, .106]
Egypt 205 49.8% 44.8 (16.8) Arabic Islam (94.1%) .234 [.228, .241]
France 206 49.0% 43.7 (16.9) French Christianity (48.5%) .079 [.075, .085]
Ireland 205 50.2% 44.8 (16.7) English Christianity (66.3%) NA
Japan 207 49.3% 47.2 (15.3) Japanese None (46.9%) .115 [.112, .119]
Kenya 205 48.3% 37.6 (12.4) English Christianity (85.4%) NA
Mexico 206 46.6% 41.9 (15.4) Spanish Christianity (53.4%) .077 [.074, .080]
Morocco 205 48.3% 41.8 (14.7) Spanish Islam (96.6%) .149 [.145, .155]
New Zealand 205 48.3% 47.4 (18.2) English None (47.3%) .053 [.050, .058]
Nigeria 205 41.4% 39.1 (13.6) English Christianity (76.6%) .130 [.126, .135]
Peru 205 37.6% 37.0 (13.8) Spanish Christianity (62.9%) .090 [.087, .094]
Russia 206 45.6% 41.7 (14.9) Russian Christianity (62.6%) .085 [.083, .088]
Saudi Arabia 207 48.3% 42.4 (15.5) Arabic Islam (96.1%) NA
South Africa 205 47.3% 41.3 (15.4) English Christianity (81.0%) .076 [.073, .079]
Switzerland 205 50.2% 46.7 (16.8) French Christianity (52.7%) .068 [.064, .074]

127

UAE 205 49.3% 43.1 (14.7) Arabic Islam (84.9%) NA

Measures. All participants first completed a few demographic questions: country of
residence, age, gender, and political ideology. Then they completed the 50-item pool of MFQ-2
prepared in Study 4.1. The order of questions was randomized. Participants then completed some
measures unrelated to this study, a single-item measure of religiosity, and demographic details.
For political ideology, we used a single-item measure, rated along a 10-point scale, that can work
equally well across cultures (“In political matters, people talk of ‘the left’ and ‘the right.’ How
would you place your views on this scale, generally speaking?”). A few other items, related to
another project, were also included at the end of the survey.
Analytic Strategy. Our statistical analyses of the data come in three separate but related
parts. In part 1, we use the Exploratory Structural Equations Modeling (ESEM) framework
(Asparouhov & Muthén, 2009; Marsh et al., 2014) as well as descriptive item analysis in order to
finalize the 36-item MFQ-2. ESEM is a synergy of EFA and CFA, incorporating the advantages
of both EFA and CFA. ESEM is effective in the psychometric examination of multidimensional
instruments and can easily be complemented with other modeling approaches. In the presence of
multidimensionality stemming from the assessment of conceptually related constructs (Morin et
al., 2016), it is possible that the restrictive assumptions of CFA are violated, and ESEM models
may outperform CFAs. In the second part, we conduct measurement invariance across all
cultures. To test measurement invariance, we use the Multi-Group Factor Analysis Alignment
method (or simply, “alignment”) which has been proposed as a new method to test metric and
scalar invariance (Asparouhov & Muthén, 2014). This method aims to address issues in Multi-

128

Group Confirmatory Factor Analysis (MGCFA) invariance testing, such as difficulties in
establishing exact scalar invariance with many groups (as is the case in the current work). The
main difference between MGCFA and alignment is that alignment does not require equality
restrictions on factor loadings and intercepts across groups. The base assumption of the
alignment method is that the number of non-invariant measurement parameters and the extent of
measurement non-invariance between groups can be held to an acceptable minimum for each
given scale through producing a solution that features many approximately invariant parameters
and few parameters with large non-invariances. Our ultimate goal is to compare latent factor
means of moral foundations across groups (here, cultures), therefore the alignment method
estimates factor loadings, item intercepts, factor means, and factor variances (Asparouhov &
Muthén, 2014). After measurement invariance is evidenced, we compare and contrast cultures
across the six dimensions of MFQ-2. We also examine the relationship between MFQ-2 scores
and WEIRDness cultural distance scores (Muthukrishna et al., 2020). In the third part, we
examine culturally variable sex, religious, and ideological differences. To do so, we rely on
multilevel models wherein participants are modeled as nested within groups.
Results and Discussion
Exploratory Structural Equations Models. We first conducted an ESEM on the
entirety of the data (CFI = .958, TLI = .958, RMSEA = .029, SRMR = .027) and discarded 14
items for having cross-loadings, while making sure that remaining items are not redundant in
content. We then conducted a secondary ESEM with the final 36 items on the whole data (CFI =
.979, TLI = .978, RMSEA = .024, SRMR = .023). All items and loadings are presented in Table
4.4. Accordingly, the final 36-item MFQ-2 has good structural validity across cultures.

129

Table 4.4.
Results of Exploratory Structural Equations Modeling (Study 4.2)
Item F1 F2 F3 F4 F5 F6
It pains me when I see someone
ignoring the needs of another human
being.
0.66 0.05 0.09 0.04 -0.02 -0.04
I am empathetic toward those people
who have suffered in their lives.
0.70 0.04 0.00 -0.01 0.01 0.05
I believe that compassion for those
who are suffering is one of the most
crucial virtues.
0.73 -0.01 0.00 -0.04 0.08 0.01
Caring for people who have suffered
is an important virtue.
0.73 -0.01 0.02 -0.01 0.05 0.04
We should all care for people who
are in emotional pain.
0.74 0.03 0.00 0.00 0.05 -0.01
Everyone should try to comfort
people who are going through
something hard.
0.64 0.05 0.04 0.07 0.03 0.00
I believe it would be ideal if
everyone in society wound up with
roughly the same amount of money.
0.01 0.81 0.00 -0.05 -0.01 0.01
When people work together toward a
common goal, they should share the
rewards equally, even if some
worked harder on it.
0.07 0.38 0.06 0.26 -0.02 -0.16
I believe that everyone should be
given the same quantity of resources
in life.
0.23 0.54 -0.02 0.07 -0.03 0.02

130

The world would be a better place if
everyone made the same amount of
money.
-0.04 0.88 -0.03 0.00 0.01 0.00
I get upset when some people have a
lot more money than others in my
country.
0.07 0.52 0.09 -0.3 0.07 0.12
Our society would have fewer
problems if people had the same
income.
-0.03 0.86 -0.04 0.04 0.00 -0.02
I feel good when I see cheaters get
caught and punished.
0.09 0.02 0.20 0.03 0.00 0.29
I think people should be rewarded in
proportion to what they contribute.
0.08 -0.04 0.03 0.06 0.09 0.54
I think people who are more hard-
working should end up with more
money.
0.00 0.02 0.01 0.01 0.04 0.72
It makes me happy when people are
recognized on their merits.
0.32 -0.07 -0.07 0.40 -0.09 0.27
In a fair society, those who work
hard should live with higher
standards of living.
-0.06 0.04 0.10 -0.04 0.01 0.72
The effort a worker puts into a job
ought to be reflected in the size of a
raise they receive.
0.13 0.06 -0.04 0.10 -0.04 0.53
I think children should be taught to
be loyal to their country.
0.00 0.00 0.78 0.12 -0.01 -0.01
I believe the strength of a sports team
comes from the loyalty of its
members to each other.
0.19 0.03 0.10 0.36 -0.04 0.13
Everyone should love their own
community.
0.15 0.06 0.37 0.21 0.08 0.01

131

Everyone should defend their
country, if called upon.
-0.02 -0.01 0.70 0.02 0.08 0.07
Everyone should feel proud when a
person in their community wins in an
international competition.
0.21 -0.03 0.27 0.28 -0.03 0.08
It upsets me when people have no
loyalty to their country.
0.04 0.00 0.83 -0.01 -0.04 -0.02
I feel that most traditions serve a
valuable function in keeping society
orderly.
-0.03 0.07 0.21 0.39 0.10 0.08
I think having a strong leader is good
for society.
0.10 -0.09 0.09 0.32 0.14 0.14
I think it is important for societies to
cherish their traditional values.
-0.04 0.05 0.22 0.44 0.06 0.08
I believe that one of the most
important values to teach children is
to have respect for authority.
-0.02 0.00 0.16 0.59 0.05 -0.03
I think obedience to parents is an
important virtue.
0.01 0.00 0.09 0.55 0.19 0.04
We all need to learn from our elders. 0.09 0.07 0.10 0.49 0.05 0.00
I believe chastity is an important
virtue.
0.05 -0.02 0.00 -0.01 0.84 0.02
I think the human body should be
treated like a temple, housing
something sacred within.
0.17 0.05 -0.02 0.37 0.21 -0.01
I admire people who keep their
virginity until marriage.
-0.01 0.01 -0.03 0.09 0.79 -0.01

132

People should try to use natural
medicines rather than chemically
identical human-made ones.
0.06 0.23 0.06 0.10 0.25 0.00
If I found out that an acquaintance
had an unusual but harmless sexual
fetish I would feel uneasy about
them.
-0.04 0.10 0.12 -0.05 0.48 0.03
It upsets me when people use foul
language like it is nothing.
0.11 -0.01 0.28 0.05 0.29 -0.01
Note. Relevant item-factor loadings are in bold.
Reliability of MFQ-2. Various reliability estimates have been proposed in the literature,
with the coefficient alpha (α) being the most prominent. However, coefficient α ignores the
measure’s internal factor structure, which should be inherent in choosing an appropriate
reliability estimate. Here, we report ωt coefficient, which by including the factor loadings in its
formula, is more suitable and stable for reporting internal structure and reliability of multi-item
scales since it corrects the underestimation bias of α when the assumption of tau-equivalence is
violated (Flora, 2020). In addition, different studies show that it is one of the best alternatives for
estimating reliability (Zinbarg et al., 2006; Revelle & Zinbarg, 2009). Here, we report
foundation-level ωt coefficients across 19 cultures (Table 5). As can be seen in Table 4.5, ωt
coefficients ranged between .73 and .95 (average ωt coefficients: Care = .90; Equality = .89;
Proportionality = .83; Loyalty = .89; Authority = .86; and Purity = .82). Hence, all six scores
computed by averaging items for the six foundations are internally consistent across cultures.

133

Table 4.5.
Omega Coefficients Across Foundations across Cultures
Country Care (ωt)
Equality
(ωt)
Proportiona
lity (ωt)
Loyalty
(ωt)
Authority
(ωt)
Purity (ωt)
Argentina 0.90 0.92 0.77 0.83 0.83 0.82
Belgium 0.92 0.90 0.78 0.88 0.80 0.83
Chile 0.92 0.88 0.82 0.90 0.88 0.83
Columbia 0.87 0.89 0.81 0.90 0.86 0.82
Egypt 0.89 0.87 0.81 0.87 0.86 0.83
France 0.92 0.90 0.80 0.90 0.83 0.76
Ireland 0.92 0.86 0.86 0.90 0.90 0.85
Japan 0.88 0.89 0.83 0.89 0.84 0.73
Kenya 0.89 0.85 0.85 0.89 0.89 0.82
Mexico 0.91 0.87 0.83 0.86 0.86 0.80
Morocco 0.91 0.88 0.89 0.89 0.86 0.82
New
Zealand
0.93 0.92 0.82 0.93 0.89 0.86
Nigeria 0.85 0.87 0.79 0.85 0.82 0.75
Peru 0.89 0.90 0.85 0.84 0.86 0.83

134

Russia 0.90 0.90 0.83 0.90 0.89 0.86
Saudi
Arabia
0.89 0.89 0.87 0.88 0.84 0.76
South
Africa
0.88 0.91 0.82 0.90 0.84 0.85
Switzerla
nd
0.90 0.95 0.83 0.91 0.89 0.84
UAE 0.94 0.89 0.93 0.91 0.90 0.87
Measurement Invariance. The alignment method proceeds in two steps (Asparouhov &
Muthén, 2014). In the first step an unconstrained configural model is fitted across all cultures. To
allow the estimation of all item loadings in the configural model, we fixed the factor means to 0
and the factor variances to 1. In the second step, we optimized the configural model using a
component loss function with the goal to minimize the non-invariance in factor means and factor
variances for each group (for a detailed mathematical description see: Asparouhov & Muthén,
2014). This optimization process terminates at a point at which “there are few large non-
invariant measurement parameters and many approximately non-invariant parameters rather than
many medium-sized non-invariant measurement parameters” (Asparouhov & Muthén, 2014, p.
497). Overall, the alignment method allows for the estimation of reliable means despite the
presence of some measurement non-invariance. Muthén and Asparouhov (2014) suggest a
threshold of 25% non-invariance as acceptable. The resulting model exhibits the same model fit
as the original configural model but is substantially less non-invariant across all parameters
considered. The percentage of non-invariant parameters in our invariance alignment method with
post-hoc item parameter constraints can be seen in Table 4.6. As can be seen, all foundations
except Purity meet the threshold of 25% non-invariance, meaning that scores on Care, Equality,

135

Proportionality, Loyalty and Authority can be reliably compared across cultural groups. For
Purity, caution should be practiced when comparing group-level means. In the present sample,
the source of non-invariance in Purity was mostly due to unique item intercepts in Argentina (6
unique parameters; 5.3%) and Chile (4 unique parameters; 3.5%). Among Purity items, the item
“I think the human body should be treated like a temple, housing something sacred within” was
most non-invariant with 10 unique parameters (8.7%). Hence, this item may elicit different
patterns of responding across different cultures.
Table 4.6.
The Measurement Invariance Alignment Results (Study 4.2)
Foundation Percentage of noninvariance
item parameters (loadings)
Percentage of noninvariance
item parameters (intercepts)
Care 0.0% 5.3%
Equality 0.0% 21.9%
Proportionality 0.0% 11.4%
Loyalty 0.0% 24.6%
Authority 0.0% 16.7%
Purity 2.6% 39.5%
Note. A threshold of 25% non-invariance as acceptable (Muthén & Asparouhov, 2014).
The Equality-Proportionality Link. One of the novel aspects of the present work is
theoretically distinguishing between Equality and Proportionality. If the two constructs are
distinct and psychometrically non-redundant, we should find only small to moderate correlations
between them. We examined the correlation between Equality and Proportionality across all 19

136

cultures and we found support for our prediction. Indeed, Equality and Proportionality were
weakly positively correlated (average Pearson correlation coefficient = .21, SD = .13). Equality
and Proportionality were most related to one another in the UAE (r = .47, p < .001) while the
smallest correlation was observed in Belgium (r = .04, p = .556). The correlations and their 95%
CI are visually presented in Figure 4.1. Based on these findings, Equality and Proportionality
may be considered orthogonal to one another, or only slightly positively correlated. For example,
in countries such as New Zealand, Belgium, and Switzerland, people’s scores on Equality do not
tell us anything about their concerns regarding merit and deservingness.
Figure 4.1.
The Correlations between Equality and Proportionality (Study 4.2)

137

Note. The error bars represent 95% Confidence Interval. The vertical line represents a zero
correlation.
Cultural Differences. After measurement invariance was evidenced, we proceeded to
examine cultural differences since theoretically predictable cultural variation is one of the
fundamental claims of MFT (see Table 4.7). Cultural differences in all foundations are visually
presented in Figure 4.2. We then examined the relationship between WEIRDness cultural
distance and culture-level moral foundations. Less WEIRD cultures had substantially higher
concerns for Purity (r = .80, 95%CI = [.47, .93], p < .001) and Loyalty (r = .54, 95%CI = [.01,
.83], p = .046). The correlations between WEIRDness cultural distance and other foundations
were not statistically significant (rs < .47, ps > .093).
Table 4.7.
Means and Standard Deviations of Moral Foundations Across 19 Cultures (Study 4.2)
country Care Equality
Proportio
nality
Loyalty Authority Purity
Argentina
3.84
(0.77)
2.81 (1.01)
3.91
(0.66)
3.58 (0.82) 3.67 (0.73) 2.6 (0.82)
Belgium
3.91
(0.73)
3.2 (0.94)
3.91
(0.56)
3.62 (0.77) 3.7 (0.64) 3.01 (0.74)
Chile
3.77
(0.82)
2.77 (0.88) 3.7 (0.69) 3.45 (0.88) 3.67 (0.81) 2.54 (0.85)
Columbia
3.83
(0.71)
2.91 (0.9)
3.69
(0.68)
3.67 (0.82) 3.84 (0.68) 2.98 (0.86)
Egypt
4.38
(0.60)
3.56 (0.94)
4.37
(0.58)
4.42 (0.62) 4.18 (0.68) 4.19 (0.63)

138

France
4.08
(0.68)
3.23 (0.92)
4.12
(0.54)
3.86 (0.74) 3.88 (0.62) 3.09 (0.74)
Ireland
4.01
(0.79)
2.94 (0.93)
3.73
(0.77)
3.29 (0.98) 3.49 (0.91) 2.51 (0.93)
Japan
3.03
(0.77)
2.27 (0.78)
3.14
(0.73)
2.66 (0.82) 2.67 (0.66) 2.63 (0.69)
Kenya 4.2 (0.77) 2.88 (0.97)
3.78
(0.79)
3.95 (0.90) 4.07 (0.80) 3.58 (0.83)
Mexico
3.77
(0.79)
2.87 (0.91) 3.8 (0.70) 3.78 (0.75) 3.94 (0.67) 2.81 (0.81)
Morocco
4.21
(0.78)
3.36 (0.97)
4.18
(0.71)
4.16 (0.82) 3.95 (0.76) 3.93 (0.73)
New
Zealand
3.84
(0.78)
2.61 (1.02)
3.61
(0.71)
3.22 (1.00) 3.48 (0.87) 2.58 (0.98)
Nigeria
4.32
(0.64)
2.9 (1.03)
4.14
(0.67)
4.11 (0.74) 4.21 (0.61) 3.8 (0.77)
Peru
3.62
(0.73)
2.63 (0.92)
3.75
(0.69)
3.73 (0.76) 3.81 (0.69) 3 (0.82)
Russia
3.96
(0.75)
3.24 (0.87)
4.27
(0.48)
3.87 (0.81) 3.68 (0.76) 3.25 (0.80)
Saudi
Arabia
4.24
(0.75)
3.32 (0.93)
4.18
(0.69)
4.2 (0.78) 4.07 (0.73) 3.98 (0.72)
South
Africa
4.21
(0.69)
3.01 (0.92)
4.03
(0.64)
3.85 (0.86) 4 (0.73) 3.4 (0.94)
Switzerla
nd
3.95
(0.68)
3.27 (0.98)
3.84
(0.64)
3.58 (0.85) 3.52 (0.81) 2.95 (0.79)
UAE
4.01
(0.92)
3.28 (0.93)
3.96
(0.89)
4.02 (0.91) 3.91 (0.89) 3.74 (0.85)

139

Figure 4.2.
Mean Endorsement of Moral Foundations and their 95% Confidence Interval (Study 4.2)

Sex Differences. In this section, we examined culturally variable sex differences in moral
foundations. Notably, only 1.3% of our sample (n = 50) identified as non-binary, hence we did
not have adequate statistical power to explore this population, and only included participants
identifying as either “woman” or “man.” Based on the findings of Atari et al. (2020c), we
expected to find female-favoring scores on Care and Purity. We did not have any a priori
predictions regarding sex differences in Equality and Proportionality. We estimated a random-
intercept model allowing countries to vary in sex differences in each of the foundations. For

140

Care, the fixed effect of sex was in line with our prediction, but was not statistically significant
(B = -0.03, SE = 0.024, p = .259), indicating that prior findings regarding sex differences in Care
are smaller when measured using MFQ-2 rather than MFQ. This might also be attributable to
some particular MFQ items tapping into neighboring constructs such as compassion and
nurturing tendencies, while MFQ-2 items are more focused on generic alleviation of pain and
suffering. Women scored substantially higher than men on Equality (B = -0.16, SE = 0.03, p <
.001) and Purity (B = -0.09, SE = 0.026, p < .001). Men, on the other hand, scored significantly
higher than women on Proportionality (B = 0.09, SE = 0.022, p < .001), Loyalty (B = 0.06, SE =
0.027, p = .038), and Authority (B = 0.06, SE = 0.024, p = .009).
Further, we calculated Mahalanobis’ D (and its 95% CI based on 10,000 bootstrap
iterations), which estimates the size of global (i.e., multivariate) sex differences (Del Giudice,
2009, 2019). Since D can overestimate sex differences in small samples and underestimate them
when using unreliable measurements, we corrected for both biases by calculating disattenuated,
bias-corrected difference, known as Dcu (Del Giudice, 2019). Multivariate sex differences in
moral foundations were smallest in France (Dcu = 0.357) and largest in Mexico (Dcu = 2.130).
Across 19 cultures, Dcu was large in size, M = 1.06, Md = 0.92, SD = 0.55. Larger Dcu values
indicate more sex differentiation in overall pattern of moral judgments (Atari et al., 2020c). We
examined the correlation coefficient between the WEIRDness cultural distance and Dcu and
found no evidence that WEIRDer countries show different multivariate sex differences (r = .43,
p = .124).
Religious Differences. We first examined moral foundations as a function of religious
affiliation. Since we did not have enough data on individuals affiliating with Juduism (n = 25),
Hinduism (n = 14), and “other” affiliations (n = 244) in our data, we excluded these participants,

141

leaving 3,618 individuals associating with Christianity (n = 1803), Islam (n = 909), Buddhism (n
= 91), and no religious affiliation (n = 815). One participant chose not to report their religious
affiliation. For Care, an ANOVA suggested significant between-religion differences (Welch-
corrected F = 73.70, ω
2
= 0.35, p < .001), such that Muslims (M = 4.18) scored higher than non-
religious individuals (M = 3.72, Holm-corrected p < .001), Christians (M = 4.03, Holm-corrected
p < .001), and Buddhists (M = 3.22, Holm-corrected p < .001). For Equality, there was a
significant difference between religious affiliations (Welch-corrected F = 67.82, ω
2
= 0.32, p <
.001), such that Muslims (M = 3.37) scored higher than non-religious individuals (M = 2.81,
Holm-corrected p < .001), Christians (M = 2.95, Holm-corrected p < .001), and Buddhists (M =
2.45, Holm-corrected p < .001). For Proportionality, there was a significant difference between
groups (Welch-corrected F = 63.92, ω
2
= 0.31, p < .001) with Muslims (M = 4.15) scoring higher
than non-religious individuals (M = 3.74, Holm-corrected p < .001), Christians (M = 3.92, Holm-
corrected p < .001), and Buddhists (M = 3.27, Holm-corrected p < .001). For Loyalty, there was
a significant difference between groups (Welch-corrected F = 202.81, ω
2
= 0.59, p < .001) with
Muslims (M = 4.16) scoring higher than non-religious individuals (M = 3.19, Holm-corrected p <
.001), Christians (M = 3.85, Holm-corrected p < .001), and Buddhists (M = 3.01, Holm-corrected
p < .001). For Authority, there was a significant difference between groups (Welch-corrected F =
198.21, ω
2
= 0.59, p < .001) with Muslims (M = 4.03) scoring higher than non-religious
individuals (M = 3.26, Holm-corrected p < .001), Christians (M = 3.95, Holm-corrected p =
.049), and Buddhists (M = 2.96, Holm-corrected p < .001). Finally, for Purity, there was a
significant difference between groups (Welch-corrected F = 538.10, ω
2
= 0.79, p < .001) with
Muslims (M = 3.94) scoring higher than non-religious individuals (M = 2.45, Holm-corrected p <

142

.001), Christians (M = 3.21, Holm-corrected p = .049), and Buddhists (M = 2.84, Holm-corrected
p < .001). These differences are shown in Figure 4.3.
Figure 4.3.
Endorsement of Moral foundations across Religious Affiliations (Study 4.2)

We then examined the role relationship between religiosity and all six foundations using
a cross-classified, random-intercept multilevel model wherein participants are nested in their
cultures (19 groups) as well as religions (4 groups). The results suggested that Care (B = -0.13,
SE = 0.063, p = .041) and Proportionality (B = -0.44, SE = 0.068, p < .001) were negatively
associated with religiosity, while Equality (B = 0.11, SE = 0.042, p = .007), Loyalty (B = 0.16,

143

SE = 0.068, p = .020), Authority (B = 0.21, SE = 0.079, p = .008), and Purity (B = 0.75, SE =
0.059, p < .001) were positively associated with religiosity.
Ideological Differences. We conducted a random-intercept, multi-level model to predict
political ideology based on all six moral foundations. We found Care (B = -0.25, SE = 0.070, p <
.001) and Equality (B = -0.57, SE = 0.048, p < .001) to be negatively correlated with political
conservatism, while Proportionality (B = 0.30, SE = 0.075, p < .001), Loyalty (B = 0.24, SE =
0.075, p = .001), Authority (B = 0.55, SE = 0.087, p < .001), and Purity (B = 0.13, SE = 0.063, p
= .039) were positively associated with right-wing ideology. Average endorsements of moral
foundations are plotted by political ideology in Figure 4.4. As a robustness check, we
dichotomized political ideology by categorizing people who scored 1-3 as “liberal” and 8-10 as
“conservative” and excluded middle-of-the-road individuals. We then conducted a logistic
multilevel model with the same specifications as above. Again, we found that Care (B = -0.31,
SE = 0.096, p = .001) and Equality (B = -0.76, SE = 0.069, p < .001) negatively predicted being a
conservative, while Proportionality (B = 0.33, SE = 0.101, p = .001), Loyalty (B = 0.31, SE =
0.103, p = .003), and Authority (B = 0.60, SE = 0.121, p < .001) positively predicted being a
conservative. Purity (B = 0.14, SE = 0.088, p = .102) was also in the same direction as before but
did not reach significance.
That Care is associated with liberal ideology, and that Loyalty, Authority, and Purity are
associated with conservative ideology are consistent with prior work (Graham et al., 2009;
Kivikangas et al., 2021). We also present novel findings with regard to the differential
relationships between two novel subscales and political ideology. In line with our theorizing and
prior work, we find that liberals are more concerned with Equality and conservatives are more
concerned with Proportionality. When middle-of-the-road individuals were dropped and we

144

compared highly liberal and highly conservative participants, all relationships held, except for
Purity. These findings suggest that our efforts in “de-politicizing” Purity items (while keeping
fidelity to the conceptual breadth of the concept) have worked.
Figure 4.4.
Average Endorsement of Moral Foundations Plotted by Political Ideology (Study 4.2)

Note. On the x-axis, 1-5 are blue-shaded (representing liberal-leaning individuals) and 6-10 are
red-shaded (representing conservative-leaning individuals). Solid blue = Care; Dashed Purple =
Equality; Dashed gray = Proportionality; Solid red = Loyalty; Solid orange = Authority; Solid
pink = Purity.
Individualizing vs. Binding Foundations. On an exploratory basis, we examined how
the individualizing vs. binding distinction looks like using the new 6-dimensional model. Given
some recent methodological reservations about higher-order CFA (see Lee & Cadogan, 2013),
we relied on Exploratory Graph Analysis (EGA; Golino & Epskamp, 2017) to estimate the

145

number of higher-order dimensions in MFQ-2. Since Equality and Proportionality were not
present in Graham et al. (2009), we performed community detection analyses to examine which
moral foundations strongly cluster together. We used the “walktrap” algorithm for community
detection as it assigns nodes to a single cluster, has been demonstrated to yield reliable results
(Pons & Latapy, 2006), and performs well on self-report data (Golino & Epskamp, 2017). We
ran the walktrap algorithm via EGA. We estimated the Gaussian graphical model using graphical
least absolute shrinkage and selection operator (GLASSO; Friedman et al., 2008) with extended
Bayesian information criterion to select optimal regularization parameter. Similar to latent-
variable modeling (applied in Graham et al., 2009), EGA identifies the grouping of nodes (here,
foundations) within a network; however, it either outperforms or is equal to other dimension
estimating methods (e.g., parallel analysis, Kaiser-Guttman rule; Golino & Epskamp, 2017).
Moreover, network analysis provides additional information about the relations among moral
foundations while controlling for all possible relationships between pairs of foundations. Finally,
since prior work shows that higher-order networks of moral foundations may differ between
cultures (Atari et al., 2020b), we ran 19 different EGAs for the 19 cultures we had data from.
All exploratory networks are presented in Figure 4.5. In all networks, γ and λmin values
were set to 0.5 and 0.1, respectively. The EGA analyses revealed one dimension in 15 cultures
and two dimensions in four cultures (Ireland, New Zealand, and Peru). Hence, the
individualizing-binding distinction may not be how moral foundations are organized universally;
rather the inter-relations between the foundations should be considered culture-dependent. These
cultural differences are in line with the findings of Atari et al. (2020b) and Turner-Zwinkels et al.
(2021), demonstrating that moral foundations’ network differs between groups. In the three
cultures in which we found a two-dimensional network, there was a somewhat consistent pattern.

146

In Ireland and New Zealand, Care and Equality formed one dimension and the rest of the
foundations formed a second dimension. In Peru, however, we found a dimension underlying
Equality and Purity, while the rest of the foundations formed a second dimension. In all these
models with two dimensions, the two sub-networks were moderately related to one another, and
we found no evidence for complete segregation of these sub-networks. Accordingly, future
research using MFQ-2 should be mindful of the cultural context when using higher-order
dimensions proposed by Graham et al. (2011) based on latent-variable models based on primarily
North American and English-speaking participants.

147

Figure 4.5.
Higher-Order Networks Displaying the EGA-Identified Dimensions

Study 4.3
Study 4.3 was designed with three aims: (a) to establish the convergence of MFQ-2
scores with those of MFQ (Graham et al., 2011), (b) to examine substantive relations with and
capacity to predict criterion variables; and (c) to compare the predictive power of MFQ-2 and
MFQ in predicting the amount of variance in external scale scores. We selected three external

148

scales as criterion variables for each foundation (see Measures). As such, Study 4.3 provides
evidence that MFQ-2 accurately quantifies its intended latent constructs (i.e., six moral
foundations), shares theoretically appropriate associations with other variables in moral
foundations’ nomological network, and does not capture confounding extraneous constructs.
Participants and Procedure. Since there were 18 criteria tested in this study, it was not
practically feasible to have all participants complete all measures. Therefore, we collected six
different samples, in which participants completed both MFQ-2, MFQ, along with a battery of
criterion scales, theorized to lie within moral foundations’ nomological network. We aimed to
collect a sample of 1,500 participants from the U.S., India, and Canada on TurkPrime. After
removing participants who failed any of the three attention checks, 1,410 participants remained
for analysis, mostly from the U.S. (82.1%). In terms of gender distribution, 642 participants
identified as women, 762 identified as man, and 6 identified as non-binary. Among American
participants, most individuals identified as White (75.7%). Based on our theoretical framework
and prior research, we predicted 18 relationships. The measures we used across these 6 samples
appear below.
Moral Foundations Questionnaire-2 (MFQ-2). We administered the 36-item MFQ-2
developed in Studies 4.1 and 4.2 (see Appendix 1). All 36 items were rated along a 5-point scale
ranging from 1 (Does not describe me at all) to 5 (Describes me extremely well). In the present
sample, the α coefficients were .89, .87, .78, .85, .87, and .86, for Care, Equality, Proportionality,
Loyalty, Authority, and Purity, respectively.
Moral Foundations Questionnaire (MFQ; Graham et al., 2011). All participants
complete the MFQ. Respondents rated the Relevance items provided using a 6-point scale from 0
(Not at all relevant) to 5 (Extremely relevant). The Judgements items were rated along 0

149

(Strongly disagree) to 5 (Strongly agree). In the present sample, the internal consistency
coefficients were .70, .67, .84, .81, and .87 for Care, Fairness, Loyalty, Authority, and Purity,
respectively.
Schwartz Values Survey (SVS; Schwartz, 1992). The SVS identifies ten personal values.
We report some of the SVS values that were previously used to examine the criterion validity of
MFQ by Graham et al. (20211). All items were rated from -1 (Opposed to my values) to +5 (Of
supreme importance), where 0 indicates this value is “not important” for the person.
Interpersonal Reactivity Index (IRI; Davis, 1983). We used the Empathic Concern
subscale of the IRI. Scores on this subscale are computed by averaging five items. This subscale
measures other-oriented feelings of compassion for the misfortune of others (e.g., “I often have
tender, concerned feelings for people less fortunate than me”). Items were rated along a 5-point
scale ranging from 1 (Does not describe me well) to 5 (Described me extremely well). In the
present sample, the internal consistency coefficient was .71.
Levenson Self-Report Psychopathy Scale (LSRPS; Levenson et al., 1995). The LSRPS
was developed to assess psychopathic traits and behaviors in the general population. The scale
includes 26 items rated along a 4-point Likert-type scale from 1 (Strongly disagree) to 4
(Strongly agree). It was developed to reflect the dual-factor model of psychopathy, assessing
primary psychopathy characterized by emotional deficits and manipulative behavior, and
secondary psychopathy, reflecting impulsivity, and antisocial behavior. An example item is “I
find myself in the same kinds of trouble, time after time.” In the present sample, the overall
internal consistency coefficient was .92.

150

Support for Redistribution Scale (SRS; Peterson et al., 2013). We used the 6-item SRS
to measure participants’ support for economic redistribution. All 6 items were rated along a 7-
point scale from 1 (Strongly disagree) to 7 (Strongly agree). An example item was “The
government should increase taxes and thus give more help to the poor.” In the present sample,
the internal consistency coefficient was .78.
Social Dominance Orientation (SDO; Ho et al., 2015). We used the extensively
validated 16-item SDO-7 Scale (Ho et al., 2015), responding to items such as “An ideal society
requires some groups to be on top and some to be on the bottom” (1 = Strongly oppose, 7 =
Strongly favor). In the present sample, the internal consistency coefficient was .91.
Preference for the Merit Principle Scale (PMPS; Davey et al., 1999). We used the
PMPS, which assesses the extent to which people believe that outcomes and resources should be
distributed based on qualifications or achievements rather than other determinants such as need
or seniority. Representative items include “Qualifications ought to be given more weight than
seniority when making promotion decisions”. Items were rated along a 7-point scale ranging
from 1 (Strongly disagree) to 7 (Strongly agree). In the present sample, the internal consistency
coefficient was .70.
Belief in a Just World (BJW; Dalbert, 1999). We measured Belief in a Just World with
Dalbert's (1999) General (i.e., BJW-other) BJW subscales, which has 6 items. Items were rated
along a 7-point scale ranging from 1 (Strongly disagree) to 7 (Strongly agree). An example item
is “I am confident that justice always prevails over injustice.” In the present sample, the internal
consistency coefficient was .85.

151

Group Loyalty Scale (GLS; Beer & Watson, 2009). We measured group loyalty using
the GLS which has 8 items (e.g., “I would describe myself as a team player”). Items were rated
along a 5-point scale ranging from 1 (Strongly disagree) to 5 (Strongly agree). In the present
sample, the internal consistency coefficient was .92.
Individualism and Collectivism Scale (ICS; Triandis & Gelfand, 1998). Individualism
and collectivism were measured using Triandis and Gelfand’s (1998) scale. Participants rated the
extent to which 16 items described them. All items were rated along a 5-point scale ranging from
1 (Never) to 5 (Always). Four items measured vertical individualism (e.g., “It is important that I
do my job better than others”), four measured horizontal individualism (e.g., “My personal
identity, independent of others, is very important to me”), four measured vertical collectivism
(e.g., “It is important to me that I respect the decisions made by my groups”), and four measured
horizontal collectivism (e.g., “I feel good when I cooperate with others”). Here we only report a
composite collectivism score (α = .85).
Right-Wing Authoritarianism (RWA; Altemeyer, 2006). The RWA scale measures the
degree to which people defer to established authorities, show aggression toward out-groups when
authorities sanction that aggression, and support traditional values endorsed by authorities. We
used the most recent version of the RWA scale (Altemeyer, 2006) which has 22 items. In
completing the RWA scale, participants respond to a series of statements (e.g., “Women should
have to promise to obey their husbands when they get married”) on a nine-point scale ranging
from 1 (Strongly disagree) to 9 (Strongly agree). In the present sample, the internal consistency
coefficient was .94.
Disgust Scale-Revised (DS-R; Olatunji et al., 2007). The DS-R is a revised version of
the 32-item Disgust Scale (Haidt et al., 1994). The DS-R consists of 25 items that measure how

152

disgusting people find various concepts. The scale consists of three subscales: contamination
disgust, animal remainder disgust, and core disgust. In the first part of the measure, people
indicate their agreement with items along a 5-point scale ranging from 1 (Strongly disagree) to 5
(Strongly agree). In the second part of the measure, participants indicate how disgusting an
experience would be (1 = Not disgusting at all; 5 = Extremely disgusting). Here, we report an
overall disgust sensitivity score (α = .86).
Duke University Religion Index (DUREL; Koenig et al., 1997). The DUREL is a five-
item measure developed for assessment of three main aspects of religiosity: Organized religious
activities (1 item), non-organizational religious activities (1 item), and intrinsic religiosity (1
item). The first two items are rated along a 6-point scale ranging from 1 (Never) to 6 (More than
once a week/day). The last three items, however, are rated along a five-point Likert-type scale
ranging from 1 (Definitely not true) to 5 (Definitely true of me). Total scores of the DUREL can
range between 5 and 27. In the present sample, the internal consistency coefficient was .92.
Left-Wing Authoritarianism (LWA; Costello et al., 2021). LWA has been
conceptualized as authoritarianism (e.g., aggression, submission, conventionalism) among
individuals who oppose traditional established hierarchies of moral and practical authority.
Despite Right-Wing Authoritarianism receiving considerably more attention in the moral
psychology literature, the conceptualization and measurement of LWA has only recently been
done (Costello et al., 2021). We used the 39-item measure of LWA (e.g., “If I could remake
society, I would put people who currently have the most privilege at the very bottom”). All items
were rated along a 7-point scale ranging from 1 (Strongly disagree) to 7 (Strongly agree). In the
present sample, the internal consistency coefficient was .95.

153

Demographics. At the end of the survey, all participants completed a set of demographic
questions including age, gender, education, religious affiliation, political ideology, and country
of residence. All these questions were identical to those administered in Study 4.2.
Analytic Strategy. First, we examine the correlations between MFQ-2 scores and MFQ
scores using Pearson correlations. We also used a linear model to tease apart unique relationships
between MFQ foundation scores and MFQ-2 scores. Second, we examine the correlations
between MFQ-2 foundation scores and the 18 criterion variables (3 per foundation). Third, we
broke down all external measure scores to their relevant subscale scores and used R-squared to
quantify and compare the predictive power of both MFQ-2 and MFQ in predicting these subscale
scores. In this way, we examined how powerful MFQ-2 and MFQ are in predicting related
psychological variables in their nomological network. Finally, since we did not have data from
the U.S. in Study 4.2, we replicated our foundation-level network analysis using exclusively
American participants.
Results
Convergence with MFQ. The correlation coefficients between MFQ-2 foundation scores
and MFQ foundations scores are summarized in Table 4.8. As can be seen, all foundations
strongly relate to their predecessor subscale. In the case of Fairness, it appears that MFQ’s
Fairness captures both Equality and Proportionality, its relationship to Equality was stronger than
that of Proportionality. This makes sense because some of the items in MFQ directly tap into
judgments about equality of outcomes (e.g., “I think it’s morally wrong that rich children inherit
a lot of money while poor children inherit nothing”). However, it is noteworthy that MFQ’s
Fairness scores are moderately correlated with both Equality and Proportionality, positively. Of
note, the correlation between Equality and Proportionality in the present sample was r = .02, p =

154

.400, consistent with the results of Study 4.2 wherein we found that in more WEIRD countries,
these two constructs tend to be more orthogonal compared with less WEIRD countries.

Table 4.8.
The correlations and regression coefficients between MFQ-2 and MFQ scores
MFQ-Care
MFQ-
Fairness
MFQ-Loyalty
MFQ-
Authority
MFQ-Purity
MFQ-2-Care .57***/.51*** .45***/.12*** -.02/-.08* -.01/-.10* .02/.04
MFQ-2-
Equality
.25***/.04 .33***/.29*** .19***/.28*** .09**/-.25*** .14***/.08*
MFQ-2-
Proportionalit
y
.20***/.08* .20***/.10** .23***/-.07 .31***/.35*** .24***/-.01
MFQ-2-
Loyalty
.13***/.04 .07**/-.07** .70***/.46*** .67***/.27*** .59***/.04
MFQ-2-
Authority
.11***/.06* .01/-.14*** .64***/.16*** .70***/.46*** .63***/.17***
MFQ-2-
Purity
.11***/-.01 .05/-.06* .64***/.14*** .65***/.08* .76***/.60***
Note. Figures on the left side of the dash represent bivariate Pearson correlation and
figures on the right side of the dash represent standardized regression coefficients in which all
MFQ foundation scores are accounted for.

155

Nomological Network. The correlation coefficients between MFQ-2 foundation scores
and criterion variables are presented in Table 4.9. Out of our 18 predicted relationships, 17 were
supported. The only correlation inconsistent with our predictions was between MFQ-2’s
Authority and Left-Wing Authoritarianism (r = .06, p = .355). The correlation was similar when
we only examined U.S. participants (r = .06, p = .394). Since this scale is mostly focused on anti-
authority and anti-tradition sentiment (e.g., “Certain elements in our society must be made to pay
for the violence of their ancestors”; see Costello et al., 2021), we predicted a negative
relationship; however, we observed a positive, non-significant relationship. Other correlations
supported the notion that MFQ-2’s foundations have substantive relations with criterion
variables.
People who score highly on MFQ-2’s Care show higher levels of empathic concern, take
benevolence to be a guiding principle in their lives, and are less likely to have psychopathic
traits. People who score highly on MFQ-2’s Equality show substantial support for redistributing
resources in the society, have substantially less desire for some groups to be actively oppressed
by others, have a stronger preference for intergroup equality, and consider social equality as a
guiding principle in their life. People who score highly on MFQ-2’s Proportionality consider
success as an important guiding principle to navigate their life, have a strong preference for
merit, and believe that the world is generally a fair and orderly place wherein what happens to
people is what they deserve. People who score highly on MFQ-2’s Loyalty tend to value
nationality and loyalty, tend to meet the duties and obligations of one’s social role to maintain
group harmony, and report to have remained loyal to their ingroup. People who score highly on
MFQ-2’s Authority tend to consider respect and obedience as important virtues as guiding
principles and tend to value authoritarian submission, authoritarian aggression, and

156

conventionalism. Finally, people who score highly on MFQ-2’s Purity tend to report higher
levels of sensitivity toward disgusting things (e.g., animal remains, corpses, rotten food), value
self-discipline and cleanliness, and report higher frequency of attending religious rituals, both
organizationally (e.g., in a church), and non-organizationally (e.g., saying prayers at home).

Table 4.9.
The Correlation Coefficients between MFQ-2 Scores and Criterion Variables
Criterion Variables Care Equality Proportion
ality
Loyalt
y
Authorit
y
Purity
Empathic Concern 0.63*** -0.01 0.14* -0.03 -0.09 -0.2**
SVS: Benevolence 0.50*** 0.13 0.35*** 0.39*** 0.48*** 0.39***
Psychopathy -0.30*** 0.32*** -0.06 0.21** 0.18** 0.43***
Support for Redistribution 0.03 0.56*** -0.1 0.05 -0.04 0.14*
Social Dominance
Orientation
-0.36*** -0.18** 0.06 0.36*** 0.4*** 0.5***
SVS: Equality 0.51*** 0.29*** 0.23*** 0.11 0.09 0.12
SVS: Success 0.09 0.01 0.22*** 0.37*** 0.42*** 0.31***

157

Preference for the Merit
Principle
0.26*** 0.42*** 0.50*** 0.5*** 0.44*** 0.47***
Belief in a Just World -0.03 0.14* 0.29*** 0.51*** 0.53*** 0.53***
SVS: Loyalty 0.32*** -0.03 0.39*** 0.5*** 0.58*** 0.4***
Collectivism 0.40*** 0.18** 0.39*** 0.59*** 0.60*** 0.42***
Group Loyalty 0.02 0.03 0.34*** 0.78*** 0.7*** 0.6***
SVS: Authority 0.12 0.15* 0.33*** 0.71*** 0.76*** 0.68***
Right-Wing
Authoritarianism
-0.32*** -0.03 0.2** 0.61*** 0.69*** 0.73***
Left-Wing Authoritarianism -0.03 0.58*** 0.02 0.10 0.06 0.30***
SVS: Purity 0.21** 0.22*** 0.25*** 0.53*** 0.6*** 0.72***
Disgust Sensitivity 0.17* 0.23*** 0.26*** 0.3*** 0.31*** 0.40***
Religiosity 0.01 0.18** 0.11 0.5*** 0.53*** 0.72***
Note. ***p < .001, **p < .01, *p < .05

158

Predictive Power. We used both MFQ’s and MFQ-2’s scores in predicting subscal-level
scores of all external measures. We collectively used 30 scores from SVS (Self-Transcendence,
Conservation, Self-Enhancement, and Openness to Change), LWA (Anti-Hierarchical
Aggression, Anti-Conventionalism, and Top-Down Censorship), Empathic Concern, Group
Loyalty, LSRP (Primary Psychopathy and Secondary Psychopathy), BJW, DSR (Core Disgust,
Animal Remainder, Contamination), Support for Redistribution, ICS (Horizontal Individualism,
Vertical Individualism, Horizontal Collectivism, and Vertical Collectivism), SDO (Pro-
Dominance, Con-Dominance, Pro-Antiegalitarianism, Con-Antiegalitarianism), Preference for
the Merit Principle, RWA, DUREL (Organizational Religiosity, Non-Organizational Religiosity,
and Intrinsic Religiosity), and Political Orientation. Across 30 regressions, MFQ-2 explained, on
average, 37% of the variance in outcome variables (Md = 38%); however, MFQ predicted, on
average, 30% of the variance in all outcomes (Md = 26%). The distribution of R
2
values and
inferential statistics are presented in Figure 4.6. A paired t-test indicated that MFQ-2 could
explain significantly more variance in outcomes compared with MFQ (t = 3.30, p = .003,
Hedges’s g = .59).

159

Figure 4.6.
The Predictive Power of MFQ and MFQ-2 in Predicting Outcomes

Replicating the Individualizing-Binding Distinction. As we demonstrated in Study 4.2,
the Individualizing-Binding distinction may depend on the cultural context and should not be
treated as a universal structure of moral foundations. Since we did not have data from the U.S. in
Study 4.2, and many MFT-relevant studies have relied on American participants, we explore this
higher-order structure using the same method as in Study 4.2, that is, network analysis. Using
exclusively American participants (n = 1,157), we found that Care and Equality form a

160

dimension while Loyalty, Authority, and Purity form a second dimension. Interestingly,
however, Proportionality was disconnected from the network, even though it was moderately
positively linked with Authority, Loyalty, and Care, while negligibly inversely linked with
Equality (see Figure 4.7). Accordingly, future research in the U.S. may consider the Care-
Equality dimension as the “individualizing island”, the Loyalty-Authority-Purity dimension as
the “binding island”, and Purity as the “bridge”.
Figure 4.7.
Higher-Order Network Displaying the EGA-Identified Dimensions in the U.S. (Study 4.3)

161

General Discussion
MFT (Haidt & Joseph, 2004; Graham et al., 2013) was developed by selecting the closest
links between evolutionary accounts of human sociality and anthropological accounts of the
breadth and variability of the moral domain (Fiske, 1992; Shweder et al., 1997). The original
operationalization of MFT offers five moral foundations (Care, Fairness, Loyalty, Authority, and
Purity). For the past decade, the Moral Foundations Questionnaire has been the primary tool with
which these foundations have been measured (Graham et al., 2011). Here, we revisited the
assumptions and conceptualization of MFT and, based on data from 24 cultures, we developed a
new tool, MFQ-2, which proves to be psychometrically adequate across these cultural settings.
We had five major goals in mind: (a) refining MFT’s view on Fairness by breaking it into
Equality and Proportionality, and incorporating this theoretical refinement into MFQ-2; (b)
development and validation of MFQ-2 across cultures using local languages, by combining
convenience and stratified samples; (c) empirically testing the structural validity and
comparability of MFQ-2 scores across cultures to make sure that MFQ-2 is truly a cross-
culturally meaningful and pragmatic tool; (d) conceptually replicating prior work on group
differences (cultural, ideological, sex, and religious differences) using the novel MFQ-2; (e)
establishing external validity of the MFQ-2 by examining associations between criterion scales
meant to capture relevant constructs, hence broadening the nomological network of moral
foundations using the new measure.
In three consecutive phases (see Flake et al., 2017), we report how MFQ-2 fares in
capturing the moral domain. We aimed to have 6 items per subscale, as is the case in MFQ
(Graham et al., 2011). In Studies 4.1a, 4.1b, and 4.1c, we finalized a 50-item pool based on data
from diverse cultural backgrounds (India, Iran, Ecuador, China, U.S.). It is crucial for a true non-

162

WEIRD science of morality to start from non-WEIRD contexts in order to make sure that our
measurements are not WEIRD themselves (Atari et al., 2021). This has been encouraged by
theoretical roots of MFT (Schweder & Haidt, 1993), but has remained ignored mostly because of
lack of easy access (or expensive access) to non-WEIRD populations (see Moshontz et al.,
2018). In Study 4.2, we diversified our samples even more, by recruiting nationally stratified
data from 19 cultures, most of which remain understudied in social and personality psychology
(Thalmayer et al., 2021). We test structural validity of MFQ-2, its measurement invariance, and
group differences in endorsement of moral foundations across these 19 cultures. In Study 4.3, we
examine how moral foundations, measured using MFQ-2, relate to relevant constructs and what
pattern of relationships may replicate findings based on the old MFQ scores.
Equality and Proportionality as Distinct Paths to Understanding Fairness
Our principal theoretical revision in this article is revisiting the concept of Fairness in
light of recent empirical findings in psychology, anthropology, economics, and political science.
We break down Fairness to more narrowly defined constructs in order to sharpen MFT’s view on
Fairness. We defined Equality in terms of a motive for balanced reciprocity, equal treatment,
equal say, and equal outcome. Proportionality, on the other hand, is a psychological mechanism
concerned with rewards and punishments to be proportionate to merit and deservingness, and
benefits to be calibrated to the amount of contribution. While we are hardly the first to recognize
these flavors of Fairness, our novel contribution is simultaneously incorporating both into MFT,
and measuring both in the new measure.
In our scale-development procedure, we made sure that (a) items representing these two
constructs were not Eurocentric (achieved by recursively soliciting feedback from a diverse
group of social and personality psychologists, see Medin et al., 2010); and (b) items were not

163

written with a particular political tone, which may inflate foundations’ correlation with political
ideology (e.g., some MFQ’s Fairness items have been shown to be particularly relevant in the
American political tenor, which may have contributed to especially strong correlations between
foundation scores and political ideology; see Kivikangas et al., 2021).
Notably, “addition” of foundations should come as no surprise; MFT theorists have
explicitly welcomed new foundations to be added to their framework as methods and theory co-
develop in moral psychology. Specifically, with regard to addition of new foundations, Graham
et al. (2013, p.58) rhetorically posited that they “do not know how many moral foundations there
really are. There may be 74, or perhaps 122, or 27, or maybe only 5, but certainly more than
one.” Graham et al. (2011) posited that what their map of the moral domain originally offered
(the five foundations) was “surely incomplete” (p. 382). These authors proposed that their
empirical support for the theory was a good initial map of the major moral continents; however,
“it is quite possible that later research, using different items or different methods, would reveal
that one of these continents is, like Eurasia, really two continents” (Graham et al., 2011, p. 382).
That is exactly what we have found and proposed in the current work, taking one more step
toward mapping the moral domain fully. This can open doors to many future investigations and
novel theoretical questions. This proposition is a direct response to Graham et al.’s (2011)
speculation that “whether a single foundation underlies intuitions about equality of opportunities
and those about equality of outcomes [remains an open question]” (p. 382).
Moral Foundations Questionnaire-2
In the past few years, MFQ has been rightly subjected to psychometric criticism
regarding its structural validity as well as internal consistency, especially in diverse, non-
Western samples (e.g., Atari et al., 2020b; Davis et al., 2017; Harper & Rhodes, 2021; Iurino &

164

Saucier, 2020; Moreira et al., 2019). In most of these studies, the original factor structure was not
replicated, and foundation-level internal consistency coefficients were lower than conventional
thresholds. This clearly called for MFT to have a psychometrically superior, and truly cross-
cultural and cross-linguistic, instrument, particularly because poor measurement qualities of
common measures in social and personality psychology are central culprits to the replication
crisis, something that in its own account is sometimes referred to as the measurement crisis.
Neglecting to address potential issues with measurement quality of common scales in social and
personality psychology has indeed helped contribute to the replication crisis in psychological
science (Flake & Fried, 2020).
In the entirety of the process of item reduction, we avoided relying on a single culture to
avoid cultural biases shaping the final battery of items in any form. The final 36-item MFQ-2
was not only developed with a diverse set of participants (Henrich et al., 2010) and by a diverse
set of researchers (Medin et al., 2010), but also different methodological strategies each of which
has its own benefits and limitations. This multi-methodological approach pushes against biases
and inclinations inherent in particular methodological choices. For example, ESEM balances the
advantages and disadvantages of EFA and CFA, Item-Response-Theory-based methods such as
the alignment method alleviate concerns about CFA-based methods in testing measurement
invariance across many groups, and network psychometrics is a helpful toolbox to complement
classical test theory (Golino et al., 2020).
In sum, MFQ-2 has desirable psychometric properties across almost all cultures from
which we had data in the current research. In terms of structural validity, different factor-analytic
approaches were used and converged on good fit indices for the a priori six-dimensional model
we hypothesized. In terms of internal structure, omega coefficients were adequate across 19

165

cultures for Care (M = .90, SD = .02), Equality (M = .89, SD = .02), Proportionality (M = .83, SD
= .04), Loyalty (M = .89, SD = .03), Authority (M = .86, SD = .03), and Purity (M = .82, SD =
.04). MFQ-2 scores also proved to be meaningfully comparable across cultures as measurement
invariance was evidenced, yet caution should be practiced when comparing Purity scores (more
so than other subscales), which can be attributable to patterns of responding (rather than latent
scores) being variable across cultures. For example, it might be the case that in some cultures,
strong norms around bodily and spiritual contamination encourage individuals to morally
condemn Purity transgressions for the sake of conformity. Broadly, the measurement invariance
evidence for MFQ-2 supports comparability of true foundation scores across cultures, something
we explored in this research in relation to WEIRDness, exploring what moral foundations are
more salient in WEIRD and non-WEIRD cultures.
Non-WEIRD Morality
Graham and colleagues (2011) contended that “one does not need to travel to non-
Western nations to find [MFT’s] broader conception of morality” (p. 380); however, we argue —
in contrast — that one certainly needs to collect high-quality data from non-Western nations, and
by non-Western researchers, to ascertain that moral psychological theories hold firmly across
various human populations, not just a small slice of human populations. This was our motivation
in recruiting a diverse group of participants across our studies. MFT was created as an
evolutionarily informed cultural theory of human morality, hence, it is imperative that its claims
be tested across non-WEIRD cultures and languages.
Breaking down Fairness into Equality and Proportionality may be considered one
theoretical step toward better understanding fairness concerns among cultures. We found, for
example, that Indians cared more about Proportionality than did their American counterparts in

166

Study 4.1. For example, Starmans et al. (2017) posit that “outside of the United States and
Europe [...] there are wide differences in fairness concerns across world cultures” (p. 3),
concluding that the preference for both equality and proportional outcomes are predominant in
many cultures.
Notably, only recently has it become possible to test non-WEIRD psychological
constructs empirically with the advent of the WEIRDness cultural distance (Muthukrishna et al.,
2020). While many researchers have speculated about non-WEIRD moral concerns, and some
researchers having erroneously dichotomized the WEIRD spectrum (e.g., Dogruyol et al., 2019),
no study to our knowledge had examined the relationship between WEIRDness and moral
foundations. In the present research, we found that culture-level endorsements of Purity and
Loyalty are strongly higher in non-WEIRD cultures. Therefore, Purity and Loyalty may be
considered least WEIRD of the moral foundations, being substantially more salient in cultures
such as Egypt, Saudi Arabia, and Morocco.
Robust Between-Group Differences
Using MFQ-2 in Study 4.2, we replicated three well-established group differences in
moral foundations: sex differences (see Atari et al., 2020c), religious differences (see Graham &
Haidt, 2010), and ideological differences (see Kivikangas et al., 2021). Our examination of
culturally variable sex differences suggested that women cared more about Equality and Purity
than did men. Men on the other hand scored slightly higher than women on Loyalty, Authority,
and Proportionality. Women’s higher emphasis on Care, Equality, and Purity may be related to
their parental care systems and disgust sensitivity, extensively researched in evolutionary
psychology (Al-Shawaf et al., 2018). These culturally variable sex differences are consistent
with prior work showing that women attribute more importance to understanding, appreciation,

167

tolerance, and protection for the welfare of all people and for nature across cultures (Schwartz &
Rubel, 2005). Relatively small sex differences in Loyalty and Authority (i.e., small in size and
variable across cultures) are consistent with Atari et al. (2020c) and suggest that motivations for
ingroup loyalty and hierarchical social structures are not substantially different between women
and men. This finding is in line with evolutionary anthropological research examining sex
differences in political leadership in small-scale societies indicating that sex differences in
leadership and coordination of ingroup members are not directly a product of differences in
motivation for status and leadership, but an indirect product of sex differences in cooperation
strategies, access to schooling, and sexual division of labor (Von Rueden et al., 2018).
With regard to religious differences, we first compared moral foundations among
Christians, Muslims, Buddhists, and non-religious individuals, finding that Muslims scored
highest in all foundations. Buddhists scored lowest in all foundations with the exception of
Purity in which non-religious individuals scored lowest. After this between-religion analysis, we
moved on to examining the relationship between religiosity and moral foundations while taking
into account that individuals are simultaneously nested within their nations and religious
affiliations. Here, we found that more religious individuals tend to score lower on Care and
Proportionality, while being more likely to score higher on Loyalty, Authority, Purity, and
Equality. These strong associations between religious affiliation, religious practices, and
endorsement of moral foundations are consistent with Graham and Haidt’s (2010) argument that
beliefs, rituals, and other facets of religious practice are best understood as means of creating a
moral community. What is new here, however, is that this preference is best understood as
emotive for an “egalitarian moral community” rather than merit-based cooperative community.

168

With regard to ideological differences, we replicated the principal findings of Graham et
al. (2009) and Kivikangas et al. (2021). In particular, we found that conservatives tend to score
higher on Proportionality, Loyalty, Authority, Purity while scoring lower on Care and Equality.
Our results are consistent with Kivikangas et al. (2021) who found that, with a few exceptions in
their meta-analysis, Care and Fairness are generally negatively, and Loyalty, Authority, and
Purity, generally positively correlate with right-wing political ideology. Indeed, prior MFT
research did not have the Equality-Proportionality distinction. With respect to differential
relationships between political ideology and these two foundations, we find that liberals tend to
value Equality while conservatives tend to prioritize Proportionality. These findings are
consistent with prior work finding that individuals on the right are more likely to endorse
recognizing people on their merits (Arts & Gellissen, 2001). While some scholars have
characterized conservative opposition to inequality as rigidity (e.g., Jost & Thompson, 2000;
Knowles, Lowery, Hogan, & Chow, 2009), conservative policies such as opposition to
affirmative action have been found to be rooted in meritocratic (as opposed to anti-egalitarian)
motivations (Bobocel et al., 1998; Nosworthy et al., 1995). Conservatives may see the prevailing
economic system as fairer which makes them feel fewer negative feelings when faced with social
inequality (Goudarzi et al., 2020). Conservatives have been found to be less willing to provide
aid to the needy, but when those in need of aid are seen to have fulfilled their social
responsibilities, conservatives are indeed willing to provide such aid (Skitka & Tetlock, 1993).
The liberal stereotype of conservatives as “heartless” may therefore be exaggerated (Brooks,
2006; Graham et al., 2012). Rather than being heartless, groups that place greater emphasis on
merit and deservingness tend to have stable productivity orientations that may trump empathic
feelings. Consistently, conservatives care about economic growth more than liberals do

169

(Rasinski, 1987) and have been shown to score higher on preference for proportional distribution
of resources (Skurka et al., 2020). MFQ-2 provides the opportunity for future research to
examine differential effects of Equality and Proportionality on an array of ideology-related
outcomes within the framework of MFT.
Nomological Network of Moral Foundations
Our findings in Study 4.3 provided compelling evidence that MFQ-2 captures more
variance in a variety of outcomes compared with MFQ. This is noteworthy given that MFQ is
already regarded a powerful tool in predicting a wide array of outcomes ranging from political
behavior (Kivikangas et al., 2021) to hate group activities (Hoover et al., 2021). Even when we
completely dropped Proportionality, MFQ-2 still significantly outperformed MFQ, indicating
that MFQ-2’s superior predictive performance is not due to having several more items or a new
subscale. This finding is promising as it opens the door to future theory-driven examination of
morally relevant behaviors and judgments, as well as modeling approaches that use MFT to
minimize out-of-sample prediction error in predicting a behavioral outcome (e.g., Karimi-
Malekabadi et al., 2021). Furthermore, Studies 4.2 and 4.3 collectively provided evidence that
the individualizing-binding distinction made in Graham et al. (2011) is actually culture-
dependent. Accordingly, one may not assume that two-dimensional higher-order structure exists
in all cultural contexts. This is a new insight into MFT which is plausible since most of Graham
et al.’s (20211) data were based on North America and English-speaking countries. Our network
approach adds to another emerging line of work indicating that moral foundations are inter-
connected in different ways depending on cultural context (Atari et al., 2020b; Turner-Zwinkels
et al., 2021).

170

Study 4.3 further expanded the nomological network of moral foundations. All six
foundations were related to theoretically relevant constructs in predictable ways. The only
exception was Left-Wing Authoritarianism which yielded a non-significant correlation with
Authority. Interestingly, LWA was also unrelated to Loyalty (in which conservatives tend to
score higher) and Care (in which liberals tend to score higher); however, LWA was very strongly
associated with Equality. These results suggest that, at least in the framework of MFT, LWA is
better seen as opposing those who doubt and disavowing the equality of all people. As such,
people who score highly on LWA might think that social groups tend to be equal on the sorts of
important characteristics and observed differences in these characteristics ought to be
dismantled. In other words, left-wing authoritarians may be motivated by Equality intuitions
rather than anti-Authority intuitions. This calls for future research examining a wide range of
interesting questions, for example, how LWA might mediate the relationship between these
moral foundations and prejudice, dogmatism, or anti-social behavior by political liberals. These
findings add to the emerging literature on left-wing authoritarianism which remains a “myth”, on
par with the “Loch Ness Monster” (Stone, 1980).
Lessons Learned and Broader Implications
The current research has important implications for future studies examining moral
psychological theories and assessment of morality. One such implication is the utility of division
of Fairness concerns into Equality and Proportionality without assuming that low scores on one
equates high concerns for the other: while we acknowledge that scholars may still disagree about
psychological roots of motives underlying Equality and Proportionality, our empirical results
show that both constructs are associated with important outcomes in theoretically predictable
ways. As such, the present work further sharpens our understanding of justice motives and

171

fairness concerns and expands the moral domain to six moral foundations. Interestingly, we
found small-to-moderate positive correlations between MFQ’s Fairness scores and both Equality
and Proportionality subscales of MFQ-2, indicating that these are tapping into different portions
of variance in fairness concerns. Hence, MFQ’s Fairness score might be considered a mix of
multiple related concepts rather than a unitary construct. An implication of this finding is that
prior work on fairness judgments and justice motives may need to be further revisited with this
new lens we presented in this work. This can open doors to novel questions and can lead to
theoretically important discoveries. For example, prior work using MFQ has found that across
cultures women scored higher on Fairness (Atari et al., 2020c); however, in the present research
we found that women do score higher on Equality, but significantly lower on Proportionality.
Importantly, we stress that our conceptualization of Equality and Proportionality is descriptive in
nature (rather than normative): MFT continues to emphasize a functional definition of morality
as a description of what motivates people to suppress selfishness and promote harmonious group
living, rather than a prescriptive definition of how one ought to behave. Our new reconsideration
of fairness judgments, implemented in the MFQ-2, can aid in our understanding of the dangers of
different flavors of fairness, for example by justifying social inequalities and systemic racial
inequality in the name of merit (e.g., Goudarzi et al., 2020), or by disregarding one’s talent and
effort in the name of equality.
A second implication of the present research is making available a psychometrically
superior measure of MFT across multiple languages. The current effort fills the need for a
theoretically grounded and psychometrically rigorous scale covering a full range of human moral
concerns. We found substantial evidence that the MFQ-2 is reliable and valid. The scale is
internally consistent while maintaining conceptual coverage of diverse manifestations of

172

foundation-related concerns. External validation of the MFQ-2 using widely used measures in
social and personality psychology, as well as theoretically relevant attitudinal constructs,
demonstrated criterion validity of the new measure across cultures. Factor-analytic approaches
confirmed our theoretical parsing of the moral domain into six concerns: The six-dimensional
model fit the data well across different cultures. Based on both theoretical refinements and
empirical findings of the present work, future research can extend other MFT-based
measurement tools. Among others, the Moral Foundations Dictionary (MFD; Graham et al.,
2009), Moral Foundations Dictionary 2.0 (MFD2; Frimer et al., 2019); Moral Foundations
Vignettes (MFV; Clifford et al., 2015); and Moral Foundations Sacredness (MFSS; Graham &
Haidt, 2012); Moral Foundations Twitter Corpus (MFTC; Hoover et al., 2020); MapYourMorals
(MYP; Hoover et al., 2021); and Moral Foundations Tradeoff Task (MFTT; Graham, 2010) can
be updated in accordance with the new findings and refinements reported here, further generating
testable hypotheses about human morality in different contexts which can be measured using
different methodologies. For example, prior research has pitched scenarios describing possible
violations of Care (e.g., assault) or Purity (e.g., consuming taboo substances, incest) against each
other, while varying intent and outcome, to examine moral condemnation or neural activity
(Borg et al., 2006; Moll et al., 2005; Young & Saxe, 2011). It is less clear how the six
foundations we present here are distinguishable behaviorally (see Young & Saxe, 2011), neurally
(see Borg et al., 2006), linguistically (see Kennedy et al., 2021), and emotionally (see Atari et al.,
2020a). As another example, using the county-level map of moral foundations (Hoover et al.,
2021), county-level endorsements of Fairness (positively) and Purity (positively) have shown to
predict regional vaccination rates during the COVID-19 pandemic (Karimi-Malekabadi et al.,

173

2021). It remains unclear how county-level concerns over Equality and Proportionality may
affect collective behaviors.
A third implication of the present research is its application in understanding and
assessing non-WEIRD morality. We achieve this by two means: first by widening our top-down
theoretical lens which better captures non-WEIRD conceptions of morality (see Willard et al.,
2020), particularly fairness; and second, by diversifying our samples using which we developed
MFQ-2 (see Apicella et al., 2020; Henrich, 2020). In addition, using Muthukrishna’s (2020)
newly validated index of WEIRDness cultural distance, we tested novel predictions about
different moral foundations in non-WEIRD cultures, finding that Purity and Loyalty are
particularly higher in more non-WEIRD cultures such as Egypt and Saudi Arabia. Our approach
has important implications for moral psychological research because moral cognition may be
more a kludge, shaped by local social norms and other features of cognition than a unified
cognitive architecture (Stich, 2006), hence it is imperative that our tools are created with this
human diversity in mind, making sure that our tools are understandable and usable across less-
WEIRD populations. In addition to collecting data from many different countries, we also
maximized, as much as possible, religious diversity in our sample. Most research linking
religious beliefs and morality has focused on participants in the Abrahamic religions, and in
Christianity in particular (Norenzayan, 2016; White, Kelly, Shariff, & Norenzayan, 2019). This
focus on Christianity, and even more narrowly, Protestantism, is in fact a common feature of the
psychology of religion, as has been observed by cross-cultural scholars of religion (e.g.,
Saroglou & Cohen, 2013). Tapping into the religious diversity we tested the relationships
between moral foundations and religious identity (i.e., Muslim, Christian, Buddhist, and non-
religious) as well as the strength of individual religiosity.

174

Limitations and Future Directions
The present research had several important methodological strengths, including its large
and demographically diverse samples from both WEIRD and non-WEIRD cultures, its use of
multiple languages instead of using English language to develop a scale and then translate it into
other “secondary” languages, its use of more than one-hundred items as a rich item pool of MFT-
relevant item content, its integration of new conceptual insights and empirical approaches to
scale development, and its broad set of criteria for evaluating the MFQ-2’s basic measurement
properties, multidimensional structure, and nomological network.
However, the present research also had some limitations that suggest important directions
for future work. One such limitation is that we currently do not have cross-culturally valid
measures of other “candidate foundations” which have been proposed as potential moral
foundations using the foundationhood criteria by Graham et al. (2013) but have not gained
consensus among researchers as foundation. Notable candidates are liberty (Iyer et al., 2012),
honor (Atari et al., 2020b), honesty, ownership, and efficiency (see Graham et al., 2013). Our
six-dimensional model is the most parsimonious model that captures the moral domain based on
the current state of the art, and MFQ-2 is shown to be the best existing tool with which these
moral intuitions can be measured. However, addition of foundations — and development of
additional scales to measure those foundations — is a great next step to broaden the pluralistic
view on human morality.
The second limitation of the present work is that we could not test developmental
predictions of MFT since our samples were all adult individuals. As a descriptive theory of
human moral concerns, MFT does not make normative claims about which kinds of concerns are
normatively superior to others or whether people would “grow out” of one foundation with

175

development. However, future work on how the foundations develop, especially from infancy to
adolescence, and change in response to environmental changes and life events could inform our
understanding of how people come to have the moral concerns they do. More work is needed on
the development of moral foundations in children and on the dynamics of moral change
throughout adolescence and adulthood (Cingel & Krcmar, 2020). Indeed, it is possible that
MFQ-2 may need to be revised or reworded to better suit particular age groups.
Third, while we collected data from 25 countries and seven languages, the present results
are still based on a chunk of these populations who were educated enough to complete the
surveys online and on their own. Our sample did not include people from traditional, small-scale
communities, whose means of living are subsistence-based with daily interactions being mainly
with local familiars (e.g., Purzycki et al., 2018). In addition, we did not have enough statistical
power to explore all group intersections within cultures (e.g., religions, ethnicities, gender, etc).
Although, to our knowledge, the present work is among the firsts to revise a commonly used
measure mostly in non-WEIRD populations, the future work is encouraged to further examine
our model in ethnographic work, cross-cultural research, and intersectional studies.
Conclusion
Both lay persons and scholars disagree about the content of the moral domain, or simply
what “morality” means. Inspired by Shweder et al. (1997), MFT was created by integrating
multiple fields of study, proposing that morality goes beyond harm and fairness. Here, we
theoretically refine MFT’s map of the moral domain by proposing six foundations: Care,
Equality, Proportionality, Loyalty, Authority, and Purity. We also present MFQ-2: a reliable,
valid, and easy-to-use self-report tool for exploring this expanded view of the moral domain. We
also provide evidence, based on data from 25 cultures and seven languages, that MFQ-2 scores

176

are internally consistent, cross-culturally valid, and pragmatically valuable in predicting
psychological and cultural phenomena.

177

Appendix
Moral Foundations Questionnaire-2 (MFQ-2)
For each of the statements below, please indicate how well each statement describes you or your
opinions. Response options: Does not describe me at all (1); Slightly describes me (2);
Moderately describes me (3); Describes me fairly well (4); Describes me extremely well (5).
1. Caring for people who have suffered is an important virtue.
2. The world would be a better place if everyone made the same amount of money.
3. I think people who are more hard-working should end up with more money.
4. I think children should be taught to be loyal to their country.
5. I think it is important for societies to cherish their traditional values.
6. I think the human body should be treated like a temple, housing something sacred within.
7. I believe that compassion for those who are suffering is one of the most crucial virtues.
8. Our society would have fewer problems if people had the same income.
9. I think people should be rewarded in proportion to what they contribute.
10. It upsets me when people have no loyalty to their country.
11. I feel that most traditions serve a valuable function in keeping society orderly.
12. I believe chastity is an important virtue.
13. We should all care for people who are in emotional pain.
14. I believe that everyone should be given the same quantity of resources in life.
15. The effort a worker puts into a job ought to be reflected in the size of a raise they receive.
16. Everyone should love their own community.
17. I think obedience to parents is an important virtue.
18. It upsets me when people use foul language like it is nothing.
19. I am empathetic toward those people who have suffered in their lives.
20. I believe it would be ideal if everyone in society wound up with roughly the same amount
of money.
21. It makes me happy when people are recognized on their merits.
22. Everyone should defend their country, if called upon.

178

23. We all need to learn from our elders.
24. If I found out that an acquaintance had an unusual but harmless sexual fetish I would feel
uneasy about them.
25. Everyone should try to comfort people who are going through something hard.
26. When people work together toward a common goal, they should share the rewards
equally, even if some worked harder on it.
27. In a fair society, those who work hard should live with higher standards of living.
28. Everyone should feel proud when a person in their community wins in an international
competition.
29. I believe that one of the most important values to teach children is to have respect for
authority.
30. People should try to use natural medicines rather than chemically identical human-made
ones.
31. It pains me when I see someone ignoring the needs of another human being.
32. I get upset when some people have a lot more money than others in my country.
33. I feel good when I see cheaters get caught and punished.
34. I believe the strength of a sports team comes from the loyalty of its members to each
other.
35. I think having a strong leader is good for society.
36. I admire people who keep their virginity until marriage.
Scoring: Average each of the following items to get six scores corresponding with the six
foundations.
Care = 1, 7, 13, 19, 25, 31
Equality = 2, 8, 14, 20, 26, 32
Proportionality = 3, 9, 15, 21, 27, 33
Loyalty = 4, 10, 16, 22, 28, 34
Authority = 5, 11, 17, 23, 29, 35

179

Purity = 6, 12, 18, 24, 30, 36

180

Supplementary Materials
Study 1
Table S.1.
Cultural Differences and Descriptive Statistics for Care Items (Study 4.1a)
Item
United States India
M SD rMFQ M SD rMFQ t p d
It pains me when I see someone ignoring
the needs of another human being.
3.99 1.07 0.47 4.14 0.98 0.28 -
2.09
0.037 -
0.14
I try very hard not to hurt anyone’s
feelings.
4.03 1.07 0.35 4.26 0.99 0.29 -
3.19
0.001 -
0.22
I am empathetic toward those people who
have suffered in their lives.
4.12 1.00 0.46 4.00 1.03 0.34 1.63 0.103 0.12
It bothers me to see someone get hurt. 4.34 0.92 0.51 4.21 0.95 0.32 1.94 0.052 0.14
It is not my problem that someone else
has suffered in their life.
2.10 1.17 -
0.34
2.55 1.35 -
0.05
-
5.05
0.000 -
0.36
When I see someone get hurt, I feel the
urge to do something about it.
3.87 1.06 0.43 4.10 1.00 0.35 -
3.20
0.001 -
0.22
I believe it is ok to use violence in some
circumstances.
2.58 1.23 -
0.16
2.78 1.35 -
0.02
-
2.13
0.034 -
0.15
I admire people who strive to relieve
human suffering.
4.31 0.94 0.52 4.14 0.98 0.34 2.63 0.009 0.19
I try to be kind toward others when they
are in pain.
4.39 0.82 0.43 4.32 0.85 0.27 1.14 0.253 0.08
I am kind toward others when they are in
need.
4.18 0.89 0.44 4.23 0.92 0.36 -
0.80
0.422 -
0.06
I believe that compassion for those who
are suffering is one of the most crucial
virtues.
4.08 1.09 0.52 3.87 1.12 0.41 2.66 0.008 0.19
Caring for people who have suffered is an
important virtue.
4.21 0.93 0.52 4.22 0.94 0.38 -
0.21
0.835 -
0.01
We should all care for people who are in
emotional pain.
4.02 1.05 0.46 4.17 0.97 0.38 -
2.03
0.042 -
0.14
Everyone should try to comfort people
who are going through something hard.
3.96 1.02 0.44 4.07 0.97 0.31 -
1.51
0.132 -
0.10
I admire people whose occupations
relieve human suffering, for example
nurses.
4.28 0.95 0.48 4.17 1.02 0.31 1.62 0.106 0.12
MFQ-Care 3.70 0.79 1.00 3.56 0.77 1.00 2.62 0.009 0.18

181

Figure S.1.
Factor Loadings of Care Items in Two Countries (Study 4.1a)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

182

Table S.2.
Cultural Differences and Descriptive Statistics for Fairness Items (Study 4.1a)
Item
United States India
M SD rMFQ M SD rMFQ t p d
I feel good when I see cheaters get caught and
punished.
3.62 1.22 0.02 4.18 1.04 0.25 -
7.22
0.000 -
0.49
I feel a moral obligation to help people who
have helped me before.
4.19 0.91 0.37 4.29 0.93 0.27 -
1.58
0.114 -
0.11
I think people should be rewarded in proportion
to what they contribute.
3.69 1.12 0.02 4.03 1.02 0.13 -
4.60
0.000 -
0.32
I get mad when in a project, lazy members of
the group are rewarded equally to hard-working
ones.
3.82 1.16 0.09 3.83 1.15 0.19 -
0.11
0.916 -
0.01
It upsets me when I see someone not doing
their fair share of a collaborative project.
3.99 1.01 0.17 3.91 1.06 0.30 1.16 0.248 0.08
I believe it would be ideal if everyone in
society wound up with roughly the same
amount of money.
2.57 1.41 0.33 3.25 1.25 0.08 -
7.41
0.000 -
0.51
I believe everyone in a fair society should end
up with roughly the same amount of money
regardle…
2.39 1.32 0.33 3.27 1.23 0.07 -
9.84
0.000 -
0.68
When dividing up a bonus, I think the people
who contributed the most to success should get
the m…
3.73 1.11 -
0.01
3.96 1.13 0.14 -
2.94
0.003 -
0.21
In a fair society, I want people who work
harder than others to end up richer than others.
3.33 1.18 -
0.08
3.84 1.11 0.17 -
6.40
0.000 -
0.44
When people work together toward a common
goal, they should share the rewards equally,
even if so…
2.55 1.25 0.12 3.30 1.29 0.04 -
8.41
0.000 -
0.59
When dividing up a bonus, I think fairness
means equality: people should all get the same
amount…
2.11 1.22 0.12 3.02 1.39 -
0.11
-
9.91
0.000 -
0.71
MFQ-Fairness 3.59 0.76 1.00 3.56 0.74 1.00 0.73 0.468 0.05

183

Figure S.2.
Factor Loadings of Fairness Items in Two Countries (Study 4.1a)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

184

Table S.3.
Cultural Differences and Descriptive Statistics for Equality Items (Study 4.1a)
Item
United States India
M SD rMFQ M SD rMFQ t p d
I engage in activities that promote social
equality.
3.16 1.33 0.39 3.82 1.13 0.15 -7.78 0.000 -
0.53
I believe that everyone should be given the
same quantity of resources in life.
3.16 1.39 0.37 3.74 1.17 0.14 -6.55 0.000 -
0.45
I think a group prize should be divided among
the group members in the same amounts.
3.38 1.28 0.23 3.91 1.10 0.21 -6.50 0.000 -
0.45
The world would be a better place if everyone
made the same amount of money.
2.43 1.37 0.34 3.40 1.33 0.14 -
10.25
0.000 -
0.72
I feel good when I see children share their toys
equally.
4.11 0.99 0.30 4.37 0.87 0.20 -4.09 0.000 -
0.28
I get upset when some people have a lot more
money than others in my country.
2.78 1.42 0.36 2.99 1.40 0.12 -2.07 0.039 -
0.15
It upsets me when someone gives preferential
treatment to one of their children.
3.58 1.27 0.26 3.78 1.17 0.17 -2.36 0.018 -
0.16
I get upset when I see inequalities in income
among citizens.
3.19 1.36 0.47 3.62 1.23 0.17 -4.78 0.000 -
0.33
If I were to divide a reward between children, I
would try to divide rewards completely
equally.
4.14 1.06 0.35 4.29 1.00 0.24 -2.20 0.028 -
0.15
Our society would have fewer problems if
people had the same income.
2.53 1.40 0.34 3.29 1.31 0.11 -8.09 0.000 -
0.56
I think those who are well-off have a duty to
help those who are less fortunate.
3.67 1.22 0.41 3.89 1.03 0.34 -2.87 0.004 -
0.20
In a fair society, basic services such as health
care should be provided for everyone free of
cha…
3.90 1.36 0.46 4.19 1.01 0.25 -3.49 0.001 -
0.23
When serving food to several adults, I take
extra time to ensure every plate has perfectly
equal…
3.12 1.32 0.23 3.79 1.22 0.13 -7.62 0.000 -
0.53
I believe everyone should have equal
opportunities in life.
4.30 0.96 0.42 4.26 0.96 0.23 0.58 0.565 0.04
MFQ-Fairness 3.59 0.76 1.00 3.56 0.74 1.00 0.73 0.468 0.05

185

Figure S.3.
Factor Loadings of Equality Items in Two Countries (Study 4.1a)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

186

Table S.4.
Cultural Differences and Descriptive Statistics for Proportionality Items (Study 4.1a)

Item
United States India
M SD rMFQ M SD rMFQ t p d
I try to make sure everyone gets what they
deserve.
3.33 1.21 0.25 4.06 0.97 0.24 -
9.63
0.000 -
0.65
I believe that everyone should be given
resources based on their needs.
3.63 1.21 0.46 3.99 1.03 0.20 -
4.60
0.000 -
0.31
I think officials should allocate resources to
different areas according to which areas have
the…
3.86 1.10 0.40 4.07 0.95 0.26 -
2.92
0.004 -
0.20
I believe that national resources should be
divided to different areas according to their
need.
3.63 1.16 0.41 3.99 0.97 0.27 -
4.89
0.000 -
0.33
I think people who are more hard-working
should end up with more money.
3.77 1.10 0.00 4.08 1.04 0.12 -
4.17
0.000 -
0.29
The government should aid unfortunate
families more than well-off ones.
3.88 1.21 0.45 3.96 1.10 0.32 -
1.04
0.298 -
0.07
I get upset when somebody obtains something
without effort.
3.01 1.30 0.04 3.48 1.28 0.01 -
5.20
0.000 -
0.36
I feel that it is unfair to give some people more
than others just so they end up equal.
2.81 1.45 -
0.28
3.34 1.22 0.11 -
5.75
0.000 -
0.39
It makes me happy when people are recognized
on their merits.
4.35 0.85 0.18 4.26 0.95 0.24 1.44 0.151 0.10
I believe a hard-working person deserves to
succeed in life.
4.42 0.83 0.26 4.40 0.89 0.24 0.20 0.839 0.01
I feel uneasy when people get a large reward
without trying hard enough.
3.22 1.22 0.06 3.58 1.24 0.15 -
4.13
0.000 -
0.29
In a fair society, those who work hard should
live with higher standards of living.
3.34 1.19 -
0.05
3.81 1.09 0.13 -
5.87
0.000 -
0.41
I think that children who help their parents
more, are deserving of more inheritance.
3.02 1.32 0.05 3.73 1.19 0.08 -
8.11
0.000 -
0.56
I think highly skilled individuals deserve to
make more money than less skilled people.
3.59 1.19 -
0.12
3.80 1.15 0.09 -
2.64
0.008 -
0.18
I feel good when people who do their job well
rise to the top.
4.15 0.98 0.04 4.41 0.81 0.27 -
4.20
0.000 -
0.28
I believe people ought to get what they deserve. 3.53 1.15 0.07 3.99 0.98 0.22 -
6.18
0.000 -
0.42
The effort a worker puts into a job ought to be
reflected in the size of a raise they receive.
3.99 1.01 0.18 3.93 1.02 0.22 0.76 0.445 0.05
I believe that the world would be a better place
if we let lazy people suffer the consequences.
2.50 1.31 -
0.26
3.12 1.33 0.06 -
6.67
0.000 -
0.47
MFQ-Fairness 3.59 0.76 1.00 3.56 0.74 1.00 0.73 0.468 0.05

187

Figure S.4.
Factor Loadings of Proportionality Items in Two Countries (Study 4.1a)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

188

Table S.5.
Cultural Differences and Descriptive Statistics for Loyalty Items (Study 4.1a)
Item
United States India
M SD rMFQ M SD rMFQ t p d
It is more important to be a team player than to
express oneself.
2.73 1.22 0.51 3.85 1.11 0.36 -
13.74
0 -
0.95
People should be loyal to their close friends, even
when they have done something wrong.
2.66 1.21 0.41 3.29 1.30 0.22 -7.09 0 -
0.50
I believe that one of the most important values to
teach children is to be loyal to their families.
3.08 1.36 0.62 4.24 0.95 0.41 -
14.61
0 -
0.96
I think it is important that people remain loyal to
their families.
3.24 1.31 0.63 4.30 0.91 0.41 -
13.79
0 -
0.91
I think children should be taught to be loyal to
their country.
2.75 1.45 0.68 4.23 1.08 0.37 -
17.05
0 -
1.14
I believe the strength of a family comes from the
loyalty of its members to each other.
3.37 1.28 0.57 4.27 0.91 0.35 -
11.93
0 -
0.79
In a dispute I tend to take my friend’s side even
before I learn exactly what happened.
2.45 1.12 0.19 2.91 1.32 0.10 -5.28 0 -
0.38
I feel angry when someone insults my country. 2.48 1.45 0.64 4.19 1.10 0.33 -
19.46
0 -
1.30
It bothers me when someone criticizes my
country.
2.48 1.43 0.63 4.11 1.14 0.38 -
18.22
0 -
1.23
People should remain loyal to their family even
when some family members are doing something
wrong.
2.28 1.21 0.56 3.31 1.29 0.25 -
11.69
0 -
0.83
People who betray their group should get kicked
out of the group.
2.98 1.25 0.35 3.84 1.24 0.27 -9.86 0 -
0.69
Everyone should love their own country. 2.84 1.47 0.65 4.30 1.05 0.34 -
16.70
0 -
1.11
Everyone should defend their country, if called
upon.
2.62 1.47 0.61 4.09 1.07 0.42 -
16.66
0 -
1.11
Everyone should feel proud when a person in their
country wins in an international competition.
3.28 1.38 0.52 4.30 1.03 0.34 -
12.29
0 -
0.82
I admire people who stick by their group even if it
would serve them better to leave.
2.39 1.28 0.44 3.62 1.08 0.35 -
15.09
0 -
1.03
It upsets me when people have no loyalty to their
country.
2.57 1.45 0.64 3.93 1.20 0.39 -
14.78
0 -
1.00
I wish the world did not have nations or borders
and we were all part of one big group.
2.61 1.49 -
0.15
3.77 1.26 0.15 -
12.12
0 -
0.83
I identify more closely with the people of the
world at large than with the people in my own
coun…
2.59 1.36 -
0.14
3.24 1.30 -
0.01
-6.97 0 -
0.48
It bothers me when someone quits a company
that’s been good to them for years, to go work for
a c…
2.08 1.28 0.42 3.25 1.31 0.27 -
12.88
0 -
0.91
MFQ-Loyalty 2.38 1.05 1.00 3.43 0.75 1.00 -
16.97
0 -
1.12

189

Figure S.5.
Factor Loadings of Loyalty Items in Two Countries (Study 4.1a)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

190

Table S.6.
Cultural Differences and Descriptive Statistics for Authority Items (Study 4.1a)
Item
United States India
M SD rMFQ M SD rMFQ t p d
I believe social order should be prioritized to
keep a society safe.
2.87 1.29 0.52 3.93 1.09 0.31 -
12.76
0.000 -
0.87
I generally like people who don’t feel much
respect for authority.
1.97 1.19 -
0.41
2.53 1.41 -
0.01
-6.06 0.000 -
0.44
Children should never disrespect their parents. 3.19 1.45 0.61 4.20 1.10 0.31 -
11.51
0.000 -
0.77
I think everyone should trust the judgment of
the proper authorities.
2.66 1.22 0.57 3.78 1.12 0.43 -
13.77
0.000 -
0.95
In general, I think the best way to do things is
the traditional way.
2.36 1.21 0.59 3.54 1.20 0.32 -
13.98
0.000 -
0.98
I feel that most traditions serve a valuable
function in keeping society orderly.
2.83 1.25 0.62 3.82 1.07 0.27 -
12.30
0.000 -
0.84
I believe that employees should do what their
bosses tell them to do (as long as it is legal)
eve…
2.95 1.19 0.46 3.38 1.20 0.36 -5.03 0.000 -
0.35
I think having a strong, determined leader is
good for society.
3.80 1.10 0.40 4.36 0.92 0.24 -8.01 0.000 -
0.55
I like it when rebels in society face the
consequences of their actions.
2.58 1.32 0.54 3.42 1.24 0.29 -9.44 0.000 -
0.66
It makes me angry when students disrespect
their teachers.
3.70 1.22 0.46 4.10 1.10 0.27 -4.92 0.000 -
0.34
I think it is important for societies to cherish
their traditional values.
2.99 1.29 0.65 4.01 1.02 0.38 -
12.70
0.000 -
0.86
I think it is sometimes justified to rebel
against authorities.
3.48 1.29 -
0.50
3.42 1.20 0.03 0.72 0.474 0.05
I think it can be ok to insult your parents if
they insult you first.
2.25 1.36 -
0.40
2.03 1.40 -
0.06
2.29 0.023 0.16
I believe that one of the most important values
to teach children is to have respect for
authority.
3.07 1.37 0.72 3.98 1.03 0.33 -
10.97
0.000 -
0.73
I fear that society would tumble into chaos if
everyone started doing as they pleased.
3.57 1.35 0.57 3.75 1.15 0.24 -2.12 0.034 -
0.14
Authorities should care for those beneath
them at all costs.
3.89 1.10 -
0.19
4.03 0.99 0.26 -1.85 0.064 -
0.13
I believe it is important for us to honor our
ancestors.
3.27 1.23 0.50 4.13 1.02 0.35 -
10.95
0.000 -
0.74
I think obedience to parents is an important
virtue.
3.23 1.36 0.69 4.29 0.98 0.35 -
13.12
0.000 -
0.87
It angers me when authorities fail to resolve
disputes.
3.58 1.11 -
0.01
4.01 1.01 0.22 -5.80 0.000 -
0.40
We all need to learn from our elders. 3.57 1.17 0.51 4.12 1.00 0.34 -7.18 0.000 -
0.49

191

Figure S.6.
Factor Loadings of Authority Items in Two Countries (Study 4.1a)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

192

Table S.7.
Cultural Differences and Descriptive Statistics for Purity Items (Study 4.1a)
Item
United States India
M SD rMFQ M SD rMFQ t p d
It bothers me when people do something
disgusting, even if no one is harmed.
3.01 1.25 0.55 3.73 1.14 0.32 -8.63 0.000 -0.60
I believe chastity is an important virtue. 2.29 1.42 0.68 3.71 1.20 0.46 -
15.61
0.000 -1.06
I think the human body should be treated like
a temple, housing something sacred within.
2.96 1.42 0.56 3.76 1.19 0.31 -8.82 0.000 -0.60
I look down on people who don’t treat their
body with the respect it deserves.
2.25 1.25 0.32 3.47 1.23 0.30 -
14.00
0.000 -0.98
It bothers me when people think nothing is
sacred in this world.
3.16 1.41 0.48 3.51 1.32 0.40 -3.72 0.000 -0.26
I admire people who keep their virginity until
marriage.
2.40 1.53 0.64 3.78 1.40 0.31 -
13.57
0.000 -0.94
I would call some acts wrong on the grounds
that they are unnatural.
2.46 1.30 0.65 3.44 1.20 0.41 -
11.22
0.000 -0.78
I think that keeping one’s impulses in check is
an important virtue.
3.75 1.10 0.32 3.95 0.98 0.34 -2.75 0.006 -0.19
People should try to use natural medicines
rather than chemically identical human-made
ones.
2.54 1.40 0.40 3.97 1.10 0.31 -
16.58
0.000 -1.12
I believe that drinking alcohol pollutes your
soul.
1.76 1.25 0.44 3.37 1.52 0.40 -
16.21
0.000 -1.18
Promiscuity is one of the worst qualities a
human can have.
1.99 1.29 0.58 3.43 1.28 0.33 -
16.04
0.000 -1.12
Nature is sacred and should not be desecrated. 3.79 1.17 0.07 4.03 1.15 0.34 -2.95 0.003 -0.21
If I found out that an acquaintance had an
unusual but harmless sexual fetish I would
feel uneasy…
2.01 1.22 0.51 3.18 1.35 0.37 -
12.92
0.000 -0.92
Consuming foods with many artificial
ingredients dirties the body even if they are
not physically…
2.36 1.33 0.38 3.48 1.33 0.32 -
12.01
0.000 -0.84
If somebody touched a corpse at a funeral, out
of curiosity rather than love, I would say that
is…
2.55 1.47 0.29 2.95 1.42 0.20 -3.96 0.000 -0.28
I believe there is nothing sacred about human
body.
1.80 1.20 -
0.29
2.75 1.50 0.06 -9.86 0.000 -0.72
It upsets me when people use foul language
like it is nothing.
2.22 1.37 0.57 3.78 1.27 0.42 -
16.93
0.000 -1.17
I think having sex with many people is
disgusting.
2.33 1.51 0.62 3.60 1.52 0.33 -
11.97
0.000 -0.84
I think that sexual promiscuity (“sleeping
around”) is disgusting.
2.33 1.47 0.62 3.34 1.42 0.35 -9.93 0.000 -0.69
MFQ1-Authority 2.23 1.32 1.00 3.27 0.85 1.00 -
13.90
0.000 -0.90

193

Figure S.7.
Factor Loadings of Purity Items in Two Countries (Study 4.1a)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

194

Study 1b
Table S.8.
Cultural Differences and Descriptive Statistics for Care Items (Study 4.1b)
Item
United States India
M SD rMFQ M SD rMFQ t p d
It pains me when I see someone ignoring the
needs of another human being.
3.91 1.04 0.47 3.82 1.03 0.26 1.23 0.219 0.08
I try very hard not to hurt anyone’s feelings. 3.90 1.07 0.27 4.07 1.03 0.38 -
2.37
0.018 -
0.16
I am empathetic toward those people who have
suffered in their lives.
4.12 0.92 0.45 3.80 1.01 0.39 4.81 0.000 0.33
It bothers me to see someone get hurt. 4.24 0.89 0.38 3.96 1.02 0.34 4.35 0.000 0.30
When I see someone get hurt, I feel the urge to
do something about it.
3.79 1.00 0.40 3.85 0.99 0.41 -
0.91
0.363 -
0.06
I admire people who strive to relieve human
suffering.
4.28 0.89 0.38 4.00 1.01 0.43 4.21 0.000 0.29
I am kind toward others when they are in need. 4.07 0.86 0.32 4.07 0.94 0.37 -
0.07
0.943 0.00
I believe that compassion for those who are
suffering is one of the most crucial virtues.
4.13 0.93 0.53 3.68 1.01 0.44 6.68 0.000 0.46
Caring for people who have suffered is an
important virtue.
4.11 0.93 0.47 4.03 0.95 0.47 1.18 0.237 0.08
We should all care for people who are in
emotional pain.
3.99 1.00 0.49 4.05 0.92 0.38 -
0.93
0.352 -
0.06
Everyone should try to comfort people who are
going through something hard.
3.92 0.98 0.40 3.82 0.95 0.39 1.56 0.120 0.11
I admire people whose occupations relieve
human suffering, for example nurses.
4.26 0.91 0.32 4.06 1.00 0.43 3.05 0.002 0.21
MFQ-Care 3.77 0.77 1.00 3.50 0.85 1.00 4.82 0.000 0.33

195

Figure S.8.
Factor Loadings of Care Items in Two Countries (Study 4.1b)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

196

Table S.9.
Cultural Differences and Descriptive Statistics for Fairness Items (Study 4.1b)
Item
United States India
M SD rMFQ M SD rMFQ t p d
I feel good when I see cheaters get caught and
punished.
3.59 1.16 0.00 4.03 1.08 0.21 -
5.88
0.000 -
0.40
I think people should be rewarded in proportion
to what they contribute.
3.54 1.06 -
0.06
3.87 0.98 0.29 -
4.66
0.000 -
0.32
I get mad when in a project, lazy members of
the group are rewarded equally to hard-working
ones.
3.58 1.19 -
0.04
3.59 1.16 0.14 -
0.05
0.957 0.00
I believe it would be ideal if everyone in
society wound up with roughly the same
amount of money.
2.50 1.28 0.31 3.20 1.13 0.04 -
8.53
0.000 -
0.57
When dividing up a bonus, I think the people
who contributed the most to success should get
the m…
3.50 1.13 -
0.08
3.93 0.94 0.23 -
6.10
0.000 -
0.41
In a fair society, I want people who work
harder than others to end up richer than others.
3.27 1.14 -
0.11
3.70 1.04 0.23 -
5.76
0.000 -
0.39
When people work together toward a common
goal, they should share the rewards equally,
even if so…
2.63 1.12 0.16 3.33 1.16 0.10 -
8.98
0.000 -
0.62
MFQ-Fairness 3.63 0.75 1.00 3.45 0.81 1.00 3.37 0.001 0.23

197

Figure S.9.
Factor Loadings of Fairness Items in Two Countries (Study 4.1b)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

198

Table S.10.
Cultural Differences and Descriptive Statistics for Equality Items (Study 4.1b)
Item
United States India
M SD rMFQ M SD rMFQ t p d
I engage in activities that promote social
equality.
3.09 1.30 0.37 3.75 1.06 0.17 -8.28 0.000 -
0.55
I believe that everyone should be given the
same quantity of resources in life.
3.23 1.28 0.37 3.69 1.11 0.24 -5.68 0.000 -
0.38
I think a group prize should be divided among
the group members in the same amounts.
3.34 1.19 0.23 3.83 1.13 0.23 -6.15 0.000 -
0.42
The world would be a better place if everyone
made the same amount of money.
2.43 1.29 0.28 3.45 1.27 0.09 -
11.58
0.000 -
0.79
I get upset when some people have a lot more
money than others in my country.
2.63 1.35 0.40 3.03 1.21 0.13 -4.56 0.000 -
0.31
I get upset when I see inequalities in income
among citizens.
3.23 1.30 0.47 3.51 1.10 0.16 -3.42 0.001 -
0.23
If I were to divide a reward between children, I
would try to divide rewards completely
equally.
4.06 1.06 0.24 4.08 1.08 0.33 -0.31 0.758 -
0.02
Our society would have fewer problems if
people had the same income.
2.54 1.31 0.29 3.23 1.19 0.12 -8.19 0.000 -
0.55
I think those who are well-off have a duty to
help those who are less fortunate.
3.58 1.21 0.42 3.67 1.01 0.34 -1.15 0.252 -
0.08
In a fair society, basic services such as health
care should be provided for everyone free of
cha…
3.84 1.37 0.37 4.12 0.96 0.31 -3.55 0.000 -
0.23
When serving food to several adults, I take
extra time to ensure every plate has perfectly
equal…
3.08 1.27 0.26 3.63 1.10 0.21 -6.88 0.000 -
0.46
I believe everyone should have equal
opportunities in life.
4.19 0.99 0.40 4.11 0.99 0.32 1.17 0.242 0.08
MFQ-Fairness 3.63 0.75 1.00 3.45 0.81 1.00 3.37 0.001 0.23

199

Figure S.10.
Factor Loadings of Equality Items in Two Countries (Study 4.1b)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

200

Table S.11.
Cultural Differences and Descriptive Statistics for Proportionality Items (Study 4.1b)
Item
United States India
M SD rMFQ M SD rMFQ t p d
I think people who are more hard-working
should end up with more money.
3.61 1.10 -
0.11
3.87 1.02 0.22 -3.50 0.000 -
0.24
I get upset when somebody obtains something
without effort.
2.90 1.26 0.13 3.29 1.19 0.16 -4.78 0.000 -
0.32
I feel that it is unfair to give some people
more than others just so they end up equal.
2.71 1.35 -
0.30
3.29 1.10 0.12 -6.99 0.000 -
0.47
It makes me happy when people are
recognized on their merits.
4.17 0.88 0.15 4.06 0.97 0.25 1.78 0.075 0.12
I believe a hard-working person deserves to
succeed in life.
4.31 0.79 0.21 4.20 0.99 0.36 1.84 0.066 0.13
I feel uneasy when people get a large reward
without trying hard enough.
3.09 1.24 0.08 3.53 1.13 0.18 -5.39 0.000 -
0.36
In a fair society, those who work hard should
live with higher standards of living.
3.35 1.08 -
0.08
3.81 0.99 0.27 -6.49 0.000 -
0.44
I think that children who help their parents
more, are deserving of more inheritance.
3.02 1.22 0.04 3.77 1.11 0.19 -9.50 0.000 -
0.64
I think highly skilled individuals deserve to
make more money than less skilled people.
3.43 1.16 -
0.17
3.67 1.12 0.20 -3.17 0.002 -
0.22
I feel good when people who do their job well
rise to the top.
4.16 0.90 0.07 4.21 0.90 0.31 -0.80 0.425 -
0.05
I believe people ought to get what they
deserve.
3.38 1.11 0.04 3.83 1.00 0.17 -6.32 0.000 -
0.43
The effort a worker puts into a job ought to be
reflected in the size of a raise they receive.
3.90 0.96 0.08 3.77 0.97 0.29 1.97 0.049 0.13
I believe that the world would be a better
place if we let lazy people suffer the
consequences.
2.29 1.23 -
0.23
3.13 1.19 0.17 -
10.09
0.000 -
0.69
MFQ-Fairness 3.63 0.75 1.00 3.45 0.81 1.00 3.37 0.001 0.23

201

Figure S.11.
Factor Loadings of Proportionality Items in Two Countries (Study 4.1b)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

202

Table S.12.
Cultural Differences and Descriptive Statistics for Loyalty Items (Study 4.1b)
Item
United States India
M SD rMFQ M SD rMFQ t p d
It is more important to be a team player than to
express oneself.
2.62 1.16 0.50 3.69 1.08 0.28 -
13.99
0 -
0.95
People should be loyal to their close friends, even
when they have done something wrong.
2.60 1.13 0.37 3.44 1.21 0.34 -
10.43
0 -
0.72
I believe that one of the most important values to
teach children is to be loyal to their families.
3.09 1.20 0.61 4.13 0.96 0.38 -
14.25
0 -
0.95
I think it is important that people remain loyal to
their families.
3.21 1.22 0.62 4.11 0.96 0.40 -
12.20
0 -
0.81
I think children should be taught to be loyal to their
country.
2.79 1.38 0.68 4.18 1.00 0.42 -
17.25
0 -
1.13
I believe the strength of a family comes from the
loyalty of its members to each other.
3.33 1.19 0.49 4.04 1.00 0.40 -9.44 0 -
0.63
I feel angry when someone insults my country. 2.48 1.42 0.65 4.12 1.09 0.31 -
19.20
0 -
1.27
It bothers me when someone criticizes my country. 2.47 1.38 0.63 3.94 1.15 0.40 -
17.11
0 -
1.14
People should remain loyal to their family even
when some family members are doing something
wrong.
2.32 1.17 0.57 3.37 1.21 0.37 -
12.92
0 -
0.89
People who betray their group should get kicked
out of the group.
2.88 1.19 0.38 3.86 1.09 0.20 -
12.62
0 -
0.85
Everyone should love their own country. 2.85 1.35 0.68 4.30 1.01 0.35 -
18.14
0 -
1.20
Everyone should defend their country, if called
upon.
2.74 1.37 0.61 3.97 1.07 0.38 -
14.84
0 -
0.98
Everyone should feel proud when a person in their
country wins in an international competition.
3.13 1.28 0.51 4.14 1.06 0.41 -
12.84
0 -
0.86
I admire people who stick by their group even if it
would serve them better to leave.
2.47 1.16 0.46 3.41 1.04 0.32 -
12.58
0 -
0.85
It upsets me when people have no loyalty to their
country.
2.65 1.40 0.68 3.89 1.17 0.33 -
14.23
0 -
0.95
It bothers me when someone quits a company
that’s been good to them for years, to go work for a
c…
1.97 1.16 0.42 3.28 1.13 0.19 -
16.79
0 -
1.14
MFQ-Loyalty 2.37 1.05 1.00 3.46 0.78 1.00 -
17.51
0 -
1.15

203

Figure S.12.
Factor Loadings of Loyalty Items in Two Countries (Study 4.1b)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

204

Table S.13.
Cultural Differences and Descriptive Statistics for Authority Items (Study 4.1b)
Item
United States India
M SD rMFQ M SD rMFQ t p d
I believe social order should be prioritized to
keep a society safe.
2.78 1.22 0.55 3.86 1.06 0.27 -
13.86
0.000 -
0.93
Children should never disrespect their parents. 3.22 1.31 0.65 4.09 1.12 0.32 -
10.51
0.000 -
0.70
I think everyone should trust the judgment of
the proper authorities.
2.53 1.11 0.52 3.64 1.05 0.38 -
15.19
0.000 -
1.03
In general, I think the best way to do things is
the traditional way.
2.33 1.16 0.62 3.47 1.18 0.23 -
14.20
0.000 -
0.97
I feel that most traditions serve a valuable
function in keeping society orderly.
2.82 1.19 0.64 3.76 1.02 0.33 -
12.48
0.000 -
0.84
I believe that employees should do what their
bosses tell them to do (as long as it is legal)
eve…
2.89 1.15 0.51 3.29 1.13 0.29 -5.25 0.000 -
0.36
I think having a strong, determined leader is
good for society.
3.71 1.12 0.47 4.21 0.99 0.33 -6.97 0.000 -
0.47
I like it when rebels in society face the
consequences of their actions.
2.55 1.26 0.57 3.33 1.15 0.24 -9.52 0.000 -
0.64
It makes me angry when students disrespect
their teachers.
3.61 1.18 0.46 3.95 1.08 0.27 -4.41 0.000 -
0.30
I think it is important for societies to cherish
their traditional values.
2.92 1.27 0.65 3.92 1.03 0.42 -
12.89
0.000 -
0.86
I believe that one of the most important values
to teach children is to have respect for
authority.
3.13 1.27 0.74 3.87 1.08 0.42 -9.29 0.000 -
0.62
I fear that society would tumble into chaos if
everyone started doing as they pleased.
3.50 1.33 0.50 3.62 1.09 0.30 -1.55 0.122 -
0.10
I believe it is important for us to honor our
ancestors.
3.30 1.20 0.52 3.98 1.05 0.37 -8.92 0.000 -
0.60
I think obedience to parents is an important
virtue.
3.26 1.29 0.70 4.19 0.94 0.37 -
12.31
0.000 -
0.81
We all need to learn from our elders. 3.56 1.16 0.52 4.03 1.00 0.35 -6.46 0.000 -
0.43
MFQ-Authority 2.69 1.03 1.00 3.41 0.78 1.00 -
11.76
0.000 -
0.78

205

Figure S.13.
Factor Loadings of Authority Items in Two Countries (Study 4.1b)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

206

Table S.14.
Cultural Differences and Descriptive Statistics for Purity Items (Study 4.1b)
Item
United States India
M SD rMFQ M SD rMFQ t p d
It bothers me when people do something
disgusting, even if no one is harmed.
3.01 1.23 0.52 3.56 1.12 0.31 -6.90 0.000 -
0.47
I believe chastity is an important virtue. 2.31 1.38 0.65 3.65 1.16 0.41 -
15.55
0.000 -
1.04
I think the human body should be treated like a
temple, housing something sacred within.
2.92 1.25 0.49 3.71 1.11 0.30 -9.83 0.000 -
0.66
I look down on people who don’t treat their
body with the respect it deserves.
2.04 1.11 0.28 3.34 1.17 0.30 -
16.58
0.000 -
1.14
It bothers me when people think nothing is
sacred in this world.
3.17 1.32 0.46 3.33 1.17 0.36 -1.94 0.052 -
0.13
I admire people who keep their virginity until
marriage.
2.44 1.51 0.62 3.77 1.31 0.30 -
13.88
0.000 -
0.93
I would call some acts wrong on the grounds
that they are unnatural.
2.45 1.25 0.64 3.28 1.13 0.38 -
10.29
0.000 -
0.69
I think that keeping one’s impulses in check is
an important virtue.
3.59 1.02 0.36 3.56 0.97 0.26 0.54 0.587 0.04
People should try to use natural medicines
rather than chemically identical human-made
ones.
2.46 1.31 0.30 3.76 1.17 0.26 -
15.42
0.000 -
1.04
I believe that drinking alcohol pollutes your
soul.
1.71 1.18 0.35 3.39 1.41 0.33 -
18.72
0.000 -
1.31
Promiscuity is one of the worst qualities a
human can have.
1.96 1.20 0.53 3.32 1.14 0.34 -
16.96
0.000 -
1.15
If I found out that an acquaintance had an
unusual but harmless sexual fetish I would feel
uneasy…
1.90 1.16 0.49 3.09 1.19 0.28 -
14.78
0.000 -
1.01
Consuming foods with many artificial
ingredients dirties the body even if they are
not physically…
2.38 1.29 0.27 3.44 1.23 0.30 -
12.40
0.000 -
0.84
It upsets me when people use foul language
like it is nothing.
2.20 1.38 0.52 3.58 1.14 0.31 -
16.22
0.000 -
1.08
I think that sexual promiscuity (“sleeping
around”) is disgusting.
2.26 1.36 0.58 3.19 1.33 0.31 -
10.17
0.000 -
0.69
MFQ-Purity 2.22 1.23 1.00 3.27 0.83 1.00 -
14.85
0.000 -
0.97

207

Figure S.14.
Factor Loadings of Purity Items in Two Countries (Study 4.1b)

Dashed lines represent 0.4 (and -0.4) factor loadings. The bold dashed line represents a loading
of 0. The Outermost dashed line (in blue) represents perfect loading (𝜆 = 1.00).

208

References
Abdel-Khalek, A. M. (2006). Measuring happiness with a single-item scale. Social Behavior and
Personality: An International Journal, 34, 139–150.
Abdel-Khalek, A. M. (2007). Assessment of intrinsic religiosity with a single-item measure in a
sample of Arab Muslims. Journal of Muslim Mental Health, 2, 211–215.
Adams, J. S. (1963). Toward an understanding of equity. Journal of Abnormal and Social
Psychology, 67, 422–436.
Adams, J. S. (1965). Inequity in social exchange. In L. Berkowitz (Ed.), Advances in
experimental social psychology (Vol. 2, pp. 267–299). New York: Academic Press.
Afhami, R., Mohammadi-Zarghan, S., & Atari, M. (2017). Self-rating of religiosity (SRR) in
Iran: Validity, reliability, and associations with the Big Five. Mental Health, Religion &
Culture, 20, 879–887.
Aghababaei, N., Mohammadtabar, S., & Saffarinia, M. (2014). Dirty Dozen vs. the H factor:
Comparison of the Dark Triad and Honesty–Humility in prosociality, religiosity, and
happiness. Personality and Individual Differences, 67, 6-10.
Al-Shawaf, L., Lewis, D. M., & Buss, D. M. (2018). Sex differences in disgust: Why are women
more easily disgusted than men?. Emotion review, 10(2), 149-160.
Alexander, R. M. (2019). The Spanish flu and the sanitary dictatorship: Mexico’s response to the
1918 influenza pandemic. The Americas, 76, 443–465.
Almås, I., Cappelen, A. W., Sørensen, E. Ø., & Tungodden, B. (2010). Fairness and the
development of inequality acceptance. Science, 328, 1176–1178.

209

Altemeyer, B. (2006). The Authoritarians. Unpublished manuscript, Department of Psychology,
University of Manitoba, Winnipeg, Canada.
Amin, A. B., Bednarczyk, R. A., Ray, C. E., Melchiori, K. J., Graham, J., Huntsinger, J. R., &
Omer, S. B. (2017). Association of moral values with vaccine hesitancy. Nature Human
Behaviour, 1(12), 873-880.
Anderson, B. W. (1998). Understanding the Old Testament. Upper Saddle River, NJ: Prentice
Hall.
Apicella, C. L., & Barrett, H. C. (2016). Cross-cultural evolutionary psychology. Current
Opinion in Psychology, 7, 92–97.
Apicella, C., Norenzayan, A., & Henrich, J. (2020). Beyond WEIRD: A review of the last
decade and a look ahead to the global laboratory of the future. Evolution and Human
Behavior, 41(5), 319-329.
Arts, W., & Gelissen, J. (2001). Welfare states, solidarity and justice principles: does the type
really matter?. Acta Sociologica, 44(4), 283-299.
Aslani, S., Ramirez-Marin, J., Brett, J., Yao, J., Semnani-Azad, Z., Zhang, Z.-X., ... Adair, W.
(2016). Dignity, face, and honor cultures: A study of negotiation strategy and outcomes in
three cultures. Journal of Organizational Behavior, 37, 1178–1201.
Asparouhov, T., & Muthén, B. (2009). Exploratory structural equation modeling. Structural
Equation Modeling: A Multidisciplinary Journal, 16(3), 397-438.
Asparouhov, T., & Muthén, B. (2014a). Multiple-group factor analysis alignment. Structural
Equation Modeling: A Multidisciplinary Journal, 21(4), 495-508.

210

Atari, M. (2018). Culture of honor. In V. Zeigler-Hill & T. K. Shackelford (Eds.), Encyclopedia
of Personality and Individual Differences (pp. 1–4). Cham: Springer International
Publishing.
Atari, M., & Jamali, R. (2016). Dimensions of women’s mate preferences: Validation of a mate
preference scale in Iran. Evolutionary Psychology, 14, 1474704916651443.
Atari, M., Barbaro, N., Sela, Y., Shackelford, T. K., & Chegeni, R. (2017). The Big Five
personality dimensions and mate retention behaviors in Iran. Personality and Individual
Differences, 104, 286–290.
Atari, M., Barbaro, N., Shackelford, T. K., & Chegeni, R. (2017). Psychometric evaluation and
cultural correlates of the mate retention inventory–short form (MRI-SF) in Iran.
Evolutionary Psychology, 15, 1474704917695267.
Atari, M., Graham, J., & Dehghani, M. (2020b). Foundations of morality in Iran. Evolution and
Human Behavior, 41(5), 367-384.
Atari, M., Graham, J., & Dehghani, M. (2021). Toward a cumulative science of non-WEIRD
morality: a reply to Rezvani Nejad et al. Evolution and Human Behavior, 42(3), 277-278.
Atari, M., Lai, M. H., & Dehghani, M. (2020c). Sex differences in moral judgements across 67
countries. Proceedings of the Royal Society B, 287(1937), 20201201.
Atari, M., Mostafazadeh Davani, A., & Dehghani, M. (2020). Body Maps of Moral Concerns.
Psychological Science, 31, 160-169.
Atran, S., & Norenzayan, A. (2004). Religion's evolutionary landscape: Counterintuition,
commitment, compassion, communion. Behavioral and Brain Sciences, 27(6), 713-730.

211

Axelrod, R., & Hamilton, W. D. (1981). The evolution of cooperation. Science, 211(4489),
1390-1396.
Bastian, B., Vauclair, C.-M., Loughnan, S., Bain, P., Ashokkumar, A., Becker, M., . . . Swann,
W. B. (2019). Explaining illness with evil: pathogen prevalence fosters moral vitalism.
Proceedings of the Royal Society B: Biological Sciences, 286, 20191576.
Bauch, C. T., & McElreath, R. (2016). Disease dynamics and costly punishment can foster
socially imposed monogamy. Nature Communications, 7, 11219.
Beer, A., & Watson, D. (2009). The individual and group loyalty scales (IGLS): Construction
and preliminary validation. Journal of Personality Assessment, 91(3), 277-287.
Bell, A. V., Richerson, P. J., & McElreath, R. (2009). Culture rather than genes provides greater
scope for the evolution of large-scale human prosociality. Proceedings of the National
Academy of Sciences, 106, 17671–17674.
Berniūnas, R. (2020). Mongolian yos surtakhuun and WEIRD “morality”. Journal of Cultural
Cognitive Science, 4, 59–71.
Billing, J., & Sherman, P. W. (1998). Antimicrobial functions of spices: why some like it hot.
The Quarterly Review of Biology, 73, 3–49.
Bloom, P. (2012). Religion, morality, evolution. Annual review of psychology, 63, 179-199.
Bobocel, D. R., Davey, L. M., Son Hing, L. S., & Zanna, M. P. (2001). The concern for justice
and reactions to affirmative action: Cause or rationalization. Justice in the Workplace, 2,
121-143.
Boehm, C. (1999). Hierarchy in the forest. Harvard University Press.

212

Bojanowski, P., Grave, E., Joulin, A., & Mikolov, T. (2017). Enriching word vectors with
subword information. Transactions of the Association for Computational Linguistics, 5,
135–146.
Borg, J. S., Hynes, C., Van Horn, J., Grafton, S., & Sinnott-Armstrong, W. (2006).
Consequences, action, and intention as factors in moral judgments: An FMRI
investigation. Journal of Cognitive Neuroscience, 18(5), 803–817.
Borsboom, D., & Cramer, A. O. (2013). Network analysis: An integrative approach to the
structure of psychopathology. Annual Review of Clinical Psychology, 9, 91–121.
Bowles, S., Smith, E. A., & Borgerhoff Mulder, M. (2010). The emergence and persistence of
inequality in premodern societies. Current Anthropology, 51, 7–17.
Boyd, R., Richerson, P. J., & Henrich, J. (2011). The cultural niche: Why social learning is
essential for human adaptation. Proceedings of the National Academy of Sciences, 108,
10918–10925.
Bräuer, J., & Hanus, D. (2012). Fairness in non-human primates? Social Justice Research, 25(3),
256–276.
Brislin, R. W. (1970). Back-translation for cross-cultural research. Journal of Cross-Cultural
Psychology, 1, 185–216.
Brooks, A. C. (2008). Gross national happiness: Why happiness matters for America—and how
we can get more of it. New York: Basic Books.
Brosnan, S. F. (2013). Justice-and fairness-related behaviors in nonhuman primates. Proceedings
of the National Academy of Sciences, 110, 10416–10423.

213

Brosnan, S. F., & de Waal, F. B. (2003). Monkeys reject unequal pay. Nature, 425(6955), 297–
299.
Brosnan, S. F., & de Waal, F. B. (2014). Evolution of responses to (un)fairness. Science,
346(6207), 1251776.
Burke, E. (1790/2003). Reflections on the revolution in France. New Haven, CT: Yale
University Press.
Buss, D. M. (1988). From vigilance to violence: Tactics of mate retention in American
undergraduates. Ethology and Sociobiology, 9, 291–317.
Buss, D. M., Shackelford, T. K., & McKibbin, W. F. (2008). The mate retention inventory-short
form (MRI-SF). Personality and Individual Differences, 44, 322-334.
Buuren, S. v., & Groothuis-Oudshoorn, K. (2010). mice: Multivariate imputation by chained
equations in R. Journal of Statistical Software, 45, 1–68.
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the
multitrait-multimethod matrix. Psychological Bulletin, 56, 81–105.
Chao, F., Gerland, P., Cook, A. R., & Alkema, L. (2019). Systematic assessment of the sex ratio
at birth for all countries and estimation of national imbalances and regional reference
levels. Proceedings of the National Academy of Sciences, 116, 9303–9311.
Chegeni, R., Pirkalani, R. K., & Dehshiri, G. (2018). On love and darkness: The Dark Triad and
mate retention behaviors in a non-Western culture. Personality and Individual Differences,
122, 43-46.

214

Chen, F. F. (2008). What happens if we compare chopsticks with forks? the impact of making
inappropriate comparisons in cross-cultural research. Journal of Personality and Social
Psychology, 95, 1005–1018.
Christensen, A. P., Golino, H., & Silvia, P. J. (2020). A psychometric network perspective on the
validity and validation of personality trait questionnaires. European Journal of
Personality, 34(6), 1095-1108.
Chudek, M., & Henrich, J. (2011). Culture–gene coevolution, norm-psychology and the
emergence of human prosociality. Trends in cognitive sciences, 15(5), 218-226.
Cingel, D. P., & Krcmar, M. (2020). Considering Moral Foundations Theory and the Model of
Intuitive Morality and Exemplars in the context of child and adolescent development.
Annals of the International Communication Association, 44(2), 120-138.
Clifford, S., & Jerit, J. (2013). How words do the work of politics: Moral foundations theory and
the debate over stem cell research. The Journal of Politics, 75(3), 659-671.
Clifford, S., Iyengar, V., Cabeza, R., & Sinnott-Armstrong, W. (2015). Moral foundations
vignettes: A standardized stimulus database of scenarios based on moral foundations
theory. Behavior Research Methods, 47(4), 1178-1198.
Cohen, D., Nisbett, R. E., Bowdle, B. F., & Schwarz, N. (1996). Insult, aggression, and the
southern culture of honor: An “experimental ethnography”. Journal of Personality and
Social Psychology, 70, 945–960.

215

Collobert, R., Weston, J., Bottou, L., Karlen, M., Kavukcuoglu, K., & Kuksa, P. (2011). Natural
language processing (almost) from scratch. Journal of Machine Learning Research, 12,
2493–2537.
Costantini, G., Epskamp, S., Borsboom, D., Perugini, M., Mõttus, R., Waldorp, L. J., & Cramer,
A. O. (2015). State of the aRt personality research: A tutorial on network analysis of
personality data in R. Journal of Research in Personality, 54, 13–29.
Costello, T. H., Bowes, S. M., Stevens, S. T., Waldman, I. D., Tasimi, A., & Lilienfeld, S. O.
(2021). Clarifying the structure and nature of left-wing authoritarianism. Journal of
Personality and Social Psychology. Advance online publication.
https://doi.org/10.1037/pspp0000341
Culpepper, P. D., Havlíček, J., Leongómez, J. D., & Roberts, S. C. (2018). Visually activating
pathogen disgust: A new instrument for studying the behavioural immune system.
Frontiers in Psychology, 9, 1397.
Curtis, V., De Barra, M., & Aunger, R. (2011). Disgust as an adaptive system for disease
avoidance behaviour. Philosophical Transactions of the Royal Society B: Biological
Sciences, 366, 389–401.
Dalbert, C. (1999). The world is more just for me than generally: About the personal belief in a
just world scale's validity. Social justice research, 12(2), 79-98.
Damasio, A. (2019). The strange order of things: Life, feeling, and the making of cultures.
Vintage.

216

Damasio, A., & Carvalho, G. B. (2013). The nature of feelings: Evolutionary and
neurobiological origins. Nature Reviews Neuroscience, 14, 143–152.
Daryaee, T. (2009). The study of ancient Iran in the twentieth century. Iranian Studies, 42, 579–
589.
Davey, L. M., Bobocel, D. R., Hing, L. S. S., & Zanna, M. P. (1999). Preference for the Merit
Principle Scale: An individual difference measure of distributive justice preferences.
Social Justice Research, 12(3), 223-240.
Davies, C. L., Sibley, C. G., & Liu, J. H. (2014). Confirmatory factor analysis of the Moral
Foundations Questionnaire. Social Psychology, 45, 431-436.
Davis, D. E., Dooley, M. T., Hook, J. N., Choe, E., & McElroy, S. E. (2017). The purity/sanctity
subscale of the Moral Foundations Questionnaire does not work similarly for religious
versus non-religious individuals. Psychology of Religion and Spirituality, 9(1), 124.
Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a
multidimensional approach. Journal of Personality and Social Psychology, 44, 113–126.
Debove, S., Baumard, N., & André, J.-B. (2017). On the evolutionary origins of equity. PLoS
One, 12(3), e0173636.
Dehghani, M., Atran, S., Iliev, R., Sachdeva, S., Medin, D., & Ginges, J. (2010). Sacred values
and conflict over Iran's nuclear program. Judgment and Decision Making, 5, 540-546.
Dehghani, M., Johnson, K., Hoover, J., Sagi, E., Garten, J., Parmar, N. J., . . . Graham, J. (2016).
Purity homophily in social networks. Journal of Experimental Psychology: General, 145,
366–375.

217

Del Giudice, M. (2009). On the real magnitude of psychological sex differences. Evolutionary
Psychology, 7, 264-279.
Del Giudice, M. (2019). Measuring sex differences and similarities. In D. P. VanderLaan & W.
I. Wong (Eds.), Gender and sexuality development: Contemporary theory and research.
New York, NY: Springer.
Delkhamosh, M. T. (2004). Identification of Iranian students’ values and norms. Journal of
Iranian Psychologists, 5, 15–34.
Doğruyol, B., Alper, S., & Yilmaz, O. (2019). The five-factor model of the moral foundations
theory is stable across WEIRD and non-WEIRD cultures. Personality and Individual
Differences, 151, 109547.
Douglas, M. (1966). Purity and danger. London, England: Routledge.
Epskamp, S., Cramer, A. O., Waldorp, L. J., Schmittmann, V. D., & Borsboom, D. (2012).
qgraph: Network visualizations of relationships in psychometric data. Journal of Statistical
Software, 48, 1–18.
Farahani, M., Farzad, V., & Fotoohi, M. (2004). A lexical study of personality factors in Farsi.
Journal of Psychology, 8, 183–204.
Farzianpour, F., Eshraghian, M., Emami, A., Hosseini, S., Hosseini, S., & Farhud, D. (2011). An
estimate of happiness among students of Tehran University of Medical Sciences: A means
for policy making in management of health system. Iranian Red Crescent Medical
Journal, 13, 841.

218

Federico, C. M., Weber, C. R., Ergun, D., & Hunt, C. (2013). Mapping the connections between
politics and morality: The multiple sociopolitical orientations involved in moral intuition.
Political Psychology, 34(4), 589-610.
Fehr, E., Bernhard, H., & Rockenbach, B. (2008). Egalitarianism in young children. Nature,
454(7208), 1079–1083.
Fincher, C. L., & Thornhill, R. (2012). Parasite-stress promotes in-group assortative sociality:
The cases of strong family ties and heightened religiosity. Behavioral and Brain Sciences,
35, 61–79.
Fincher, C. L., Thornhill, R., Murray, D. R., & Schaller, M. (2008). Pathogen prevalence
predicts human cross-cultural variability in individualism/collectivism. Proceedings of the
Royal Society B: Biological Sciences, 275, 1279–1285.
Fiske, A. P. (1992). Four elementary forms of sociality: Framework for a unified theory of social
relations. Psychological Review, 99, 689 –723.
Flake, J. K., & Fried, E. I. (2020). Measurement schmeasurement: Questionable measurement
practices and how to avoid them. Advances in Methods and Practices in Psychological
Science, 3(4), 456-465.
Flake, J. K., Pek, J., & Hehman, E. (2017). Construct validation in social and personality
research: Current practice and recommendations. Social Psychological and Personality
Science, 8(4), 370-378.
Flannery, K., & Marcus, J. (2012). The creation of inequality. Harvard University Press.

219

Flora, D. B. (2020). Your coefficient alpha is probably wrong, but which coefficient omega is
right? A tutorial on using R to obtain better reliability estimates. Advances in Methods and
Practices in Psychological Science, 3(4), 484-501.
Friedman, J., Hastie, T., & Tibshirani, R. (2008). Sparse inverse covariance estimation with the
graphical lasso. Biostatistics, 9(3), 432-441.
Frimer, J. A. (2018). Moral heroes are puppets. In K. Gray & J. Graham (Eds.), Atlas of moral
psychology. New York, NY: Guilford Press.
Frimer, J. A., & Sinclair, L. (2016). Moral heroes look up and to the right. Personality and
Social Psychology Bulletin, 42, 400–410.
Frimer, J. A., Boghrati, R., Haidt, J., Graham, J., & Dehghani, M. (2019). Moral foundations
dictionary for linguistic analyses 2.0 [Unpublished manuscript].
Gächter, S., & Schulz, J. F. (2016). Intrinsic honesty and the prevalence of rule violations across
societies. Nature, 531, 496.
Garg, N., Schiebinger, L., Jurafsky, D., & Zou, J. (2018). Word embeddings quantify 100 years
of gender and ethnic stereotypes. Proceedings of the National Academy of Sciences, 115,
E3635–E3644.
Garten, J., Hoover, J., Johnson, K. M., Boghrati, R., Iskiwitch, C., & Dehghani, M. (2018).
Dictionaries and distributions: Combining expert knowledge and large scale textual data
content analysis. Behavior Research Methods, 50, 344–361.

220

Gelfand, M., Jackson, J. C., Pan, X., Nau, D., Dagher, M. M., & Chiu, C.-y. (2020). Cultural and
institutional factors predicting the infection rate and mortality likelihood of the COVID-19
pandemic. PsyArXiv. Retrieved from psyarxiv.com/m7f8a
Gelfand, M., Raver, J. L., Nishii, L., Leslie, L. M., Lun, J., Lim, B. C., . . . Yamaguchi, S.
(2011). Differences between tight and loose cultures: A 33-nation study. Science, 332,
1100–1104.
Gheissari, A. (2009). Contemporary Iran: Economy, society, politics. New York, NY: Oxford
University Press.
Golino, H. F., & Epskamp, S. (2017). Exploratory graph analysis: A new approach for
estimating the number of dimensions in psychological research. PloS one, 12(6),
e0174035.
Golino, H., Shi, D., Christensen, A. P., Garrido, L. E., Nieto, M. D., Sadana, R., Thiyagarajan, J.
A., & Martinez-Molina, A. (2020). Investigating the performance of exploratory graph
analysis and traditional techniques to identify the number of latent factors: A simulation
and tutorial. Psychological Methods, 25(3), 292–320.
Gosling, S. D., Rentfrow, P. J., & Swann Jr, W. B. (2003). A very brief measure of the Big-Five
personality domains. Journal of Research in Personality, 37, 504–528.
Goudarzi, S., Pliskin, R., Jost, J. T., & Knowles, E. D. (2020). Economic system justification
predicts muted emotional responses to inequality. Nature communications, 11(1), 1-9.
Graham, J. (2010). Left gut, right gut: Ideology and automatic moral reactions, Doctoral
dissertation. Retrieved from ProQuest Dissertations and Theses (AAT 3437423).

221

Graham, J., & Haidt, J. (2010). Beyond beliefs: Religions bind individuals into moral
communities. Personality and Social Psychology Review, 14, 140–150.
Graham, J., & Haidt, J. (2012). Sacred values and evil adversaries: A Moral Foundations
approach. In P. Shaver & M. Mikulincer (Eds.), The social psychology of morality:
Exploring the causes of good and evil. New York: APA Books.
Graham, J., Haidt, J., & Nosek, B. A. (2009). Liberals and conservatives rely on different sets of
moral foundations. Journal of Personality and Social Psychology, 96, 1029–1046.
Graham, J., Haidt, J., Koleva, S., Motyl, M., Iyer, R., Wojcik, S. P., & Ditto, P. H. (2013). Moral
foundations theory: The pragmatic validity of moral pluralism. Advances in Experimental
Social Psychology, 47, 55–130.
Graham, J., Meindl, P., Koleva, S., Iyer, R., & Johnson, K. M. (2015). When values and
behavior conflict: Moral pluralism and intrapersonal moral hypocrisy. Social and
Personality Psychology Compass, 9, 158-170.
Graham, J., Nosek, B. A., Haidt, J., Iyer, R., Koleva, S., & Ditto, P. H. (2011). Mapping the
moral domain. Journal of Personality and Social Psychology, 101, 366–385.
Graham, J., Nosek, B.A., & Haidt, J. (2012). The moral stereotypes of liberals and
conservatives: Exaggeration of differences across the political spectrum. PLoS One, 7,
e50092.
Grave, E., Bojanowski, P., Gupta, P., Joulin, A., & Mikolov, T. (2018). Learning word vectors
for 157 languages. In Proceedings of the international conference on language resources
and evaluation (LREC 2018).

222

Greene, J. D., Sommerville, R. B., Nystrom, L. E., Darley, J. M., & Cohen, J. D. (2001,
September 14). An fMRI investigation of emotional engagement in moral judgment.
Science, 293, 2105–2108.
Greenfield, P. M. (2013). The changing psychology of culture from 1800 through 2000.
Psychological Science, 24, 1722–1731.
Grosjean, P. (2014). A history of violence: The culture of honor and homicide in the US South.
Journal of the European Economic Association, 12, 1285-1316.
Grossmann, I., & Varnum, M. E. (2015). Social structure, infectious diseases, disasters,
secularism, and cultural change in America. Psychological Science, 26, 311–324.
Guerra, V. M., Giner-Sorolla, R., & Vasiljevic, M. (2013). The importance of honor concerns
across eight countries. Group Processes & Intergroup Relations, 16, 298-318.
Gurven, M. D. (2018). Broadening horizons: Sample diversity and socioecological theory are
essential to the future of psychological science. Proceedings of the National Academy of
Sciences, 115(45), 11420-11427
Haidt, J. (2001). The emotional dog and its rational tail: A social intuitionist approach to moral
judgment. Psychological Review, 108, 814 – 834.
Haidt, J. (2007). The new synthesis in moral psychology. Science, 316, 998–1002.
Haidt, J. (2012). The righteous mind: Why good people are divided by politics and religion.
Vintage.
Haidt, J., & Graham, J. (2007). When morality opposes justice: Conservatives have moral
intuitions that liberals may not recognize. Social Justice Research, 20, 98 –116.

223

Haidt, J., & Joseph, C. (2004). Intuitive ethics: How innately prepared intuitions generate
culturally variable virtues. Daedalus, 133, 55–66.
Haidt, J., & Joseph, C. (2008). The moral mind: How five sets of innate intuitions guide the
development of many culture-specific virtues, and perhaps even modules. In P. Carruthers,
S. Laurence, & S. Stich (Eds.), Evolution and cognition: The innate mind Vol. 3,
Foundations and the future (pp. 367–391). New York, NY: Oxford University Press.
Haidt, J., & Joseph, C. (2011). How Moral Foundations Theory succeeded in building on sand:
A response to Suhler and Churchland. Journal of Cognitive Neuroscience, 23, 2117–2122.
Haidt, J., McCauley, C., & Rozin, P. (1994). Individual differences in sensitivity to disgust: A
scale sampling seven domains of disgust elicitors. Personality and Individual Differences,
16, 701–713.
Hanretty, C., Lauderdale, B. E., & Vivyan, N. (2016). Comparing strategies for estimating
constituency opinion from national survey samples. Political Science Research and
Methods, 6, 571–591.
Harper, C. A., & Rhodes, D. (2021). Reanalysing the factor structure of the moral foundations
questionnaire. British Journal of Social Psychology. In press.
Hatemi, P. K., Crabtree, C., & Smith, K. B. (2019). Ideology justifies morality: Political beliefs
predict moral foundations. American Journal of Political Science, 63(4), 788-806.
Hauser, D. J., & Schwarz, N. (2016). Attentive Turkers: MTurk participants perform better on
online attention checks than do subject pool participants. Behavior Research Methods,
48(1), 400-407.

224

Haynie, H. J., Kavanagh, P. H., Jordan, F. M., Ember, C. R., Gray, R. D., Greenhill, S. J., ... &
Gavin, M. C. (2021). Pathways to social inequality. Evolutionary Human Sciences, 3.
Hayton, J. C., Allen, D. G., & Scarpello, V. (2004). Factor retention decisions in exploratory
factor analysis: A tutorial on parallel analysis. Organizational Research Methods, 7, 191–
205.
Henrich J. (2020). The WEIRDest People in the World: How the West Became Psychologically
Peculiar and Particularly Prosperous. New York: Farrar, Straus & Giroux.
Henrich, J. (2015). The secret of our success: How culture is driving human evolution,
domesticating our species, and making us smarter. Princeton University Press.
Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world?
Behavioral and Brain Sciences, 33, 61–83.
Herfehdoust, M., Falsafinejad, M., Delavar, A., Sohrabi, F., & Tamimdari, A. (2015). Extracting
the personality factors based on lexical study of the Persian language. Educational
Measurement, 6, 83–112.
Ho, A. K., Sidanius, J., Kteily, N., Sheehy-Skeffington, J., Pratto, F., Henkel, K. E., ... &
Stewart, A. L. (2015). The nature of social dominance orientation: Theorizing and
measuring preferences for intergroup inequality using the new SDO₇ scale. Journal of
Personality and Social Psychology, 109(6), 1003.
Hofstede, G. (2010). National cultural dimensions. Retrieved from https://www.hofstede-
insights.com/country-comparison/

225

Hogan, R., & Nicholson, R. A. (1988). The meaning of personality test scores. American
Psychologist, 43, 621– 626.
Hogan, R., Johnson, J. A., & Emler, N. P. (1978). A socioanalytic theory of moral development.
In W. Damon (Ed.), New directions for child development: Moral development (Vol. 2, pp.
1–18). San Francisco, CA: Jossey-Bass.
Homans, G. C. (1961). Social behavior: Its elementary forms. New York: Harcourt, Brace, and
World.
Hooshyari, Z., Delavar, A., Minaee, A., & Eskandari, H. (2018). Comparison of two approaches
for measuring moral development: Neo-Kohlbergian approach and moral study challenges.
Iranian Journal of Psychiatry and Clinical Psychology, 24, 30–43.
Hoover, J., & Dehghani, M. (2020). The big, the bad, and the ugly: Geographic estimation with
flawed psychological data. Psychological Methods, 25, 412–429.
Hoover, J., Atari, M., Davani, A. M., Kennedy, B., Portillo-Wightman, G., Yeh, L., & Dehghani,
M. (2021). Investigating the role of group-based morality in extreme behavioral
expressions of prejudice. Nature Communications, 12(1), 1-13.
Hoover, J., Atari, M., Mostafazadeh Davani, A., Kennedy, B., Portillo-Wightman, G., Yeh, L., . .
. Dehghani, M. (2019). Bound in hatred: The role of group-based morality in acts of hate.
PsyArXiv. Retrieved from psyarxiv.com/359me
Hoover, J., Portillo-Wightman, G., Yeh, L., Havaldar, S., Davani, A. M., Lin, Y., ... &
Dehghani, M. (2020). Moral Foundations Twitter Corpus: A collection of 35k tweets

226

annotated for moral sentiment. Social Psychological and Personality Science, 11(8), 1057-
1071.
Hruschka, D. J., & Henrich, J. (2013). Institutions, parasites and the persistence of in-group
preferences. PLoS One, 8, 1-9.
Hu, L.-t., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling: A
Multidisciplinary Journal, 6, 1–55.
Huppert, E., Cowell, J. M., Cheng, Y., Contreras‐Ibáñez, C., Gomez‐Sicard, N., Gonzalez‐
Gadea, M. L., ... & Decety, J. (2019). The development of children's preferences for
equality and equity across 13 individualistic and collectivist cultures. Developmental
Science, 22(2), e12729.
IJzerman, H., van Dijk, W. W., & Gallucci, M. (2007). A bumpy train ride: A field experiment
on insult, honor, and emotional reactions. Emotion, 7, 869.
International Monetary Fund. (2019). World Economic Outlook Database. Retrieved from:
https://www.imf.org/external/pubs/ft/weo/2019/01/weodata/download.aspx
ISNA. (2019). Illiteracy rates in Iran from 60 years ago to today. Iranian Students’ News
Agency. Retrieved 2019-07-29, from https://www.isna.ir/news/97120603224/
Iurino, K., & Saucier, G. (2020). Testing measurement invariance of the Moral Foundations
Questionnaire across 27 countries. Assessment, 27(2), 365-372.
Iyer, R., Koleva, S., Graham, J., Ditto, P., & Haidt, J. (2012). Understanding libertarian morality:
The psychological dispositions of self-identified libertarians. PLoS One, 7, e42366.

227

Jackson, A. W. (1896). The moral and ethical teachings of the ancient zoroastrian religion. The
International Journal of Ethics, 7, 55–62.
Jackson, J. C., Gelfand, M., De, S., & Fox, A. (2019). The loosening of American culture over
200 years is associated with a creativity–order trade-off. Nature Human Behaviour, 3,
244–250.
Jamalzadeh, M. A. (1966). Kholghiate Ma Iranian. [Temperaments of us, Iranians]. Tehran:
Foroughi Bookstore.
Janoff-Bulman, R., & Carnes, N. C. (2013a). Moral context matters: A reply to Graham.
Personality and Social Psychology Review, 17, 242–247.
Janoff-Bulman, R., & Carnes, N. C. (2013b). Surveying the moral landscape: Moral motives and
group-based moralities. Personality and Social Psychology Review, 17, 219–236.
Johnson, K. A., Hook, J. N., Davis, D. E., Van Tongeren, D. R., Sandage, S. J., & Crabtree, S.
A. (2016). Moral foundation priorities reflect US Christians' individual differences in
religiosity. Personality and Individual Differences, 100, 56-61.
Jost, J. T., & Thompson, E. P. (2000). Group-based dominance and opposition to equality as
independent predictors of self-esteem, ethnocentrism, and social policy attitudes among
African Americans and European Americans. Journal of Experimental Social Psychology,
36(3), 209-232.
Jost, J. T., Federico, C. M., & Napier, J. L. (2009). Political ideology: Its structure, functions,
and elective affinities. Annual Review of Psychology, 60, 307-337.

228

Jost, J. T., Hawkins, C. B., Nosek, B. A., Hennes, E. P., Stern, C., Gosling, S. D., & Graham, J.
(2014). Belief in a just God (and a just society): A system justification perspective on
religious ideology. Journal of Theoretical and Philosophical Psychology, 34(1), 56.
Kanagawa, C., Cross, S. E., & Markus, H. R. (2001). “Who am I?” the cultural psychology of
the conceptual self. Personality and Social Psychology Bulletin, 27, 90–103.
Kandrik, M., Jones, B. C., & DeBruine, L. M. (2015). Scarcity of female mates predicts regional
variation in men’s and women’s sociosexual orientation across us states. Evolution and
Human Behavior, 36, 206–210.
Karimi-Malekabadi, F., & Esmaeilinasab, M. (2019). Religiosity, intrasexual rivalry, and mate
retention behaviors in Iran. Personality and Individual Differences, 149, 135–140.
Karimi-Malekabadi, F., Trager, J., Reimer, N., Kennedy, B., Atari, M., Graham, J., & Dehghani,
M. (2021). County-level moral values predict COVID-19 vaccination rate in the United
States. PsyArXiv. https://psyarxiv.com/z6kxm/
Katouzian, H. (1981). The political economy of modern Iran: Despotism and pseudo-modernism,
1926–1979. Springer.
Kennedy, B., Atari, M., Davani, A. M., Hoover, J., Omrani, A., Graham, J., & Dehghani, M.
(2021). Moral concerns are differentially observable in language. Cognition, 212, 104696.
Khodayarifard, M., Abdollahi, A., Akbari-Zardkhaneh, S., Afrooz, G., Rahiminezhad, A.,
Paknejad, M., ... Hood, R. W. (2018). Abrahamic religiosity scale: Development and initial
validation. Mental Health, Religion & Culture, 1–11.

229

Khodayarifard, M., Rahiminezhad, A., Ghobari Bonab, B., Shokohi Yekta, M., Alinaghiye, F.,
Azarbayjani, M., ... Bahrami-Ehsan, H. (2010). Religiosity measuring model and
preparation of religiosity scale to Iranian population. Journal of Applied Psychological
Research, 1, 1–24.
Kim, H. S., Sherman, D. K., & Updegraff, J. A. (2016). Fear of ebola: The influence of
collectivism on xenophobic threat responses. Psychological Science, 27, 935–944.
Kitayama, S., Park, H., Sevincer, A. T., Karasawa, M., & Uskul, A. K. (2009). A cultural task
analysis of implicit independence: Comparing North America, Western Europe, and East
Asia. Journal of Personality and Social Psychology, 97, 236–255.
Kivikangas, J. M., Fernández-Castilla, B., Järvelä, S., Ravaja, N., & Lönnqvist, J. E. (2021).
Moral foundations and political orientation: Systematic review and meta-analysis.
Psychological Bulletin, 147(1), 55.
Klein, R. A., Vianello, M., Hasselman, F., Adams, B. G., Adams Jr, R. B., Alper, S., ... &
Sowden, W. (2018). Many Labs 2: Investigating variation in replicability across samples
and settings. Advances in Methods and Practices in Psychological Science, 1(4), 443-490.
Knauft, B. M., Abler, T. S., Betzig, L., Boehm, C., Dentan, R. K., Kiefer, T. M., ... & Rodseth,
L. (1991). Violence and sociality in human evolution [and comments and replies]. Current
Anthropology, 32(4), 391-428.
Knowles, E. D., Lowery, B. S., Hogan, C. M., & Chow, R. M. (2009). On the malleability of
ideology: Motivated construals of color blindness. Journal of Personality and Social
Psychology, 96(4), 857.

230

Koenig, H. G., Meador, K. G., & Parkerson, G. (1997). Religion index for psychiatric research.
American Journal of Psychiatry, 154, 885–886.
Koleva, S. P., Graham, J., Iyer, R., Ditto, P. H., & Haidt, J. (2012). Tracing the threads: How
five moral concerns (especially Purity) help explain culture war attitudes. Journal of
Research in Personality, 46(2), 184-194.
Koleva, S., Selterman, D., Iyer, R., Ditto, P., & Graham, J. (2014). The moral compass of
insecurity: Anxious and avoidant attachment predict moral judgment. Social Psychological
and Personality Science, 5, 185–194.
Konow, J. (2000). Fair shares: Accountability and cognitive dissonance in allocation decisions.
American Economic Review, 90(4), 1072-1091.
Koppes, C. R. (1976). Captain Mahan, general Gordon, and the origins of the term ‘Middle
East’. Middle Eastern Studies, 12, 95–98.
Krämer, N., Schäfer, J., & Boulesteix, A.-L. (2009). Regularized estimation of large-scale gene
association networks using graphical gaussian models. BMC Bioinformatics, 10, 384.
Krosnick, J. A., & Fabrigar, L. R. (1997). Designing rating scales for effective measurement in
surveys. In L. Lyberg, P. Biemer, M. Collins, E. De Leeuw, C. Dippo, N. Schwarz, & D.
Trewin (Eds.), Survey measurement and process quality (pp. 141-164). New York: John
Wiley & Sons, Inc.
Krupka, E. L., & Weber, R. A. (2013). Identifying social norms using coordination games: Why
does dictator game sharing vary?. Journal of the European Economic Association, 11(3),
495-524.

231

Laajaj, R., Macours, K., Hernandez, D. A. P., Arias, O., Gosling, S. D., Potter, J., ... Vakis, R.
(2019). Challenges to capture the Big Five personality traits in non-WEIRD populations.
Science Advances, 5, eaaw5226.
Landy, J. F., & Bartels, D. M. (2018). An empirically-derived taxonomy of moral concepts.
Journal of Experimental Psychology: General, 147, 1748–1761.
Lang, M., Purzycki, B. G., Apicella, C. L., Atkinson, Q. D., Bolyanatz, A., Cohen, E., ...
Henrich, J. (2019). Moralizing gods, impartiality and religious parochialism across 15
societies. Proceedings of the Royal Society B, 286, 20190202.
Larijani, B., & Zahedi, F. (2008). Contemporary medical ethics: An overview from Iran.
Developing World Bioethics, 8, 192–196.
Lee, N., & Cadogan, J. W. (2013). Problems with formative and higher-order reflective
variables. Journal of Business Research, 66(2), 242-247.
Leemann, L., & Wasserfallen, F. (2017). Extending the use and prediction precision of
subnational public opinion estimation: Extending use and precision of MrP. American
Journal of Political Science, 61, 1003–1022.
Levenson, M. R., Kiehl, K. A., & Fitzpatrick, C. M. (1995). Assessing psychopathic attributes in
a noninstitutionalized population. Journal of Personality and Social Psychology, 68, 151–
158.
Lewis, G. J., & Bates, T. C. (2011). From left to right: How the personality system allows basic
traits to influence politics via characteristic moral adaptations. British Journal of
Psychology, 102, 546–558.

232

Liénard, P., Chevallier, C., Mascaro, O., Kiura, P., & Baumard, N. (2013). Early understanding
of merit in Turkana children. Journal of Cognition and Culture, 13(1-2), 57–66.
Litman, L., Robinson, J., & Abberbock, T. (2017). TurkPrime. com: A versatile crowdsourcing
data acquisition platform for the behavioral sciences. Behavior Research Methods, 49(2),
433-442.
Marcus, G. (2004). The birth of the mind. New York: Basic.
Marsh, H. W., Morin, A. J., Parker, P. D., & Kaur, G. (2014). Exploratory structural equation
modeling: An integration of the best features of exploratory and confirmatory factor
analysis. Annual Review of Clinical Psychology, 10, 85-110.
Massen, J. J., Van Den Berg, L. M., Spruijt, B. M., & Sterck, E. H. (2012). Inequity aversion in
relation to effort and relationship quality in long-tailed macaques (macaca fascicularis).
American Journal of Primatology, 74(2), 145–156.
McAdams, D. P., & Guo, J. (2015). Narrating the generative life. Psychological Science, 26,
475–483.
McAdams, D. P., Albaugh, M., Farber, E., Daniels, J., Logan, R. L., & Olson, B. (2008). Family
metaphors and moral intuitions: How conservatives and liberals narrate their lives. Journal
of Personality and Social Psychology, 95(4), 978.
McCrae, R. R., & John, O. P. (1992). An introduction to the five-factor model and its
applications. Journal of Personality, 60, 175–215.
McKay, R., & Whitehouse, H. (2015). Religion and morality. Psychological Bulletin, 141(2),
447.

233

Medin, D. (2017). Psychological science as a complex system: Report card. Perspectives on
Psychological Science, 12, 669-674.
Medin, D., Bennis, W., & Chandler, M. (2010). Culture and the home-field disadvantage.
Perspectives on Psychological Science, 5, 708–713.
Medin, D., Ojalehto, B., Marin, A., & Bang, M. (2017). Systems of (non-)diversity. Nature
Human Behaviour, 1, 0088.
Mesoudi, A. (2019). Cultural evolution and cultural psychology. In D. Cohen & S. Kitayama
(Eds.), Handbook of cultural psychology (2nd ed., pp. 144–162). The Guilford Press.
Métayer, S., & Pahlavan, F. (2014). Validation de l'adaptation française du questionnaire des
principes moraux fondateurs. Revue internationale de psychologie sociale, 27, 79-107.
Michel, J.-B., Shen, Y. K., Aiden, A. P., Veres, A., Gray, M. K., Pickett, J. P., . . . Aiden, E. L.
(2011). Quantitative analysis of culture using millions of digitized books. Science, 331,
176–182.
Mikolov, T., Sutskever, I., Chen, K., Corrado, G. S., & Dean, J. (2013). Distributed
representations of words and phrases and their compositionality. In Advances in neural
information processing systems (pp. 3111–3119).
Mill, J. S. (2003). On liberty. New Haven, CT: Yale University Press. (Original work published
1859)
MIT Election Data and Science Lab. (2018). County presidential election returns 2000-2016.
Harvard Dataverse. doi: 10.7910/DVN/VOQCHQ

234

Moll, J., de Oliveira-Souza, R., Moll, F. T., Ignácio, F. A., Bramati, I. E., Caparelli-Dáquer, E.
M., & Eslinger, P. J. (2005). The moral affiliations of disgust: A functional MRI study.
Cognitive and Behavioral Neurology, 18(1), 68-78.
Moreira, L. V., Souza, M. L. D., & Guerra, V. M. (2019). Validity evidence of a Brazilian
version of the Moral Foundations Questionnaire. Psicologia: Teoria e Pesquisa, 35.
Morin, A. J., Arens, A. K., & Marsh, H. W. (2016). A bifactor exploratory structural equation
modeling framework for the identification of distinct sources of construct-relevant
psychometric multidimensionality. Structural Equation Modeling: A Multidisciplinary
Journal, 23(1), 116-139.
Mortensen, C. R., Becker, D. V., Ackerman, J. M., Neuberg, S. L., & Kenrick, D. T. (2010).
Infection breeds reticence: The effects of disease salience on self-perceptions of
personality and behavioral avoidance tendencies. Psychological Science, 21, 440–447.
Moshontz, H., Campbell, L., Ebersole, C. R., IJzerman, H., Urry, H. L., Forscher, P. S., ... &
Chartier, C. R. (2018). The Psychological Science Accelerator: Advancing psychology
through a distributed collaborative network. Advances in Methods and Practices in
Psychological Science, 1(4), 501-515.
Mujtaba, B. G., Tajaddini, R., & Chen, L. Y. (2011). Business ethics perceptions of public and
private sector Iranians. Journal of Business Ethics, 104, 433–447.
Murray, D. R., & Schaller, M. (2010). Historical prevalence of infectious diseases within 230
geopolitical regions: A tool for investigating origins of culture. Journal of Cross-Cultural
Psychology, 41, 99–108.

235

Murray, D. R., & Schaller, M. (2016). The behavioral immune system: Implications for social
cognition, social interaction, and social influence. Advances in Experimental Social
Psychology, 53, 75–129.
Murray, D. R., Kerry, N., & Gervais, W. M. (2019). On disease and deontology: Multiple tests
of the influence of disease threat on moral vigilance. Social Psychological and Personality
Science, 10, 44–52.
Murray, D. R., Schaller, M., & Suedfeld, P. (2013). Pathogens and politics: Further evidence that
parasite prevalence predicts authoritarianism. PloS One, 8, e62275.
Muthén, B., & Asparouhov, T. (2014). IRT studies of many groups: The alignment method.
Frontiers in Psychology, 5, 978
Muthukrishna, M., & Henrich, J. (2019). A problem in theory. Nature Human Behaviour, 3, 221-
229.
Muthukrishna, M., Bell, A. V., Henrich, J., Curtin, C. M., Gedranovich, A., McInerney, J., &
Thue, B. (2020). Beyond Western, Educated, Industrial, Rich, and Democratic (WEIRD)
psychology: Measuring and mapping scales of cultural and psychological distance.
Psychological Science, 31, 678–701.
Navarrete, C. D., & Fessler, D. M. (2006). Disease avoidance and ethnocentrism: The effects of
disease vulnerability and disgust sensitivity on intergroup attitudes. Evolution and Human
Behavior, 27, 270–282.
Nejat, P., & Hatami, J. (2019). Psychometric properties of the Persian version of Moral
Foundations Questionnaire in three Iranian samples. Social Cognition, 8, 107-124.

236

Nejat, P., Bagherian, F., Hatami, J., & Shokri, O. (2015). Examining social-cognitive
characteristics of the mental representation of moral concerns: Grouping, motivational
orientation and relational context of moral foundations. Social Cognition, 4, 109–126.
Neuberg, S. L., Kenrick, D. T., & Schaller, M. (2011). Human threat management systems: Self-
protection and disease avoidance. Neuroscience & Biobehavioral Reviews, 35, 1042–1051.
Nilsson, A., & Erlandsson, A. (2015). The Moral Foundations taxonomy: Structural validity and
relation to political ideology in Sweden. Personality and Individual Differences, 76, 28-32.
Nilsson, A., Erlandsson, A., & Västfjäll, D. (2020). Moral foundations theory and the
psychology of charitable giving. European Journal of Personality, 34(3), 431-447.
Norenzayan, A. (2013). Big Gods: How religion transformed cooperation and conflict.
Princeton, NJ: Princeton University Press.
Norenzayan, A. (2016). Theodiversity. Annual review of psychology, 67, 465-488.
Nosworthy, G. J., Lea, J. A., & Lindsay, R. C. L. (1995). Opposition to Affirmative Action:
Racial Affect and Traditional Value Predictors Across Four Programs 1. Journal of
Applied Social Psychology, 25(4), 314-337.
Novin, S., & Oyserman, D. (2016). Honor as cultural mindset: Activated honor mindset affects
subsequent judgment and attention in mindset-congruent ways. Frontiers in Psychology, 7,
1921.
Nowak, A., Gelfand, M. J., Borkowski, W., Cohen, D., & Hernandez, I. (2016). The
evolutionary basis of honor cultures. Psychological Science, 27, 12–24.

237

Oishi, S., & Graham, J. (2010). Social ecology: Lost and found in psychological science.
Perspectives on Psychological Science, 5, 356–377.
Olatunji, B. O., Williams, N. L., Tolin, D. F., Abramowitz, J. S., Sawchuk, C. N., Lohr, J. M., &
Elwood, L. S. (2007). The Disgust Scale: item analysis, factor structure, and suggestions
for refinement. Psychological Assessment, 19(3), 281.
Oyserman, D. (2011). Culture as situated cognition: Cultural mindsets, cultural fluency, and
meaning making. European Review of Social Psychology, 22(1), 164-214.
Oyserman, D. (2017). Culture three ways: Culture and subcultures within countries. Annual
Review of Psychology, 68, 435-463.
Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and
collectivism: evaluation of theoretical assumptions and meta-analyses. Psychological
Bulletin, 128, 3–72.
Park, D. K., Gelman, A., & Bafumi, J. (2004). Bayesian multilevel estimation with
poststratification: State-level estimates from national polls. Political Analysis, 12, 375-
385.
Parvini, S., & Simani, E. (2019). Are Arabs and Iranians white? census says yes, but many
disagree. Los Angeles Times. Retrieved 2019-07-11, from https://
www.latimes.com/projects/la-me-census-middle-east-north-africa-race/
Peristiany, J. G. (1965). Honour and shame: The values of Mediterranean society. London:
Weidenfeld and Nicolson.

238

Petersen, M. B., Sznycer, D., Sell, A., Cosmides, L., & Tooby, J. (2013). The ancestral logic of
politics: Upper-body strength regulates men’s assertion of self-interest over economic
redistribution. Psychological Science, 24(7), 1098-1103.
Piazza, J., & Sousa, P. (2014). Religiosity, political orientation, and consequentialist moral
thinking. Social Psychological and Personality Science, 5(3), 334-342.
Pinker, S. (2003). The blank slate: The modern denial of human nature. Penguin.
Pons, P., & Latapy, M. (2006). Computing communities in large networks using random walks.
Journal of Graph Algorithms and Applications, 10, 191-218.
Proctor, D., Williamson, R. A., de Waal, F. B., & Brosnan, S. F. (2013). Chimpanzees play the
ultimatum game. Proceedings of the National Academy of Sciences, 110(6), 2070–2075
Purzycki, B. G., Pisor, A. C., Apicella, C., Atkinson, Q., Cohen, E., Henrich, J., … Xygalatas, D.
(2018). The cognitive and cultural foundations of moral behavior. Evolution and Human
Behavior, 39, 490–501.
R Core Team. (2017). R: A language and environment for statistical computing. Vienna,
Austria: R Foundation for Statistical Computing. Retrieved from https://www.R-
project.org/
Rad, M. S., Martingano, A. J., & Ginges, J. (2018). Toward a psychology of Homo Sapiens:
Making psychological science more representative of the human population. Proceedings
of the National Academy of Sciences, 115, 11401–11405.
Rai, T. S., & Fiske, A. P. (2011). Moral psychology is relationship regulation: moral motives for
unity, hierarchy, equality, and proportionality. Psychological Review, 118(1), 57.

239

Rasinski, K. A. (1987). What's fair is fair—Or is it? Value differences underlying public views
about social justice. Journal of Personality and Social Psychology, 53(1), 201.
Razavi, P., Shaban-Azad, H., & Srivastava, S. (2020). Gheirat as a Complex Emotional Reaction
to Relational Boundary Violations. https://doi.org/10.31234/osf.io/8ev23
Revelle, W. R. (2017). psych: Procedures for personality and psychological research. Retrieved
from www.scholars.northwestern.edu/en/publications/ psych-procedures-for-personality-
and-psychological-research
Revelle, W., & Zinbarg, R. E. (2009). Coefficients alpha, beta, omega, and the glb: Comments
on Sijtsma. Psychometrika, 74(1), 145-154.
Richardson, S. S., Reiches, M. W., Bruch, J., Boulicault, M., Noll, N. E., & Shattuck-Heidorn,
H. (2020). Is there a gender-equality paradox in science, technology, engineering, and
math (STEM)? Commentary on the study by Stoet and Geary (2018). Psychological
Science, 31, 338-341.
Riebler, A., Sørbye, S. H., Simpson, D., & Rue, H. (2016). An intuitive Bayesian spatial model
for disease mapping that accounts for scaling. Statistical Methods in Medical Research,
25, 1145–1165.
Rosseel, Y. (2012). Lavaan: An R package for structural equation modeling and more. version
0.5–12 (Beta). Journal of Statistical Software, 48, 1–36.
Rozin, P. (2001). Social psychology and science: Some lessons from Solomon Asch. Personality
and Social Psychology Review, 5, 2-14.

240

Rozin, P., Haidt, J., & McCauley, C. R. (2008). Disgust. In M. Lewis, J. M. Haviland-Jones, &
L. F. Barret (Eds.), Handbook of emotions (pp. 757–776). New York, NY, US: The
Guilford Press.
Rzymski, P., & Nowicki, M. (2020). Preventing covid-19 prejudice in academia. Science, 367,
1313–1313.
Sachdeva, S., Iliev, R., Ekhtiari, H., & Dehghani, M. (2015). The role of self-sacrifice in moral
dilemmas. PLoS One, 10, e0127409.
Saroglou, V., & Cohen, A. B. (2013). Cultural and cross-cultural psychology of religion. In R. F.
Paloutzian & C. L. Park (Eds.), Handbook of the psychology of religion and spirituality
(pp. 330–354). The Guilford Press.
Schäfer, M., Haun, D. B., & Tomasello, M. (2015). Fair is not fair everywhere. Psychological
Science, 26(8), 1252-1260.
Schaller, M. (2006). Parasites, behavioral defenses, and the social psychological mechanisms
through which cultures are evoked. Psychological Inquiry, 17, 96–101.
Schaller, M., & Murray, D. R. (2011). Infectious disease and the creation of culture. In M.
Gelfand, C.-Y. Chiu, & Y.-Y. Hong (Eds.), Advances in culture and psychology (Vol. 1,
pp. 99–151). New York: Oxford University Press.
Schaller, M., & Park, J. H. (2011). The behavioral immune system (and why it matters). Current
Directions in Psychological Science, 20, 99–103.
Schmid-Hempel, P. (2003). Variation in immune defence as a question of evolutionary ecology.
Proceedings of the Royal Society of London. Series B: Biological Sciences, 270, 357–366.

241

Schulz, J. F., Bahrami-Rad, D., Beauchamp, J. P., & Henrich, J (2019). The Church, intensive
kinship, and global psychological variation. Science, 366, eaau5141.
Schwartz, S. H. (1992). Universals in the content and structure of values: Theoretical advances
and empirical tests in 20 countries. In Advances in experimental social psychology (Vol.
25, pp. 1-65). Academic Press.
Schwartz, S. H., & Rubel, T. (2005). Sex differences in value priorities: cross-cultural and
multimethod studies. Journal of Personality and Social Psychology, 89(6), 1010.
Seifi-Ghozlu, S. J., Hamidi, O., Sharifi, G., & Khalili, S. (2015). Investigating of effectiveness of
life skills training on marital satisfaction. Journal of Family Counseling and
Psychotherapy, 5, 116–130. Retrieved from http://fcp.uok.ac.ir/article_11243.html
Selb, P., & Munzert, S. (2011). Estimating constituency preferences from sparse survey data
using auxiliary geographic information. Political Analysis, 19, 455–470.
Seth, A. K., Barrett, A. B., & Barnett, L. (2015). Granger causality analysis in neuroscience and
neuroimaging. Journal of Neuroscience, 35, 3293–3297.
Shackelford, T. K. (2005). An evolutionary psychological perspective on cultures of honor.
Evolutionary Psychology, 3, 147470490500300126.
Shams, A. (2016). Revolutionary religiosity and women’s access to higher education in the
Islamic Republic of Iran. Journal of Middle East Women’s Studies, 12, 126–138.
Shaw, M., Cloos, L., Luong, R., Elbaz, S., & Flake, J. K. (2020). Measurement practices in
large-scale replications: Insights from Many Labs 2. Canadian Psychology.

242

Shweder, R. A. (1990). In defense of moral realism: Reply to Gabennesch. Child Development,
61, 2060-2067.
Shweder, R. A., & Haidt, J. (1993). The future of moral psychology: Truth, intuition, and the
pluralist way. Psychological Science, 4(6), 360-365.
Shweder, R. A., Much, N. C., Mahapatra, M., & Park, L. (1997). The “big three” of morality
(autonomy, community, and divinity), and the “big three” explanations of suffering. In A.
Brandt & P. Rozin (Eds.), Morality and health (pp. 119 –169). New York, NY: Routledge.
Skitka, L. J. (2009). Exploring the “lost and found” of justice theory and research. Social Justice
Research, 22(1), 98-116.
Skitka, L. J., & Conway, P. (2019). Morality. In E. J. Finkel & R. F. Baumeister (Eds.),
Advanced social psychology: The state of the science. Oxford University Press.
Skitka, L. J., & Tetlock, P. E. (1993). Providing public assistance: Cognitive and motivational
processes underlying liberal and conservative policy preferences. Journal of Personality
and Social Psychology, 65(6), 1205.
Skurka, C., Winett, L. B., Jarman-Miller, H., & Niederdeppe, J. (2020). All things being
equal: Distinguishing proportionality and equity in moral reasoning. Social Psychological
and Personality Science, 11(3), 374-387.
Smaldino, P. E., Lukaszewski, A., von Rueden, C., & Gurven, M. (2019). Niche diversity can
explain cross-cultural differences in personality structure. Nature Human Behaviour,
3(12), 1276-1283.

243

Snowden, F. M. (2019). Epidemics and society: From the black death to the present. Yale
University Press.
Sowell, T. (2002). A conflict of visions: The ideological origins of political struggles. New York,
NY: Basic.
Srivastava, S. (2020). Personality structure: Who cares?. European Journal of Personality.
Retrieved from https://doi.org/10.31234/osf.io/dvb4n
Starmans, C., Sheskin, M., & Bloom, P. (2017). Why people prefer unequal societies. Nature
Human Behaviour, 1(4), 1–7.
Stich, S. (2006). Is morality an elegant machine or a kludge?. Journal of Cognition and Culture,
6(1-2), 181-189.
Stoet, G., & Geary, D. C. (2018). The gender-equality paradox in science, technology,
engineering, and mathematics education. Psychological Science, 29, 581-593.
Stone, W. F. (1980). The myth of left-wing authoritarianism. Political Psychology, 2(3/4), 3–19.
Tabachnick, B. G., & Fidell, L. S. (2013). Using multivariate statistics, 6th. New York, NY:
Pearson.
Tabatabai, M. H. (1996). Al-mizan fi tafsir al-Qur’an. Qom: Islamic Publications Society of
Seminary Teachers of Qom.
Talhelm, T., Haidt, J., Oishi, S., Zhang, X., Miao, F. F., & Chen, S. (2015). Liberals think more
analytically (more “WEIRD”) than conservatives. Personality and Social Psychology
Bulletin, 41, 250–267.

244

Tanaka, M. M., Kumm, J., & Feldman, M. W. (2002). Coevolution of pathogens and cultural
practices: A new look at behavioral heterogeneity in epidemics. Theoretical Population
Biology, 62, 111–119.
Tavakol, M., & Dennick, R. (2011). Making sense of Cronbach’s alpha. International Journal of
Medical Education, 2, 53–55.
Teddlie, C., & Tashakkori, A. (2009). Foundations of mixed methods research: Integrating
quantitative and qualitative approaches in the social and behavioral sciences. Thousand
Oaks, CA: Sage Publications.
Tezcur, G. M., Azadarmaki, T., & Bahar, M. (2006). Religious participation among Muslims:
Iranian exceptionalism. Critique: Critical Middle Eastern Studies, 15, 217-232.
Thalmayer, A. G., Toscanelli, C., & Arnett, J. J. (2021). The neglected 95% revisited: Is
American psychology becoming less American?. American Psychologist, 76(1), 116.
Thompson, E. R. (2007). Development and validation of an internationally reliable short-form of
the positive and negative affect schedule (PANAS). Journal of Cross-Cultural Psychology,
38, 227–242.
Thornhill, R., & Fincher, C. L. (2014). The parasite-stress theory of values and sociality:
Infectious disease, history and human values worldwide. Springer International Publishing
Switzerland.
Tizro, Z. (2013). Domestic violence in Iran: Women, marriage and Islam. London: Routledge.
Tomasello, M. (2018). The origins of morality. Scientific American, 319, 70–75.
Triandis, H. C. (1995). Individualism and collectivism. Boulder, CO: Westview.

245

Triandis, H. C., & Gelfand, M. J. (1998). Converging measurement of horizontal and vertical
individualism and collectivism. Journal of Personality and Social Psychology, 74(1), 118.
Turner-Zwinkels, F. M., Johnson, B. B., Sibley, C. G., & Brandt, M. J. (2021). Conservatives’
moral foundations are more densely connected than liberals’ moral foundations.
Personality and Social Psychology Bulletin, 47(2), 167-184.
Tybur, J. M., Inbar, Y., Aarøe, L., Barclay, P., Barlow, F. K., de Barra, M., . . . Žeželj, I. (2016).
Parasite stress and pathogen avoidance relate to distinct dimensions of political ideology
across 30 nations. Proceedings of the National Academy of Sciences, 113, 12408–12413.
Tybur, J. M., Lieberman, D., Kurzban, R., & DeScioli, P. (2013). Disgust: Evolved function and
structure. Psychological Review, 120, 65–84.
United Nations Development Programme. (2018). Human development report. Retrieved from
http://hdr.undp.org/en/2018-update
Uz, I. (2015). The index of cultural tightness and looseness among 68 countries. Journal of
Cross-Cultural Psychology, 46, 319–335.
van Borkulo, C., Epskamp, S., & Millner, A. (2016). Network comparison test: Statistical
comparison of two networks based on three invariance measures (R package version 2.0.
1)[computer software]. Retrieved from
cran.rproject.org/web/packages/NetworkComparisonTest/index.html
van der Linden, S., & Panagopoulos, C. (2019). The O'Reilly factor: An ideological bias in
judgments about sexual harassment. Personality and Individual Differences, 139, 198-201.

246

van Leeuwen, F., & Park, J. H. (2009). Perceptions of social dangers, moral foundations, and
political orientation. Personality and Individual Differences, 47, 169-173.
van Leeuwen, F., Park, J. H., Koenig, B. L., & Graham, J. (2012). Regional variation in
pathogen prevalence predicts endorsement of group-focused moral concerns. Evolution
and Human Behavior, 33, 429–437.
von Rueden, C. (2020). Making and unmaking egalitarianism in small-scale human societies.
Current Opinion in Psychology, 33, 167-171.
von Rueden, C. R., Redhead, D., O'Gorman, R., Kaplan, H., & Gurven, M. (2019). The
dynamics of men's cooperation and social status in a small-scale society. Proceedings of
the Royal Society B, 286(1908), 20191367.
von Rueden, C., Alami, S., Kaplan, H., & Gurven, M. (2018). Sex differences in political
leadership in an egalitarian society. Evolution and Human Behavior, 39(4), 402-411.
Wagemans, F., Brandt, M. J., & Zeelenberg, M. (2018). Disgust sensitivity is primarily
associated with purity-based moral judgments. Emotion, 18, 277–289.
Walster, E., Berscheid, E., & Walster, E. (1976). New directions in equity research. In L.
Berkowitz & E. Walster (Eds.), Advances in Experimental Social Psychology (Vol. 9, pp.
1–43). New York: Academic Press.
White, C. J., Kelly, J. M., Shariff, A. F., & Norenzayan, A. (2019). Supernatural norm
enforcement: Thinking about karma and God reduces selfishness among believers. Journal
of Experimental Social Psychology, 84, 103797.

247

Wilcox, R. R., & Tian, T. S. (2011). Measuring effect size: a robust heteroscedastic approach for
two or more groups. Journal of Applied Statistics, 38, 1359–1368.
Willard, A. K., Baimel, A., Turpin, H., Jong, J., & Whitehouse, H. (2020). Rewarding the good
and punishing the bad: The role of karma and afterlife beliefs in shaping moral norms.
Evolution and Human Behavior, 41
Yalçındağ, B., Özkan, T., Cesur, S., Yilmaz, O., Tepe, B., Piyale, Z. E., ... & Sunar, D. (2019).
An investigation of moral foundations theory in Turkey using different measures. Current
Psychology, 38(2), 440-457.
Yazdanbakhsh, M., Kremsner, P. G., & Van Ree, R. (2002). Allergy, parasites, and the hygiene
hypothesis. Science, 296, 490–494.
Yi, D., & Tsang, J. A. (2020). The relationship between individual differences in religion,
religious primes, and the moral foundations. Archive for the Psychology of Religion, 42,
161-193.
Yilmaz, O., & Saribay, S. A. (2017). Activating analytic thinking enhances the value given to
individualizing moral foundations. Cognition, 165, 88–96.
Yilmaz, O., Harma, M., Bahçekapili, H. G., & Cesur, S. (2016). Validation of the Moral
Foundations Questionnaire in Turkey and its relation to cultural schemas of individualism
and collectivism. Personality and Individual Differences, 99, 149–154.
Young, L., & Saxe, R. (2011). When ignorance is no excuse: Different roles for intent across
moral domains. Cognition, 120(2), 202-214.

248

Zhang, Z. (2020). The impact of social relationships on children’s distributive justice.
Developmental Psychology, 56(1), 103–116.
Zinbarg, R. E., Yovel, I., Revelle, W., & McDonald, R. P. (2006). Estimating generalizability to
a latent variable common to all of a scale's indicators: A comparison of estimators for ωh.
Applied Psychological Measurement, 30(2), 121-144.
Zou, H. (2006). The adaptive lasso and its oracle properties. Journal of the American Statistical
Association, 101, 1418–1429.

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Sex differences in moral judgements across 67 countries

PDF

Bound in hatred: a multi-methodological investigation of morally motivated acts of hate

PDF

How misinformation exploits moral values and framing: insights from social media platforms and behavioral experiments

PDF

Culture's consequences: a situated account

PDF

Can ideologically relevant threat shift group-oriented values? Relevant threatening tweets cause Progressives to be as prejudiced as Conservatives

PDF

Metacognitive experiences in judgments of truth and risk

PDF

Bridging possible identities and intervention journeys: two process-oriented accounts that advance theory and application of psychological science

PDF

Intuitions of beauty and truth: what is easy on the mind is beautiful and true

PDF

The spread of moral content in online social networks

PDF

Identity is the lens through which moral values predict action

PDF

Difficulty-as-sanctifying: when difficulties build character, purify the self, and elevate the soul

PDF

The morality of technology

PDF

“What difficulty means for me”: predictors and consequences of difficulty mindsets

PDF

Countering problematic content in digital space: bias reduction and dynamic content adaptation

PDF

Classrooms are game settings: learning through and with play

PDF

Three text-based approaches to the evolution of political values and attitudes

PDF

A roadmap for changing student roadmaps: designing interventions that use future “me” to change academic outcomes

PDF

Toward sustainable and resilient communities with HCI: physical structures and socio-cultural factors

PDF

The moral foundations of needle exchange attitudes

PDF

Expressing values and group identity through behavior and language

Asset Metadata

Creator Atari, Mohammad (author)

Core Title Socio-ecological psychology of moral values

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Psychology

Degree Conferral Date 2021-12

Publication Date 11/16/2021

Defense Date 09/29/2021

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag cultural evolution,culture,Moral Foundations Questionnaire,moral foundations theory,Morality,natural language processing,OAI-PMH Harvest,pathogen prevalence,Qeirat,scale development,Social Psychology,socioecological psychology

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Dehghani, Morteza (committee chair), Haidt, Jonathan (committee member), Lai, Mark (committee member), Oyserman, Daphna (committee member), Yazdiha, Hajar (committee member)

Creator Email atari@usc.edu,mohammad.attari@yahoo.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC17138483

Unique identifier UC17138483

Legacy Identifier etd-AtariMoham-10228

Document Type Dissertation

Format application/pdf (imt)

Rights Atari, Mohammad

Type texts

Source 20211117-wayne-usctheses-batch-897-nissen (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu