University of Southern California Dissertations and Theses

Culture's consequences: a situated account

(USC Thesis Other)

Culture's consequences: a situated account

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content `
CULTURE’S CONSEQUENCES: A SITUATED ACCOUNT
by
Ying Lin
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
(PSCHOLOGY )
August 2021
Copyright 2021 Ying Lin
`

ii

Acknowledgements
First and foremost, I would like to thank my advisor, Daphna Oyserman. The work
presented in this dissertation would not have been possible without her guidance and support.
Her tireless pursuit of high-quality research, hard work, and dedication to students have taught
me what a true researcher should be like. I am tremendously grateful for her wisdom and
continued support, for countless hours of meetings and conversations, for her quick and sharp
feedback, for helping me develop good research taste, and for steering me toward an exciting
next stage in my academic career.
I would like to thank Norbert Schwarz for his generous support, inspiring discussions,
and great sense of humor. He has shown me that research doesn’t have to be dry; it can be
simultaneously methodologically rigorous and fun. I also want to thank Morteza Dehghani, and
Sarah Townsend for serving on my committee and providing constructive feedback.
I wish to thank my lab mates and fellow graduate students, who have brightened up my
time in graduate school. Finally, I will always be indebted to my parents. Their decision to let me
spread my wings at 15 is what made me the person I am now.
`

iii

TABLE OF CONTENTS
Acknowledgements……………………………………………………………………………ii

Abstract………………………………………………………………………………………..iv

Introduction ................................................................................................................................ 1
Chapter 1: Culture Is Situated: A Meta-Analysis of Cultural Mindset Priming Effects ........... 6
Chapter 2: Seeing Meaning Even When None May Exist: Collectivism Increases Belief in Empty
Claims ...................................................................................................................................... 67
Chapter 3: Cultural Fluency Means All Is Okay, Cultural Disfluency Implies Otherwise ... 105
General Discussion ................................................................................................................ 137
References .............................................................................................................................. 138
Appendices ............................................................................................................................. 163
Appendix A: Chapter 1 Tables and Figures ....................................................................... 163
Appendix B: Chapter 2 Tables and Figures ....................................................................... 179
Appendix C: Chapter 3 Tables and Figures ....................................................................... 183

`

iv
Abstract
Much of cultural psychological research is rooted in identifying differences across
cultures. This approach has inevitably led to the perception that culture has stable influences.
Moving away from a static view of culture, this three-paper dissertation examines culture’s
consequences for how we act and what we believe as processes situated in context. In Chapter 1,
I synthesize the wide-ranging consequences of momentary cues that make a cultural mindset
accessible using a meta-analytical approach. A decade of cumulative evidence from cultural
priming research suggests that cultural mindsets – individualism, collectivism, and honor, are
situated, universal, and psychologically consequential. In Chapters 2 and 3, I identify two unique
consequences of culture for reasoning. Across seven studies in Chapter 2, I show that
collectivism, a universal cultural mindset that prioritizes fitting in and connecting with others,
motivates people to construct meaning and therefore find meaning even in places where meaning
might not exist. This meaning-making process has far-reaching implications for the consumption
of misinformation. In Chapter 3, I present eight experiments focusing on an important, yet
underappreciated aspect of culture – having cultural expertise. I show that having cultural
expertise allows people to make pragmatic inferences based on the immediate situation. Subtle
mismatches between culture-based expectations and situations disrupt people’s sense that current
patterns represent a natural order, calling into question whether social categories have stable
essences. Collectively, empirical findings across three chapters highlight a situated account of
culture’s consequences. Situational features that bring a cultural mindset to mind or interact with
culture-based knowledge powerfully influence how we act and what we believe.
`

1
Introduction
Culture has consequences. This claim has motivated a great body of cultural psychology
research in the past decades. Using cross-cultural comparisons and experimental methods,
cultural psychologists have fruitfully documented that culture influences how people structure
their self (self-concept; e.g., Markus & Kitayma, 1991; 2010), how they place value (values; e.g.,
Triandis, 1995), how they engage with others (relationality; e.g., Cohen et al., 1996), how they
experience feelings (emotion and wellbeing; Matsumoto, Yoo, & Nakagawa, 2008), and how
they process cognitive information (attribution, cognitive style; e.g., Nisbett & Norenzayan,
2002). This rich body of empirical evidence suggests that culture has profound and far-reaching
consequences on human behavior.
Three Ways to Conceptualize Culture
Thus far, cultural psychology researchers have focused on three primary ways to
conceptualize what culture is and what it does (Oyserman, 2017). First, culture can be considered
as a unique product of a group of interconnected individuals, typically marked by geographical
region, ethnicity, or race (Berry, 2003; Chiu & Hong, 2007; Kim & Berry, 1993). Based on this
view, each society has a unique set of cultural knowledge including beliefs, attitudes, norms,
practices, and behaviors that members of the society are socialized to use. Culture is the full set
of features learned within this society and is therefore rooted in a group (e.g., national, ethnic)
identity.
Instead of studying individual societies, the second way to conceptualize culture focuses
on organizing the world into basic cultural syndromes that reflect similarities and differences in
psychological processes: individualism, collectivism, and honor (Hofstede, 1983; Markus &
`

2
Kitayama, 1991; Markus & Kitayama, 2010; Triandis, 1995). A cultural syndrome consists of a
collection of elements including self-concepts, attitudes, social orientations, values, and
cognitive styles that are organized around a theme (e.g. individualism). According to this culture-
as-syndrome perspective, different socio-ecological environments chronically afford and foster
different social and psychological responses, giving rise to long-term, entrenched differences in
psychological and even neural processes across cultural groups (Kitayama et al., 2009; Markus
& Kitayama, 2010). Thus, the world can be dimensionalized into broader cultural groups (e.g.,
East vs. West, males vs. females) that vary in dominant cultural syndrome.
Individualism and collectivism constitute the primary cultural syndromes in the cultural
psychology literature (Oyserman et al., 2002; Oyserman & Lee, 2008; Triandis, 1995). In terms
of self-concept, individualism tends to focus on the independent self, with an emphasis on the
self being unique and autonomous. In contrast, collectivism tends to focus on the interdependent
self, with an emphasis on the self being embedded in groups and relationships. As a result,
individualism-collectivism is also referred to as independent-interdependent self-construal
(Markus & Kitayama, 1991). In terms of social orientations, individualism prioritizes
uniqueness, separateness from others, and control over the environment, whereas collectivism
prioritizes connectedness, fitting in, and adjustment to others. In terms of geographical regions,
Western societies (North America, Western Europe) are characterized as individualistic societies,
whereas Eastern societies (East Asia) are describes as collectivistic societies (Markus &
Kitayama, 1991; Oyserman et al., 2002; Varnum et al., 2010).
In addition to individualism and collectivism, cultural researchers have articulated a third
cultural syndrome, honor (Leung & Cohen, 2011; Nisbett & Cohen, 1996). Research on honor is
limited but emerging. Within honor, the self tends to be viewed in terms of reputation, family,
`

3
and in-groups. Reputation and respect are prioritized and should be vigorously defended.
Research on honor has focused on the distinct pattern of response to potentially honor-provoking
situations in certain societies and groups. For example, Southern American participants
displayed stronger aggression in the face of insults than non-Southern Americans (Cohen et al.,
1996). Research suggests that honor is the dominant cultural syndrome in the Middle East, Latin
America, and Southern America (Mosquera et al., 2002; Nisbett, 2018; Uskul et al., 2019).
Lastly, a culture-as-situated cognition perspective posits that culture is a human
universal. Cultural mindsets provide working solutions to basic human problems present across
societies– sustaining groups, maintaining relationships, and promoting individual welfare and
innovation (Oyserman, 2011, 2017). Given that these mindsets are evolutionarily necessary to
ensure survival, they should be available for use across societies. Hence, cultural mindsets --
individualism, collectivism, honor, should be universally available for use (Oyserman, 2015;
Oyserman & Lee, 2008). That is, irrespective of which society they are from, people have
available in memory these culturally-rooted associative knowledge networks and can use them,
depending on which one is more accessible in the immediate context.
Each of the three ways of conceptualizing culture provides a different but complementary
understanding of what culture is and how it influences psychological processes. Specifically, the
culture-as-everything perspective suggests that living in a society means having tacit cultural
knowledge that allows people to make automatic predictions in everyday situations. It allows
people to get by intuitively when situations match implicit predictions and shift to rule-based
systematic reasoning when they do not (Oyserman & Yan, 2018). The culture-as-syndrome
perspective suggests that basic cultural syndromes such as collectivism underlie major between-
society differences in social orientations, communication, and reasoning. Culture’s consequences
`

4
can be studied by comparing broad cultural groups (e.g., East vs. West). Lastly, the culture-as-
situated-cognition perspective highlights that culture’s consequences are situated. Which culture-
based knowledge affects meaning-making depends on what seems relevant and apt at the
moment, what has been recently brought to mind, and what is chronically cued (Oyserman,
2016). Cues in immediate situations can trigger a cultural mindset and shape thinking, feeling,
and doing.
Overview of Dissertation
This dissertation integrates and builds on these different conceptualizations of culture.
The overarching goal of the dissertation is to examine the unique psychological consequences
produced by different facets of culture while highlighting the situated nature of these
consequences. The dissertation is composed of three chapters. Chapter one utilized a meta-
analytical approach to provide a comprehensive overview of the consequences of making a
cultural mindset accessible in the moment. In doing so, I address outstanding theoretical
questions in the field of cultural psychology and quantify evidence for predictions provided by
each of the three ways of conceptualizing culture outlined earlier. Chapter two centers on a
consequence of one particular cultural mindset – collectivism, for human reasoning. Integrating
between-society comparisons, correlational analyses, and experimental methods, seven studies
show that collectivism motivates people to seek common ground and construct meaning in
empty claims. The findings highlight the cultural root of seeing meaning and the implications for
the acceptance of misinformation. Chapter three focuses on an underappreciated consequence of
having tacit, culture-based knowledge, which is allowing people to filter unexpected from
expected situations and guiding pragmatic inferences about the world. Specifically, the studies in
Chapter three focus on the effect of situational matches and mismatches with cultural
`

5
expectations on inherence – people’s sense that the way things are now is the natural order of
things, with downstream consequences for beliefs about fixed essences. Findings across chapters
highlight the situated nature of human reasoning and demonstrate how different aspects of
culture profoundly shape meaning-making and reasoning in the moment.

`

6
Chapter 1: Culture Is Situated: A Meta-Analysis of Cultural Mindset Priming Effects
Lin, Y., Arieli, S., & Oyserman, D. (in preparation) Culture Is Situated: A Meta-Analysis of
Cultural Mindset Priming Effects.

What are the consequences of culture? Decades of research have suggested that culture
influences almost every domain of psychological processes. Oyserman, Coon, and Kemmelmeier
(2002) provided a comprehensive review of culture's consequences for basic psychological
processes. The review shows that cultures differ in terms of how people make sense of
themselves (self-concept and self-esteem), how they valued (values), how they engage with
others (relationality), how they experience feelings (emotion and wellbeing), and how they
process cognitively (attribution, cognitive style). These consequences are reflected in between-
society differences that are small to medium in magnitude.
While cross-cultural research reveals important societal differences that are assumed to
be due to culture, this approach is unable to establish the causal role of culture or address the
underlying mechanism. To address this limitation, researchers have developed experimental
methods to manipulate the accessibility of a cultural mindset. This approach, termed cultural
priming, has the unique strengths of establishing the causal role of culture and will be the focus
of the current meta-analysis.
The first and only comprehensive meta-analysis on cultural priming was done by
Oyserman and Lee in 2008, which revealed a small-to-medium effect of priming individualism
and collectivism across 67 studies. With broader interest in understanding causal effects of
culture, the number of studies that have used cultural priming has increased exponentially since
2008. In the current research, we aim to provide a comprehensive meta-analysis with all
available experimental studies that primed a cultural mindset in the decade since 2008, while
incorporating prior studies using a Bayesian approach. We start by introducing what culture is
`

7
and how cultural priming works. We examine predictions for cultural priming effects based on
three different ways of conceptualizing culture. Next, we review core theoretical questions in the
field of cultural psychology that we seek to address with the present meta-analysis.
What Is Culture and How Can We Prime It?
Culture can be defined as the man-made part of the human environment (Triandis, 1964;
Oyserman, 2017). It is the set of meanings and practices that a group in a time and place develop
or come to adopt; these meanings and practices function to facilitate social coordination, clarify
group boundaries, protect oneself and one’s ingroups, and provide a space for innovation
(Geertz, 1984; Markus et al., 1996; Oyserman, 2011, 2017). Meanings and practices are
differentially common across societies. Based on this definition, cultural psychologists have
studied different ways that individuals are socialized within a society or social group (e.g.,
country, gender).
Importantly, culture sets up available knowledge in memory that guides thinking and
doing. Whether the knowledge is used depends on whether it comes to mind at the moment and
whether it is applicable to the task at hand. Building on this logic of concept accessibility,
researchers have developed and used experiments with priming methods to examine how culture
causally influences psychological processes.
The term “priming” describes the process by which exposure to environmental stimuli
alters subsequent responses by activating a mental construct that is applicable to the task at hand
(Bargh & Chartrand, 2000; Barsalou, 2016). The logic of priming suggests that if elements are
associated in memory, cueing one element should trigger a proximal, related element, and
through spreading activation, activate other related elements in the associative knowledge
network (Barsalou, 2016). Cultural priming has utilized this logic of priming: if self-concept,
`

8
values, goal, and cognitive procedures are all stored as part of an associative knowledge network
in memory, then activating one of these elements should activate all other elements within this
network. Such knowledge networks have been termed cultural mindsets.
In a cultural mindset priming study, participants typically receive ostensibly unrelated
tasks. The first task is the priming task (the independent variable), which entails a component of
a cultural mindset (e.g., the goal of standing out, American identity, analytical cognitive style).
The priming task activates the cultural mindset, which brings to mind content, goals, and
procedures within this associative network stored in memory through spreading activation. This
increased accessibility of the cultural mindset is expected to carry over to subsequent tasks (the
dependent variable) if the knowledge is relevant to the task.
Unlike between-society comparisons, by controlling which cultural mindset is accessible
in the moment and demonstrate its effect, researchers can use cultural priming to establish the
causal effect of a cultural mindset. By manipulating the assumed active ingredient underlying
cross-cultural differences, it enables researchers to examine the more proximal process of how
culture affects behavior. Importantly, cultural priming provides a test of whether observed
cultural differences are a result of differences in the availability (what can be activated) or in the
accessibility (what is typically activated) of cultural mindsets. Because knowledge cannot be
made accessible and used unless it is available in memory (Higgins, 1996), one can only be
primed with a cultural mindset if this cultural mindset is already available in memory. That is,
priming Americans with a collectivistic mindset can shift responses only if a collectivistic
mindset is available for use. Importantly, different ways of conceptualizing cultures can give rise
to unique predictions for cultural priming effects, which we consider in the following sections.
Three Ways of Conceptualizing Culture
`

9
Thus far, cultural psychologists have formulated three major ways to conceptualize
culture to highlight what culture is and does (Oyserman, 2017). While the three major ways are
not mutually exclusive, they emphasize different processes and functions of culture, which give
rise to unique predictions about culture priming effects.
First, culture can be considered as the unique product of a society; it consists of every
cultural knowledge including beliefs, attitudes, norms, practices, and behaviors developed and
adopted by members of a group at a specific time and place. Having this tacit cultural knowledge
allows people to make automatic predictions, living intuitively when observations match implicit
predictions and shifting to rule-based systematic reasoning when they do not (Oyserman & Yan,
2018). Second, culture can be represented as a set of dominant cultural syndromes --
individualism, collectivism, honor, and each society differentially socialized its members to each
of these cultural syndromes. Third, culture can be considered as a set of universal themes to
solve core problems created by humans needing others for survival. Each of the universal themes
(individualism, collectivism, honor) is available, though differentially accessible, across societies
and groups.
Culture as everything in a society. A culture-as-society perspective considers culture as
a unique product of a group of interconnected individuals, typically marked by geographical
region, ethnicity, or race (Berry, 2003; Chiu & Hong, 2007; Kim & Berry, 1993). Different
cultural forms evolve as adaptations to different physical and social environmental niches
(Cohen, 2001). Based on this view, each society has a unique set of cultural knowledge including
beliefs, attitudes, norms, and behaviors that members are socialized to use. Culture is the full set
of features learned within this society, not simply a cultural syndrome of individualism,
collectivism, or honor. Hence, this perspective uniquely predicts that when a cultural identity is
`

10
made salient, for example by seeing icons or symbols from this society, people act in ways that
fit with this culture (Chiu & Hong, 2007; Hong, 2009). This implies that any icon cue that
triggers the culture (the symbols, the language, considering one’s racial-ethnic identity) should
trigger the entire ensemble of knowledge associated with this society.
Based on this logic, researchers have used icon cues to activate the cultural knowledge
associated with the society among bicultural individuals. Cueing an icon from one’s society
should not only activate identification with that society (e.g., Fu et al., 2015), but also bring to
mind scripted society-specific knowledge (e.g., how time is represented spatially, Boroditsky et
al., 2011; which body image is more desirable, Guan et al., 2012) as well as the dominant
cultural mindset in that society (e.g., individualism, Mok & Morris, 2009). Thus, culture-as-
society perspective emphasizes the use of society-specific icons to study the consequences of the
main cultural mindset in the society. This approach has been limited to activating individualism
and collectivism; we did not find any study that uses icon cues to activate an honor mindset.
By conceptualizing culture as knowledge specific to a society, the culture-as-society
perspective predicts that cultural knowledge is only available to people who have been immersed
in or exposed to a society for an extended period. As a result, bi- or multi-culturalism is
necessary for people to be able to respond to cues other than the ones from their own society.
Therefore, cultural priming effects should be limited to individuals who have been socialized in
that culture or bicultural individuals who have had experience living in two cultures and possess
two sets of cultural knowledge (Chiu & Hong, 2007; Hong, 2009). At the same time, this
perspective implies that college students are a biased sample who are more "prime-able" because
they are more likely to be bicultural elites who have mastered different languages or have been
exposed to multiple cultures.
`

11
Another unique prediction from this perspective is the possibility of a contrast effect.
Because the use of cultural knowledge is deeply intertwined with the individual’s cultural
identity, how the individuals consider their bicultural identity should also matter (Benet-Martínez
et al., 2002). Bicultural individuals who experience identity concerns may react negatively to
cues from a society. For example, seeing Asian icons can make some Asian Americans feel that
they are not full-fledged Americans; this feeling leads to contrastive responses (Hong, 2009).
Thus, bicultural individuals who perceive their cultural identities as integrated and compatible
(high bicultural identity integration) should assimilate to cultural primes, whereas those who
perceive their identities as separated or in conflict (low bicultural identity integration) would
contrast to cultural primes (e.g., Mok & Morris, 2009; Ng et al., 2016).
Culture as Syndromes. Instead of focusing on studying individual societies, an
alternative way of conceptualizing culture focuses on organizing the world into basic cultural
syndromes that reflect similarities and differences in psychological processes: individualism,
collectivism, and honor (Hofstede, 1983; Markus & Kitayama, 1991; Markus & Kitayama, 2010;
Triandis, 1995). A cultural syndrome consists of a collection of elements including self-concepts,
attitudes, social orientations, values, and cognitive styles that is organized around a theme. This
perspective posits that because different socio-ecological environments chronically afford and
foster different social and psychological tendencies, there are long-term, entrenched differences
in psychological and even neural responses across cultural groups (Kitayama et al., 2009;
Markus & Kitayama, 2010). Thus, the world can be dimensionalized into broader cultural groups
(e.g., East vs. West, males vs. females) that vary in dominant cultural syndrome.
Individualism and collectivism constitute the primary cultural syndromes in the cultural
psychology literature (Oyserman et al., 2002; Oyserman & Lee, 2008; Triandis, 1995). In terms
`

12
of self-concept, individualism tends to focus on the independent self, with an emphasis on the
self being unique and autonomous. In contrast, collectivism tends to focus on the interdependent
self, with an emphasis on the self being embedded in groups and relationships. As a result,
individualism-collectivism is also referred to as independent-interdependent self-construal
(Markus & Kitayama, 1991). In terms of social orientations, individualism prioritizes
uniqueness, separateness from others, and control over the environment, whereas collectivism
prioritizes connectedness, fitting in, and adjustment to others. In terms of geographical regions,
cross-cultural comparisons have fruitfully documented numerous East-West differences along
the dimension of collectivism and individualism, with Western societies (North America,
Western Europe) giving priority to individualism and Eastern societies (East Asia) giving
priority to collectivism (Markus & Kitayama, 1991; Oyserman et al., 2002; Varnum et al., 2010).
Evidence also exists for gender differences in cultural syndromes, with women being more
interdependent (collectivistic) than men (Cross et al., 2000; Cross & Madson, 1997; Markus &
Oyserman, 1989).
In addition to individualism and collectivism, cultural researchers have articulated a third
cultural syndrome, honor (Leung & Cohen, 2011; Nisbett & Cohen, 1996). Research on honor is
limited but emerging. Within honor, the self tends to be viewed in terms of reputation, family,
and in-groups. Reputation and respect are prioritized and should be vigorously defended. This
body of research focuses on the distinct pattern of response to potentially honor-provoking
situations in certain societies and groups. For example, Southern American participants
displayed stronger aggression in the face of insults than non-Southern Americans (Cohen et al.,
1996). Research suggests that honor is the dominant cultural syndrome in the Middle East, Latin
America, and Southern America (Mosquera et al., 2002; Nisbett, 2018; Uskul et al., 2019).
`

13
Evidence of cultural syndromes can also be found in non-social domains – cognitive
styles. The social orientation hypothesis posits that differences in independent and
interdependent self-construal (individualism, collectivism) give rise to differences in cognitive
styles (Kitayama et al., 2019; Varnum et al., 2010). An individualistic focus on uniqueness and
separateness from others should orient people’s attention to the main point irrespective of the
context, whereas a collectivistic need to fit in and connect should orient people’s attention to
their surroundings and the context. Hence, social independence leads to heightened analytic
thinking, which entails paying attention to the main point, separating from context, relying on
taxonomical categorization, and focusing on internal, dispositional factors in attribution. In
contrast, social interdependence leads to heightened holistic thinking, which entails paying
attention to contextual information, relying on relationships among objects and events, and
focusing on external, situational factors in attribution (Nisbett & Norenzayan, 2002; Kühnen &
Oyserman, 2002). Empirical evidence of cognitive processing related to honor is scarce but there
is some evidence suggesting that honor is related to ordinal and hierarchical processing
(Oyserman, Novin, & Yan, unpublished data). Importantly, Nisbett et al. (2001) raised the
possibility of mutual reinforcement between cognitive style and social orientation within a
cultural syndrome, suggesting that the resulting cognitive style may feed back to strengthen the
social orientation that gives rise to it.
By aggregating different elements (e.g., self-concept, cognitive style) collectively into a
syndrome, the culture-as-syndrome perspective predicts that these elements go hand in hand;
activating each element should activate the whole cultural syndrome. Unique to this perspective
is the prediction that people can only use the cultural syndrome dominant in their culture. That is,
the effect of priming a cultural syndrome is limited to people who are socialized in this cultural
`

14
syndrome. Thus, priming collectivism may only trigger responses in ‘collectivistic’ Eastern
societies, priming individualism only trigger responses in ‘individualistic’ Western societies, and
priming honor only trigger responses in “honor cultures” (IJzerman & Cohen, 2011). To examine
this prediction, it is necessary to examine the comparison between a cultural mindset prime (e.g.,
collectivism, individualism, honor) with a control (no prime) across cultural groups with varied
dominant cultural syndromes.
Culture as Situated Cognition. A culture-as-situated cognition perspective posits that
cultural mindsets provide working solutions to basic human problems present across societies–
sustaining groups, maintaining relationships, and promoting individual welfare and innovation
(Oyserman, 2011, 2017). If these mindsets are evolutionarily necessary to ensure survival, they
should be available for use across societies. Hence, all societies are likely to have socialized their
members to each of these mindsets – individualism, collectivism, and honor. In this sense, these
cultural mindsets are universally available (Oyserman, 2015; Oyserman & Lee, 2008). That is,
irrespective of which society they are from, people have available in memory these culturally-
rooted associative knowledge networks and can use them, depending on which one is more
accessible in the immediate context.
A culture-as-situated-cognition perspective yields the following predictions. First, it
uniquely predicts that everyone is prime-able; priming each cultural mindset should produce an
effect for everyone and does not depend on region, gender, or biculturalism. Second, it posits
that societies do not differ in the availability of these cultural mindsets but in the probability that
each mindset is accessible across situations. To the extent that people living in one context
encounter cues of a particular cultural mindset more frequently than the cues of the other
mindsets, this cultural mindset becomes chronically accessible to them. Hence, it assumes that
`

15
accessibility of cultural mindset is the cause of between-societal differences. Therefore, it
predicts that a cultural mindset should produce comparable effects whether it is temporarily
(activated by a prime) or chronically accessible (between-country differences). Third, it assumes
that cultural mindsets are probabilistically cued as a result of activation of the particular cues that
are part of the mindsets. Thus, it predicts that the likelihood that cueing one element (e.g.,
cognitive style) activates another element (e.g., social orientation) within the cultural mindset
should depend on the frequency of the two elements occurring together.
Summary. As summarized in Table 1, each way of conceptualizing culture highlights
different predictions in terms of what can activate a cultural mindset and to whom it matters. In
the current meta-analysis, we aim to synthesize and assess evidence for each of the predictions.

Prior Meta-Analysis on Cultural Priming
Oyserman and Lee conducted the first meta-analysis on cultural priming studies in 2008.
This meta-analysis reveals a significant effect of priming individualism and collectivism across
67 studies and 6,240 participants. The mean weighted effect size of priming individualism and
collectivism was d = 0.34 (95% CI = [0.29, 0.39]). The core message from this meta-analysis
was that cues in the immediate situation can make individualistic and collectivistic mindsets
accessible and shape self-concept, relationality, values, and cognition. The size of effects did not
differ by region.
Oyserman and Lee’s meta-analysis marks the first step toward a comprehensive
understanding of cultural priming effects. However, due to the small number of source studies
available at the time, the conclusions drawn from their meta-analysis were necessarily tempered
by limited variability in sample characteristics, study designs, priming tasks, and outcome
`

16
measures in the source studies. Studies typically compared individualism-prime and
collectivism-prime conditions and rarely included a control condition, and hence inferences
about the relative size of the effect of priming individualistic mindset and the effect of priming
collectivistic mindset cannot be made. Study samples were largely limited to undergraduates and
included European Americans, Europeans, and Asians. Furthermore, studies that tested Asian
samples used priming tasks that differed from those used in studies testing other samples,
limiting the generalizability of results. Each of these limitations undermined the ability to make
sense of the process by which cultural mindset priming affects how people think, what they think
about, and what they do.
Given that the substantiveness of conclusions that can be drawn from a meta-analysis
depends on features of the source studies, a new look with more diverse samples, priming tasks,
and outcomes is in order. Our exhaustive review of the literature since 2008 reveals that the
number of studies that used a cultural priming method has grown by a factor of four, that
participants are no longer only undergraduates, that the priming methods used have become more
diverse, and that the number of countries represented in this literature has increased. Each of
these changes contributes to our ability to make more generalizable inferences and inform
theoretical debates. Indeed, samples are no longer limited to undergraduates; instead, an
increasing number of studies have recruited crowdsourced adult participants online. This
advance is important because these adult participants are less likely to be cultural elites than
undergraduates at research universities. The inclusion of more countries in Asia, Africa, the
Middle East, Latin America provides a broader basis for inference. The increased variability of
priming tasks provides opportunities to explore emerging areas in cultural priming literature such
as the consequences of priming society-specific icons and priming cognitive styles. Studies have
`

17
also expanded to examine a more diverse set of outcomes including creativity and consumer
attitudes. Finally, since 2008 the cultural priming literature has expanded to include priming of
another cultural mindset – honor.
Each of these changes calls for a fresher perspective on when and how cultural priming
affects outcomes. The current meta-analysis aims to provide an updated comprehensive review
of cultural priming studies published since 2008. By doing so, we also hope to assess evidence
for each of the three conceptualizations of culture as outlined above and address prominent
theoretical questions in the cultural priming literature, which we outline next.
What Activates A Cultural Mindset?
To test the effect of accessible cultural mindset on thinking and doing, researchers have
used a wide range of structured situational cues to activate a cultural mindset and assess the
corresponding shift in outcomes. To obtain a cultural priming effect, the situational cue should
activate a relevant knowledge network that contains the cue (cultural mindset), bringing to mind
the remaining components previously stored within the network, which then carry over to
influence the subsequent task. What kind of cues activate a cultural mindset remains a core
question in cultural psychology research. Cultural priming effects can only occur to the extent
that the prime activates the proposed cultural mindset. Thus, what can trigger a cultural mindset
is not only an empirical question but also a theoretical one. Can any aspect of the cultural
mindset trigger the entire ensemble of the cultural mindset and be the outcome of it? Is priming
icon, content, goal, or procedures equally effective in producing an effect?
An exhaustive review of the current literature reveals three general categories of priming
tasks: content and goal primes, icon primes, and cognitive style primes. The use of icon primes
stems from the conceptualization of culture as everything in a society. Icon primes consist of
`

18
situational cues that are specifically tied to a national, racial, or ethnic group. These include
cultural icons (e.g., Statue of Liberty, Great Wall of China), language (e.g., Spanish, English),
group identity cues (e.g., cues related to being American versus Asian American), and so on. The
prediction is that icons from a society can activate the dominant cultural mindset within that
society. It further assumes that participants need to have a significant amount of exposure to the
society or group to activate the knowledge associated with it and react to the icon primes. By this
assumption, studies that utilize icon primes have largely focused on bilingual or bicultural
individuals (e.g., ethnic minorities, sojourners, citizens of a society with a long colonial history).
Part of this focus is due to pragmatic reasons -- after all, participants can only be randomly
assigned to complete the study in English or Spanish if they speak both languages.
Content and goal primes were developed to activate the key self-concept, value, and
social orientation within each cultural mindset. Thus, content and goal primes for an
individualistic mindset activate an independent self-concept and make goals related to sticking
out and being unique salient. These include having participants consider their differences from
their family and friends, circle or write with singular first-person pronouns (e.g., "I"), read a
message that promotes individualism, and so on. In contrast, content and goal primes for a
collectivistic mindset activate an interdependent self and make goals related to fitting in and
connecting salient. These include having participants consider their similarities with family and
friends, circle or write with plural first-person pronouns (e.g., "we"), read a message that
promotes collectivism, and so on. Content and goal primes for an honor mindset focus on
bringing the concept of honor to mind; these tasks include having participants do a word
completion task with honor-related words or complete an honor value scale.
`

19
Lastly, an emerging set of studies have used cognitive style primes to manipulate holistic
versus analytic cognitive style. These include manipulating visual focus on the main point or the
context in a scene (e.g., Monga & John, 2007) and forcing categorical or relational pairings of
objects (e.g., Talhelm et al., 2015). Importantly, whether the relationship between social
orientation and cognitive styles is unidirectional remains an open question that has not been
addressed by the literature (Varnum et al., 2010). Does priming analytic versus holistic cognitive
style also lead to different ways of thinking about the self and relating with others? Studies that
used cognitive styles primes to examine social outcomes would provide an opportunity to test the
potential bidirectionality of social orientation and cognitive styles. In the current meta-analysis,
we will look for evidence regarding the bidirectionality of this relationship.
Is Everyone Prime-able?
Are people from different societies equally susceptible to the priming of each cultural
mindset, or does the extent to which a mindset can be primed depend on society or exposure to
more than one culture? As outlined in the previous sections, the three ways of considering
culture articulate different predictions regarding who can be primed with a cultural mindset. The
culture-as-situated-cognition perspective uniquely predicts that the effect of priming a cultural
mindset should be observed across societies and is not dependent on the sample being bicultural,
being college students, or being female. Addressing this question requires examination of studies
that compared primed individualism, primed collectivism, and primed honor each with a control
group to isolate the effect of priming a cultural mindset. The prior analysis was unable to address
this question due to the limited number of studies that involved a control group.
The recent expansion and the increased diversity of samples studied over the past decade
provides an invaluable opportunity to address this question using a meta-analytical approach. We
`

20
examine whether priming a cultural mindset depends on being in a region or group with that
cultural mindset as the dominant one. We also ask if biculturalism, participant sources, and
gender moderates the effect of cultural mindset priming. If cultural priming effects are
comparable across societies and regions of the world or are comparable for monocultural and
bicultural samples, for college student samples and non-college student samples, for males and
females, it would suggest that cultural mindsets are available to use for people across groups and
societies. This supports the notion that each society socializes its members to use different
cultural mindsets, rather than a particular one.
How does chronic accessibility affect cultural priming?
Another consideration of who can be primed is related to the chronic accessibility of a
cultural mindset. Cultural mindset priming functions by making a cultural mindset temporarily
accessible and hence more likely to be used in subsequent tasks. However, at the same time,
populations also vary in terms of which mindset is chronically accessible (Markus & Kitayama,
2010; Oyserman et al., 2002). How does the chronic accessibility of a cultural mindset affect
cultural priming effects?
According to a situated account of cognition, elements that have been experienced more
frequently with the prime become activated more easily (Barsalou, 2016; Förster & Liberman,
2007). The more accessible the construct is, the more likely it is to be applied. Indeed, empirical
evidence from priming research supports the additive effect of chronic and temporary sources of
accessibility on increasing the likelihood of construct use (Bargh et al., 1986). According to this
perspective, chronic accessibility of a cultural mindset should enhance the priming effect of this
mindset.
`

21
The alternative possibility points to the potential existence of a ceiling effect in the
dependent variable. For a cultural mindset priming effect to occur, the dependent variable needs
to be sensitive to the prime-induced effects. If a cultural mindset is chronically accessible, it is
plausible that goals, procedures, and content in this cultural mindset are already on the mind and
hence their responses are subject to ceiling effects; thus, priming a chronically accessible cultural
mindset may not further increase its accessibility. Though we did not find any study
documenting a ceiling effect in cultural priming, it is plausible in a statistical sense. This
possibility suggests that the effect of priming a cultural mindset would be attenuated if the
cultural mindset is chronically accessible.
These possibilities can be illustrated with the following example. A collectivistic mindset
prime entails considering connections with family and friends, which should bring to mind other
elements of collectivism such as the goal of fulfilling obligations and the procedure of
perspective-taking (cultural priming effect). The additive effect perspective suggests that this
process should happen more easily and more rapidly for those who frequently have family and
friends on the mind, i.e., people with a chronically accessible collectivistic mindset. In contrast,
the ceiling effect perspective implies that chronic accessibility of collectivism can mean that the
goal of fulfilling obligations and the procedure of perspective-taking is already constantly on the
mind; therefore, cueing connection with family and friends has little effect on shifting responses.
This potential ceiling effect in outcomes can lead to insensitivity to cultural mindset primes.
To test the potential effect of chronic accessibility on cultural priming effects, we contrast
populations that differ in chronic accessibility of individualism and collectivism, using region of
the world and gender as our proxies. We will focus on examining the pattern of results when
primed individualism and primed collectivism are each compared with a control group to reduce
`

22
the ambiguity of inference. Without any cue provided, participants in the control group are
assumed to use the chronically accessible cultural mindset in their group (e.g., individualism for
Americans, collectivism for Japanese). By examining the moderating effects of world region and
gender on the respective effects of priming individualism and collectivism, our analysis will shed
light on the role of chronic accessibility insensitivity to priming.
Current Meta-Analysis
The first and most recent meta-analysis on whether cultural mindsets can be primed was
conducted in 2008 (Oyserman & Lee, 2008), which included 6,240 participants (67 studies) and
revealed a small-to-medium effect size (d = 0.34). The core takeaway point was that small
changes in the situation (cultural mindset primes) can make individualistic and collectivistic
cultural mindsets accessible and affect psychological processes. The number and the variability
of cultural priming studies have expanded considerably in the past decade. To address the
question of the robustness of cultural priming effect across a more diverse set of studies and shed
light on unresolved theoretical debates, we conducted a comprehensive meta-analysis on the
decade of cultural priming studies since 2008.
In addition to the standard meta-analytical approach, we utilize a Bayesian meta-
analytical approach to more thoroughly quantify evidence for the true underlying effect of
cultural priming and incorporate earlier studies (Oyserman & Lee, 2008). The Bayesian approach
assumes that any one study is simply a random sample of data and random variation is likely to
cause random samples to manifest differently (Kruschke & Liddell, 2018). We use the Bayesian
approach to combine information from cultural priming studies since 2008 and information about
prior cultural priming studies from Oyserman and Lee’s (2008) meta-analysis. We also use the
Bayesian approach to incorporate existing knowledge about the field of Psychology (by basing
`

23
prior distributions on effect sizes in the field of Psychology) and address the uncertainty around
effect estimates (e.g., model selection).
The goal of the current meta-analysis is three-fold. First, we hope to examine the extent
to which cultural mindset priming effects are reliable and stable given the increased number of
cultural priming studies in the last decade and the diversification of study samples,
manipulations, and outcomes. We use a Bayesian approach to incorporate cultural priming
studies included in Oyserman and Lee's (2008) meta-analysis and existing knowledge about the
field of Psychology into our meta-analyses. Second, we take a theoretically-driven approach to
synthesize and evaluate evidence for predictions provided by the different conceptualizations of
culture, while addressing core theoretical questions regarding when and how culture influences
psychological processes. Finally, using moderator analyses, we capitalize on the increased
variability in source study samples, priming tasks, and outcomes to unpack the heterogeneity in
cultural priming effects. In doing so, we also hope to provide insights into the effectiveness of
cultural priming as a research tool, asking which cultural priming methods are likely to be
effective as well as when and for whom these priming tasks are effective.
Method
Interested readers can go to the Open Science Framework to find our (1) preregistered
PRISMA protocol, (2) complete list of included studies, (3) complete list of screened but
excluded articles with explanations for exclusion, (4) complete list of specific predictions that we
based effect size coding on for each study, and (5) the dataset and the R codes we used to
perform the analyses.
Article Identification
Literature search. We conducted an exhaustive literature search to locate studies that
`

24
experimentally manipulated a cultural mindset in three steps. First, we identified potential
empirical studies in the databases of PsycINFO, ProQuest Dissertations and Theses, and Google
Scholar. The keywords we used were: (cultur * OR interdependent OR independent OR honor
OR honour OR individualis* OR collectivis*) AND (prime OR priming OR vehicle OR agent*
OR activat* OR cue* OR language* OR frame*). We restricted our database searches to
experimental studies and replications in English published between January 1, 2008 and April 1,
2018. We chose this time frame because the prior comprehensive meta-analysis on cultural
priming was published in 2008. Second, we scanned the publication lists of cultural psychology
researchers (e.g., M. Morris, C.Y. Chiu, S. Shavitt) to identify more studies by these researchers.
Third, we scanned conference presentations in repositories of relevant conferences (SPSP, APS)
to identify unpublished studies.
Our search resulted in 17,653 records of journal articles and 8,659 records of
dissertations and theses after removing duplicate records. The lead author and two research
assistants trained by the lead author screened the titles and abstracts of these records to discard
the irrelevant ones. After the initial screening process, we identified and extracted 295 full-text
articles to assess eligibility. In Figure 1 we present our PRISMA flowchart documenting our
literature search and screening process.
Inclusion criteria. We used the following inclusion criteria for the full-text records:
1. Language: The article was in English.
2. Time period: The article was dated between 2008 and 2018.
3. Presence of cultural mindset prime: The study used an experimental manipulation (i.e.,
participants were randomly assigned to condition) to prime a cultural mindset (e.g.,
individualism, collectivism, honor). We excluded studies that manipulated bi- or multi-
`

25
culturalism (e.g., recalling experiences of living abroad) for conceptual clarity. We also
excluded studies that primed a specific aspect of cultural knowledge (e.g., American
stereotypes, Jewish religion) rather than a general cultural mindset.
4. Presence of a self-report or behavioral dependent measure: The study reported one or
more self-report or behavioral dependent measures. We excluded studies that only
measured neural responses (n = 8, e.g., event-related potential, fMRI activation) because
they were distinct and often described a disparate array of neural responses that we did
not have clear a priori predictions for. We included self-report or behavioral outcomes
from these neural studies if they were available.
5. A priori prediction: We included studies if we could use the theoretical frame
(individualism, collectivism, honor) to make an a priori theoretical prediction about the
effect of the prime on the obtained outcome, or if existing empirical evidence from cross-
national studies provided an a priori prediction. We excluded studies that only measured
knowledge and belief linked to a specific society (e.g., color preference, bodily ideal),
identification with a society or group (e.g., loyalty to China), or acculturation outcomes
(e.g., cultural adjustment).
6. Statistics: The study provided enough information to compute Cohen’s d (standardized
mean difference) for the difference between two cultural prime conditions, or between
the cultural prime and the control condition for an outcome. When information was
insufficient, we contacted the article’s author for relevant statistics and use this
information when provided.
Final sample. After evaluating each article against our inclusion criteria, we included
174 articles that contained 307 studies and 946 effect sizes (total N = 34,822). Among this
`

26
dataset, 27 studies (105 effect sizes) were from unpublished sources.
Effect Size Coding
We used findings from cross-national studies and cultural psychology theorizing
(Lehman et al., 2004; Markus & Kitayama, 1991; Oyserman, 2017; Oyserman et al., 2002;
Oyserman & Lee, 2008; Varnum et al., 2010) to make general predictions to guide our effect size
coding of studies. Because all but six studies focused on priming individualism and/or
collectivism, we list general theoretical predictions about individualism and collectivism priming
effects in Table 2. The specific prediction(s) for each study can be found in our OSF repository.
We used Cohen’s standardized d as our measure of effect size (Cohen, 1988). The first
author extracted statistics (e.g., means, standard deviations, F-values, p-values) relevant to
computing effect size then used the formula provided by Borenstein (2021) to compute the effect
size. We coded effects that supported theory-based directional theoretical predictions as positive
effects and effects that did not support theory-based directional predictions as negative effects.
For example, the theory-based prediction is that priming individualistic mindset should increase
considering the self as bounded and separate from others and decrease considering the self as
connected and related to others. Therefore both results would be coded as positive, supporting
the directional theory-based prediction.
For within-subject designs, the correlation between measures before and after
manipulation is required for computing Cohen’s d. When this statistic was not available, we used
the recommended correlation of .50 to compute Cohen’s d (Borenstein et al., 2021). When the
computation of an effect size required sample size in each condition and the author only reported
a total sample size, we divided the total sample size by the number of conditions and used that as
the condition sample size. When the computation of an effect size required the use of a p-value
`

27
and the author only provided a range (e.g., p < .05, p < .01), we used the upper bound of that
range to be conservative. Even when the authors reported no significant effect, we computed
effect sizes if the authors provided enough statistical information to do so. When authors did not
provide sufficient statistical information to compute a null effect, we assigned a conservative
estimate of d = 0.00, one-tailed p = .50, Z = 0.00 (Oyserman & Lee, 2008; Rosenthal, 1995).
Moderator Coding
Coding procedures. Prior to coding moderators, all three authors discussed, developed,
and approved a coding manual based on the included studies and Oyserman and Lee’s (2008)
prior meta-analysis on cultural priming. At the first stage, the first and the second authors
independently coded moderators of half of the studies in duplicate for comparison. Intercoder
agreement was excellent (𝜅 ranges from .80 to 1, ICC > .96). All authors discussed and resolved
the discrepancies and improved the coding manual to disambiguate the coding process. At the
second stage, the first author coded the remaining half of the studies using the updated coding
manual. Table 3 presents the moderators and the criteria we used for coding.
Overall Moderator.
Experimental Comparison. Whether the comparison is between two cultural primes or
between a cultural prime and control can be a source of variability in effect sizes. If
individualism- and collectivism-primes have opposite effects on outcomes, the size of the
difference between primed collectivism and individualism is likely to be larger than the size of
the difference between primed individualism and control, or primed collectivism and control. We
coded for the nature of the experimental comparison (individualistic- vs. collectivistic-mindset,
individualistic-mindset vs. control, collectivistic-mindset vs. control, honor vs. control).
Empirically, most cultural priming studies focus on the comparison between primed
`

28
individualistic and primed collectivistic mindset. Only a small subset included a control
condition or focused on comparing a single primed cultural mindset (individualism, collectivism,
or honor) with a control condition.
Sample-level moderator.
Region. We coded for the country or territory in which data collection took place. If data
were collected from more than one region, we computed effects for each country or territory
separately whenever data were available. We excluded effect sizes from experiments providing
only the combined effect across regions in moderator analyses. We grouped countries and
territories we found in the dataset into six world regions: English speaking countries (Canada,
U.S.A., the U.K.), Northern and Western Europe (Austria, France, Germany, Norway, the
Netherlands), Southern and Eastern Europe (Greece, Poland), Latin America (Brazil, Mexico),
Asia (China, Hong Kong, Japan, Korea, Taiwan, Singapore), Middle East-Africa (Israel,
Tunisia).
A large body of research has revealed an East-west difference in chronic accessibility of
individualism and collectivism (e.g., Oyserman et al., 2002; Vignoles et al., 2016). On the other
hand, there is mixed evidence on whether individualism, collectivism, or a mixture of both are
more chronically accessible in Latin America, Middle East-Africa, Southern and Eastern Europe
(Oyserman et al., 2002; San Martin et al., 2018; Vignoles et al., 2016). To explore the potential
impact of chronic accessibility of cultural mindset on cultural priming effect sizes, we further
categorized world regions as Western (English speaking countries, Northern and Western
Europe), Eastern (Asia), and the rest (Latin America, Southern and Eastern Europe, Middle East-
Africa).
Language. Assignment to language (e.g., English, Chinese) is sometimes used to prime
`

29
the cultural mindset associated with the society that uses that language (Chiu et al., 2007;
Oyserman & Lee, 2008). Even in studies where language is not used as a prime, it might
incidentally cue a cultural mindset when it is used. We coded for the language in which a study
was conducted. If language is not mentioned, we assumed that it was the dominant language in
the country in which the study took place.
Biculturalism. We coded for whether the study assessed biculturalism using a measure
of biculturalism. Only a small subset of studies (n = 14) assessed bicultural identity integration
and reported results separately for high and low BII groups. We computed effects for each group
separately whenever data were available.
Although we could not test the specific predictions related to bicultural identity
integration, we also coded for five other features of participants that might imply bicultural status
(Berry, 2003; Y. Hong, 2016). People who had none of these features were assumed to be mono-
cultural in our analyses. The five other features we coded to infer biculturalism were: being
members of a minority racial-ethnic group in the country in which data were collected, being
sojourners (international students, expats, or people who lived in a country other than their
country of origin at the time of testing), being members of a society in which more than one
language is both official and routinely used, being multi-lingual in a monolingual society, and
members of a society that had experienced lengthy, recent rule by an outside cultural group.
For example, the French controlled Tunisia for 75 years ending 1956, the British
controlled what was then the British Mandate of Palestine for 30 years ending in 1947 with the
creation of the state of Israel on part of the landmass of the mandate, and for 156 years, ending in
1997, Hong Kong was a British Crown Colony. Such colonial history may have socialized
citizens to another set of cultural norms, institutions, and languages, providing another cultural
`

30
frame.
We did not code participants who came from the U.S. or Canada as belonging to one of
the bicultural groups even though these are racially and ethnically heterogeneous societies.
Instead, if the study did not provide racial-ethnic breakdown of the sample, we coded it as mono-
cultural and if the study did provide a breakdown, we coded the sample as bicultural if most
participants in the sample (>60%) were racial-ethnic minorities.
Gender (Percentage of female). We wanted to code results by gender and computed
effect sizes separately for each gender whenever the study reported subgroup analyses based on
gender. However, only a few studies (n = 4) reported cultural priming effects for each gender
separately. Hence, we coded for the percentage of females within the sample and examined it as
a continuous moderator.
Methodological moderators.
Participant source. We examined participant source as a potential source of variation in
effect sizes by coding for participant source (college, crowdsourced, non-college or
crowdsourced, children) whenever studies provide such information. When participants were
drawn from multiple sources, we coded effect size separately for each source whenever data
were available. If only the combined effect across sources was reported, the experiment was not
included in this moderator analysis.
Priming task (experimental manipulation). We coded for the priming task used and
grouped the tasks into three major types: icon primes, content and goal primes, and cognitive
style primes.
Icon primes. The first type of cultural mindset priming tasks makes use of icons linked to
a specific geographical region or group (e.g., language, landmark) to activate the cultural
`

31
mindset typically associated with that region or group. Examples of these tasks include
completing a survey in English or Japanese, showing people cultural icons such as images of the
Great Wall of China or the Statue of Liberty, or asking people to recall personal experiences as
an Asian or an American to trigger ethnic or national identity. Because authors using these tasks
assumed that having been socialized to a certain culture is necessary for priming effect to occur,
most of these studies focused on bicultural samples so that the authors can match specific
participants to specific primes (e.g., showing Asian American participants cultural icons of Asia
versus America).
Content and goal primes. The second type of priming tasks were the content and goal
primes described in Oyserman and Lee (2008) and newly developed derivatives of these tasks.
This type of cultural mindset primes focuses on social psychological processes core to
individualism, collectivism, and honor. Individualistic psychological process primes focused on
content, procedures, and goals relevant to individualism (e.g., being different, separate, unique).
Examples of individualistic primes include writing about differences from family and friends,
focusing on first-person singular pronouns, and imagining working alone. Collectivistic
psychological process primes focus on the content, procedures, and goals relevant to collectivism
(e.g., being similar, connected, common). Examples of collectivistic primes include writing
about similarities with family and friends, focusing on first-person plural pronouns, and
imagining working as part of a team. Honor psychological process primes focused on making
honor values accessible (e.g., completing an honor value scale).
Cognitive style primes. The third type of cultural mindset priming task is cognitive style
primes, which provided a test of bidirectionality of effects when the assessed outcomes were
social orientations. These tasks focused on priming the cognitive styles associated with
`

32
individualistic and collectivistic mindsets. These cognitive style priming tasks focused on
activating analytic thinking versus holistic thinking (Monga & John, 2007; Talhelm et al., 2015).
Analytic cognitive style primes entail categorizing by isolating or using rule-based
categorizations of objects. Holistic cognitive style primes entail focusing on the background or
the relationships between objects.
Priming Domain. We separately coded the actual content domain of the priming task.
Levels of priming domains were constrained by actual priming tasks used. The content domain
of icons primes was society due to their roots in a specific society or group. The content domain
of cognitive style primes was cognitive style. The content domains of content and goal primes
included self, relationality, and values. Self-focused primes focused on change how participants
viewed themselves (e.g., self as separate and independent from others, self as embedded in
relationships) and an example was circling singular versus plural personal pronouns. Value-
focused primes focused on making individualistic, collectivistic, or honor value salient and an
example was reading an article promoting individualistic or collectivistic values. Relationality-
focused primes increased the salience of relationships and an example was the subliminal
presentation of affiliation-related words.
Outcomes. We took two steps to code the outcome moderator. First, we examined the
outcomes assessed and group them by type, and then we further grouped similar types of
outcomes into the broader outcome domains. Major domains focused on self-concepts, values,
relationality, and cognition. In addition to major domains, there are four more specific categories
with operationalizations constrained by the measures used in the source studies (emotionality,
personal characteristics, consumer attitudes, creativity)
Self-concepts. These measures captured thoughts about the self as self-concept. This
`

33
category includes the twenty-statement task (TST, Kuhn & McPartland, 1954) and variants of
this task, the self- vs. other-focused thought index (Lee et al., 2000), Singelis's (1994)self-
construal scale, private self-concept (e.g., self-efficacy, self-esteem), relational self-concept (e.g.,
interdependence), collective self-concept (e.g., identification with the university), and relational
and collective self-concept (e.g., self-construal scales that tap into both relationships and groups).
Values. These measures captured personal endorsement of a particular value set. This
category includes individualism-collectivism values (e.g.,, Individualism-Collectivism scale,
Triandis & Gelfand, 1998, Arab-French Value Survey, Boski & Youssef, 2012) and non-
individual-collectivism values (e.g., honor values, Ijzerman & Cohen, 2011).
Relational and Social attitudes. These measures captured how people approach, consider,
and relate to others. This category includes relationality (e.g., perspective-taking, intention to
help friends) and social judgments (e.g., victim-blaming, attitudes towards elderly).
Cognition. Outcomes in this category captured different patterns of thinking and
perception related to contextual and relational information. This category included two main
types of outcomes – attribution and cognitive procedures. Measures of attribution captured the
extent to which people attribute behavior or outcomes to the internal, dispositional attributes of
the target versus the external, situational forces. Within this category, outcome types include the
fish attribution task (a classic paradigm that involves judgment of whether a fish’s movement is a
result of internal or external forces; Morris & Peng, 1994), trait versus situational inferences, and
other attribution measures such as belief in karma.
Measures of cognitive procedures captured patterns of perceiving and processing
information. Within this category, outcome types include analytic-holistic reasoning (e.g., triad
task; Chiu, 1972), temporal connection (e.g., distance to future), and other connection-separation
`

34
procedures (e.g., Stroop task).
Emotionality. These measures captured affect/feelings, emotional strategies, and
wellbeing. This category includes measures of guilt, emotional suppression, and a wellbeing
inventory.
Personal characteristics. These measures captured personal attributes and tendencies
including personality traits (e.g., extraversion, assertiveness) and self-enhancement (e.g.,
narcissism, overestimation of self-performance).
Consumer attitudes. These measures captured preferences, behavior, or thoughts related
to consumer products (e.g., brand attitudes, purchase intention). The studies within this outcome
level often examined the effect of a match (or mismatch) between a cultural mindset and a
specific product or advertising content, with the assumption that a match should improve product
attitudes and vice versa. Hence, while these measures were all consumption-relevant, the
underlying processes by which cultural priming effects took place were heterogeneous.
Creativity. These measures focused on divergent or novel thinking.
Format. We coded how studies were administered as the previously common face-to-
face paper-and-pencil format has shifted to online, whether as part of a subject pool or with
children or adults. Face-to-face experiences may be more psychologically vivid, but given that
psychological studies are increasingly occurring online, we explored this as a possible moderator
of by study variation in effect sizes.
Study design. We coded whether the effect size was computed from a between-subject,
within-subject, or mixed-factorial design. Each design has both strengths and weaknesses. In a
between-subjects design, participants in each condition receive only one mindset prime, followed
by dependent variables. This removes the chance of carryover effects of the first mindset prime
`

35
on the understanding of the second prime but leaves open the possibility that variation may be
due to individual differences other than the randomized assignment to the priming task. In a
within-subject design, the same participants receive both mindset primes and measures of
dependent measures after each prime. This removes variation that may be due to individual
differences as distinct from the prime but leave the possibility of carryover from priming one
cultural mindset on the understanding of the second cultural mindset prime. This concern is
particularly pronounced in studies that gave participants opposing primes (e.g., individualism,
collectivism) within a short time.
Manipulation check measure. We coded whether the effect size was computed from a
measure that the authors labeled as a manipulation check of the priming task. Effect sizes of
manipulation check measures may be inflated compared to effect sizes of measures that are not
labeled manipulation check for three reasons. First, researchers tend to choose items that are
closely related to the priming task as a manipulation check to demonstrate that the priming task
indeed affected the underlying process; such construct proximity can lead to a stronger effect.
Second, researchers commonly choose a manipulation check measure that is central to the
cultural mindset (e.g., self-concept) or is an established correlate of the priming task. Third, if the
purpose of the manipulation check is to check that the priming task indeed manipulates the
assumed process, researchers can choose not to proceed (or report) if the priming task did not
produce a significant effect on manipulation check measures. Hence, there is a higher likelihood
for significant manipulation checks to be reported than non-significant ones.
Manipulation check presence and success. We coded whether the author conducted a
manipulation check of the priming task in the same study or a pilot test. We coded studies with a
manipulation check on whether the check was successful – that is, the prime significantly
`

36
affected the measured manipulation check in the predicted direction.
Number and order of outcomes. A number of studies included multiple dependent
measures. Because priming works by affecting what is on the mind and what is on the mind is
most likely to influence responses, cultural priming effects may be decay as a result of
intermediate tasks. Cultural priming effects may be the strongest for the outcome tasks that
appear immediately after the prime. We coded the number of culture-relevant outcomes
measured after the priming task in each study. For each effect size, we coded the order in which
the corresponding measure appeared. This allowed us to examine the extent to which effects
decay as increasing numbers of intermediate tasks separating the prime from the dependent
measure.
Other Moderators of Interest
Publication year. Effect sizes sometimes may decline over years. This declining trend
may occur due to large error variance associated with small sample sizes, under-specified
conditions, statistical or methodological artifacts that are more common in early studies (Protzko
& Schooler, 2017). We examined the possible declining trend of cultural priming effect by using
publication year as a continuous moderator.
Completeness of reporting. We coded the completeness of method reporting as a
potential indicator of study quality on a scale from 0 to 6 by assigning one point for information
on how the sample was recruited (source), country or language in which the study was conducted
(country or language), study format (format), mean age (age), gender make-up (female %), and
race-ethnic make-up (ethnicity %). We used the completeness score as a continuous moderator of
effect size estimates.
Publication status. We coded study publication status as unpublished if it was from a
`

37
thesis, dissertation, or conference report and as published if it was from a journal article.
Publication status is sometimes used to infer the possibility of publication bias (Francis, 2012).
The general assumption is that unpublished studies may have smaller effect sizes than published
studies separate from indicators of completeness of reporting or other indicators of quality.
Meta-analytic Approach
Overall effects. To assess the overall effect of cultural mindset priming, we used an
intercept-only random-effect meta-regression with robust variance estimation (RVE; Hedges et
al., 2010) and small sample corrections (Tipton, 2015). This approach allows us to adjust for
dependencies among effect sizes. Tanner-Smith and Tipton (2014) recommend specifying
weights for RVE based on the predominant type of dependencies among effect sizes. The effect
size dependency in our dataset primarily came from authors reporting multiple effect sizes within
a study (91%) rather than from authors reporting multiple studies using different samples (66%).
Based on this dependency, we used the correlated-effect weights throughout. We conducted all
analyses with RVE using the robumeta R package (Fisher & Tipton, 2015).
Bayesian meta-analyses. Because our current set of studies involve cultural priming
studies published over the decade between 2008 and 2018, our RVE approach does not account
for effect sizes of studies published before 2008, namely, the studies included in Oyserman and
Lee (2008). By keeping the already published meta-analysis separate, we could take advantage
of a Bayesian approach to use the average effect sizes and between-study variances reported in
Oyserman and Lee's 2008 meta-analysis as our informed prior distributions in a Bayesian meta-
analysis on the effect of priming individualism and collectivism. We used these priors in addition
to the priors based on plausible effect sizes and between-study heterogeneity in the general field
of Psychology (Gronau et al., 2017; Van Erp et al., 2017).
`

38
A Bayesian approach also allowed us to quantify evidence for our choice of a random-
effect model. A fixed-effect model assumes that there is only one true effect across studies,
whereas a random-effect model assumes that effect sizes could vary from study to study
(Borenstein, 2009). The choice of fixed- vs. random-effect model is non-trivial and we used
Bayesian model-averaging as a way to test whether an effect exists while accounting for the
uncertainty concerning the choice of model. Bayesian model-averaging does not impose the
assumption of a fixed- or a random-effect model, but instead, weights both models according to
their posterior possibilities in the observed data to generate a model-averaged estimate of effect
sizes (Gronau et al., 2017; Scheibehenne et al., 2017).
We used the R package, metaBMA (Heck & Gronau, 2017) to conduct Bayesian meta-
analyses. We used aggregated effect sizes in these analyses. To do so, we aggregated the
dependent effect sizes within a study using the aggregation method using a pre-registered
correlation (default r = .50) between the dependent effect sizes (Borenstein et al., 2009).
We performed all Bayesian analyses using a random-effect model unless stated
otherwise. The random-effects model has a hierarchical structure (Gronau et al., 2017). The
effect size (d) of the ith study is generated from a normal distribution with the true effect size (𝜃 i)
and within-study variance (SE
2
). The study-specific effect size is generated from a normal
distribution with a population-average effect size (𝛿 random) and between-study variance (𝜏 2
). This
hierarchical structure is depicted as:
di ~ N(𝜃 I, SE
2
)
𝜃 I ~N(𝛿 random, 𝜏 2
)
The first analysis focused on estimating an overall cultural priming effect while
incorporating existing research in the field of psychology. Therefore, for the estimate of
`

39
standardized effect size, we used default prior in Psychology – a Cauchy distribution with a scale
parameter equal to 1/√2 (Gronau et al., 2017; Morey et al., 2015). For the estimate of between-
study variability, we use an inverse gamma distribution with shape parameter equal to 1 and
scale parameter equal to 0.15, based on the distribution of between-study standard deviations for
mean-difference effect sizes in meta-analyses published in Psychological Bulletin from 1990 to
2013 (Gronau et al., 2017; Van Erp et al., 2017). This prior choice is depicted as:
𝛿 ~ Cauchy(0,
1
√2
)
𝜏 ~ Inverse-Gamma(1, 0.15)
The second analysis focused on estimating the effect of priming individualism and
collectivism while incorporating information about effect size and between-study heterogeneity
from similar cultural priming studies before 2008. To do so, we used informed prior distributions
based on Oyserman and Lee’s meta-analysis of priming individualism and collectivism.
Oyserman and Lee reported a mean weighted effect size of d = 0.34, SE = 0.02, and between-
study variability SD = 0.34. Based on this information, for the estimate of standardized effect
size, we used a normal distribution with mean equal to 0.34 and standard deviation equal to 0.02.
For the estimate of between-study variance, we used a half-t distribution centered around 0.34
with scale equal to 0.1 and degree of freedom equal to 1, based on Gelman's (2006)
recommendations for variance parameters. This informed prior choice is depicted as:
𝛿 ~ N(0.34, .02)
𝜏 ~ t(0.34, 0.1)
After each analysis, as a model check, we used Bayesian model-averaging to combine the
results of both fixed-effects and random-effects models and obtain a model-average estimate of
effect size. We compared the model-average estimates to the estimates obtained in the random-
`

40
effect models. Similar effect size estimates generated using Bayesian model-averaging and a
random-effect model would suggest that a random-effect is appropriate for the current dataset.
Moderator analyses. We used RVE with correlated-effect weights and small-sample
corrections to test the effect of each moderator (Tanner-Smith et al., 2016). For our moderator
analyses, we excluded moderator levels with effect sizes fewer than five because small-sample
correction is problematic under such circumstances (Tipton, 2015).
Most moderators were categorical: region, participant source, language, biculturalism,
BII group, priming task, outcome type, study design, format, manipulation check measure,
manipulation check presence, manipulation check success. For categorical moderators, we used
random-effects meta-regression to estimate the effect size and confidence intervals for each
subgroup. When there were two levels within a moderator, significant moderation was indicated
by the significance of the regression coefficient of the variable. When there were more than two
levels within a moderator, we performed a small-sample adjusted F-test to test the presence of
differences among its levels using the Wald test function from the R package clubSandwich
(Pustejovsky, 2016; Tipton, 2015).
Some moderators -- the percentage of females, year, outcome number, outcome order,
and completeness of method reporting were continuous variables. In these cases, we also used
random-effects meta-regression to estimate the effect of the moderator. We centered the year
variable by subtracting 2,007 from the year to allow ease of interpretation of coefficient (i.e.,
Year 1 = 2008).
Examination of publication bias.
Tests of publication bias. To examine potential publication bias operationalized as a
higher likelihood of significant effects to be found in published rather than unpublished studies,
`

41
we examined the funnel plot distribution and performed three statistical tests of publication bias:
Egger's regression intercept test (Sterne & Egger, 2005), Duval and Tweedie’s Trim and Fill
method (Duval & Tweedie, 2000), and PET-PEESE models (Stanley & Doucouliagos, 2014).
We used aggregated effect sizes to create a funnel plot to visually inspect the possibility
of publication bias. The funnel plot is a scatter plot of effect size estimates against the standard
error of these estimates. Publication bias is indicated by a rightward asymmetry in a funnel plot.
We then used Egger’s regression method to quantify the bias captured by the funnel plot.
Egger’s test regresses the standardized effect on the inverse of standard error. Small studies
generally have a large standard error. In the absence of bias, we would expect to see such studies
associated with smaller standardized effects. This would create a regression line whose intercept
approaches zero. Next, we used the Trim and Fill method to remove observations from the right
side of the funnel plot and imputes observations to even out the distribution.
We used RVE to perform the PET-PEESE method. Precision-Effect Test (PET) and
Precision-Effect Estimate with Standard Error (PEESE) models compute the relationship
between effect size and standard error (Stanley & Doucouliagos, 2014). While PET fits a linear
regression line to this relationship, PEESE fits a quadratic line. PET-PEESE accounts for the
limitations of both models by taking account of the statistical significance of the PET estimate to
decide whether the final estimate is based on PET or PEESE.
Sensitivity analyses. In addition to tests of publication bias, we performed additional
sensitivity analyses to look for evidence that selection bias could yield an overestimation of
effect size. First, we performed moderator analysis to test whether published studies, on average,
report larger effect sizes than unpublished studies. Our pool of unpublished studies included
dissertations, thesis, and conference posters, which might later be revised for publication or will
`

42
not be published due to problems in design or execution or because studies do not yield
significant results consistent with predictions. These issues may mean that published studies
overestimate the overall effect size. Second, we reran our main meta-analysis of the cultural-
priming effect excluding outcomes labeled as manipulation checks. Manipulation checks may be
more likely to be reported when they are significant and in the predicted direction, hence it is
likely that effect sizes from manipulation checks are inflated. Including manipulation checks as
outcomes may lead to an overestimation of effect size.
Results of Overall Effect
The overall effect of priming cultural mindset(s) across all studies is d = 0.41, 95% CI
[0.38, 0.45], t(306) = 22.4, p < .001 (See Table 5). This indicates that the overall cultural priming
effect is small-to-medium in size in the predicted theoretical direction.
Bayesian Meta-analyses
Bayesian meta-analyses on overall cultural priming effect. We conducted a Bayesian
meta-analysis to estimate the overall cultural priming effect in a random-effect model. Four
chains were used to approximate the posterior distributions of the parameters based on the
weakly informed priors from the literature of Psychology. The 𝑅 ̂
statistics were smaller than 1.1
for both estimates, suggesting good mixing. Effective sample sizes were 7,728 for effect size
estimation and 4,068 for tau estimation, suggesting that parameters were estimated with enough
precision. Figure 2 presents the weakly informed prior distribution of effect size and the
estimated posterior distribution of effect size based on the weakly informed priors. Based on the
model, priming a cultural mindset had an average effect of d = 0.40, 95% HPD [0.36, 0.44], SD
= 0.02, in the predicted theoretical direction. The estimate of between-study heterogeneity
(standard deviation of study effect sizes) was 0.24, 95% HPD [0.20, 0.27].
`

43
We used Bayesian model-averaging to combine fixed- and random-effect models. The
model-averaging approach yielded the same effect size estimate as the random-effect model, d =
0.40, 95% HPD [0.36, 0.44], SD = 0.02, which was distinct from the fixed-model result, d =
0.35, 95% HPD [0.33, 0.37], SD = 0.01. This suggests that a random-effect model is a more
accurate model for the current data.
Bayesian meta-analyses of priming individualism and collectivism. We conducted a
Bayesian meta-analysis to estimate the effect size of priming individualism versus collectivism
in a random-effect model. Four chains were used to approximate the posterior distributions of the
parameters using informed prior distributions based on effect size and between-study
heterogeneity in Oyserman and Lee’s (2008) meta-analysis. The 𝑅 ̂
statistics were smaller than
1.1 for both estimates, suggesting good mixing. Effective sample sizes were 6,604 for effect size
estimation and 3,770 for tau estimation, suggesting that parameters were estimated with enough
precision. Figure 2 presents the informed prior distribution of effect size (Oyserman & Lee,
2008) and the estimated posterior distribution of effect size based on the informed priors. Based
on the model, the average difference between priming individualism and collectivism was d =
0.37, 95% HPD [0.35, 0.40], SD = 0.01, in the predicted theoretical direction. The estimate of
between-study heterogeneity (standard deviation of study effect sizes) was 0.24, 95% HPD [0.21,
0.28].
We used Bayesian model-averaging to combine fixed- and random-effect models. Again,
the model-averaging approach yielded the same effect size as the random-effect model, d = 0.37,
95% HPD [0.35, 0.40], SD = 0.01, which was different from the effect size given by the fixed-
model, d = 0.35, 95% HPD [0.33, 0.37], SD = 0.01. This suggests that a random-effect model is
favored for the current data of priming individualism and collectivism.
`

44
Results of Moderator Analyses
There was substantial heterogeneity among cultural priming effect sizes (I
2
= 69.85%, 𝜏 2

= 0.09). Table 4 presents an overview of source study characteristics. We followed up with
planned moderator analyses to unpack the source of such variability.
Moderator Analyses Addressing Theoretical Questions
Experimental Comparison. To address if priming each cultural mindset has an effect,
we tested whether experimental comparison (4 levels: individualistic mindset vs. collectivistic
mindset, individualistic mindset vs. control, collectivistic mindset vs. control, and honor vs.
control) moderates the effect of priming a cultural mindset. Results suggested that there was no
significant difference in the magnitude of cultural priming effects across types of experimental
comparison, F(3, 18.6) = 2.97, p = 0.06. The difference between primed individualistic and
primed collectivistic (d = 0.44, 95% CI [0.40, 0.48], p < .001) mindset was similar in size to the
difference between primed collectivistic mindset and control (d = 0.35, 95% CI [0.27, 0.44], p
< .001) and the difference between primed honor and control (d = 0.29, 95% CI [0.08, 0.50], p
= .02), but was larger than the size of the difference between primed individualistic mindset and
control (d = 0.21, 95% CI [0.10, 0.34], p <.001). These results suggest that on average, each of
the cultural mindsets is prime-able.
Region. To perform region-level moderator analyses, we started with region as the
moderator (7 levels: English speaking countries, Northern and Western Europe, Southern and
Eastern Europe, Asia, Latin American, the Middle East and Africa). The Middle East (Israel) and
Africa (Tunisia) only contained one country but we included them for completeness. We
excluded effect sizes from studies in which participants came from more than one region (k =
12). World region did not have a significant effect on the magnitude of cultural priming effects,
`

45
F(5, 5.32) = 0.78, p = .59. Effect were similar in size in studies conducted in English-speaking
countries (d = 0.36, 95% CI = [0.32, 0.41], p <.001), in Northern and Western Europe (d = 0.40,
95% CI = [0.28, 0.53], p = <.001), in Southern and Eastern Europe (d = 0.54, 95% CI = [-0.01,
1.10], p = .05), in Asia (d = 0.41, 95% CI = [0.34, 0.48], p < .001), in Latin America (d = 0.23,
95% CI = [-2.73, 3.18], p = 0.51), in Middle East and Africa (d = 0.56, 95% CI = [0.37, 0.74], p
<.001).
Next, we tested the more specific country and territory as the moderator. We excluded
effect sizes from studies in which participants came from more than one country or territory (k =
12) or from a place with fewer than 5 effect sizes (Austria k = 4; Brazil k = 4; Japan k = 3;
Mexico k = 1; Norway k = 2; Poland, k = 3) or with effect sizes from only one study (Tunisia, k =
14, s = 1), leaving 13 countries and territories in this moderator analysis. Country and territory
had a non-significant main effect on cultural priming effect size, F(12, 13.5) = 2.43, p = .06.
Figure 4 presents effect sizes and 95% CI by country and territory.
To explore the potential role of chronic accessibility of a cultural mindset on the effect
size of priming the cultural mindset, we compared the effect sizes from studies conducted in the
West, the East, and the rest of the world. We limited our test to effect sizes from 1) the
comparisons between priming individualism and control (k = 80) and 2) the comparisons
priming collectivism and control (k = 100). This allowed us to address the question of whether
priming individualism (or collectivism) has a larger or smaller effect in a country where
individualism (collectivism) is chronically accessible. The number of studies that primed honor
was too small for subgroup analyses. Our analyses did not reveal significant differences between
studies in the West, the East, and the rest of the world in effect sizes of priming individualism
(F(2, 4.63) = 1.02, p = .43) or priming collectivism (F(2, 1.72) = 0.19, p = .84) (Figure 5).
`

46
Priming individualism produced similar effects in studies conducted in the East (d = 0.36, 95%
CI = [0.01, 0.72], p = .05) than in studies conducted in the West (d = 0.12, 95% CI = [0.01,
0.24], p = .04) and in the rest of the world (d = 0.16, 95% CI = [-0.21, 0.53], p = .20), these
differences were not significant (ps > .10) as the 95% confidence intervals overlapped with each
other. Priming collectivism achieved similar magnitudes of effects in the East (d = 0.40, 95% CI
= [0.16, 0.64], p = .01), the West (d = 0.33, 95% CI = [0.24, 0.42], p < .001), and the rest of the
world (d = 0.32, 95% CI = [-0.92, 1.55], p = .19). Therefore, we did not find any evidence that a
mindset is more prime-able in one region than another based on which cultural mindset is
potentially more chronically accessible in that region.
Biculturalism. We excluded effect sizes from participants from samples that included
participants with different levels of biculturalism (k = 14) to allow simplicity of inference. First,
we tested bicultural group as moderator. The monocultural group was set as the reference level in
the moderator analysis. There was a significant effect of bicultural group, F(4, 9.2) = 3.83, p
= .04. However, effect size was not smaller for monocultural participants. Effect sizes were
similar among monocultural participants (d = 0.42, 95% CI = [0.38, 0.47], p < .001), racial-
ethnic minorities (d = 0.37, 95% CI = [0.20, 0.54], p = .02), participants from societies with a
colonial history (d = 0.43, 95% CI = [0.34, 0.54], p < .001), and bilinguals in monocultural
societies (d = 0.27, 95% CI = [-0.73, 1.28], p = 0.35). The only bicultural group with an effect
size significantly smaller than the monocultural participants was sojourners (d = 0.11, 95% CI =
[-0.11, 0.32], p = .26). Follow-up analysis suggests that this low effect size among sojourners
may be driven by the low BII groups. Effect size for sojourners (d = 0.31, 95% CI = [-0.12,
0.74], p = .12) was comparable with other groups when low BII participants (k = 3) were
excluded.
`

47
We also examined the effects of biculturalism separately for the comparison between
primed individualism and control and the comparison between primed collectivism and control.
The number of studies that primed honor was too small to warrant any subgroup analysis. Effects
of priming individualism (F(2, 10) = 0.27, p = .77) or priming collectivism (F(2, 6.91) = 0.30, p
= .75) did not vary as a function of biculturalism. Priming individualism yielded similar effect
sizes for monocultural participants (d = 0.21, 95% CI = [0.04, 0.39], p = .02), participants from
societies with a colonial history (d = 0.32, 95% CI = [-0.09, 0.73], p = 0.10), and participants
who were racial-ethnic minorities (d = 0.17, 95% CI = [-0.03, 0.37], p = .08). Priming
collectivism also yielded comparable effect sizes for monocultural participants (d = 0.36, 95% CI
= [0.27, 0.47], p < .001), participants from societies with a colonial history (d = 0.28, 95% CI =
[-0.06, 0.62], p = .08), and participants who were racial-ethnic minorities (d = 0.36, 95% CI = [-
0.06, 0.66], p = .03). This result suggests that monocultural people are equally affected by
cultural mindset priming as the various bicultural groups.
Bicultural identity integration. Next, we tested bicultural identity integration (BII: high
vs. low) as a moderator. Among the 14 studies that assessed BII and reported results separately
for high and low BI groups, BII was a significant moderator of cultural mindset priming effects,
B = 1.01, t(10.5) = 5.94, p < .001. Cultural priming yielded a positive effect among high BII
samples (d = 0.46, 95% CI = [0.27, 0.64], p <.001) but a negative effect among low BII samples
(d = -0.53, 95% CI = [-0.77, -0.29], p <.001). This result suggests that people with low BII
demonstrated contrast effects to cultural priming.
Participant source. We excluded effect sizes from studies with multiple participant
sources (k = 22) and studies with children (k = 4) because there were fewer than five effect sizes
for children. We did not find a significant effect of participant source on the magnitude of
`

48
cultural priming effects, F(2, 13.8) = 0.42, p = .67. Effect sizes were similar among college
student samples (d = 0.41, 95% CI = [0.37, 0.45], p < .001), crowdsourced adult samples (d =
0.36, 95% CI = [0.20, 0.52], p <.001), and non-crowdsourced, non-college-student adult samples
(d = 0.31, 95% CI = [-0.03, 0.66], p = .07). This implies that adults, whether students, or crowd-
sourced or other adults are equally sensitive to cultural mindset priming.
Percentage of females. We examined the effect of gender by testing whether the
percentage of females in a sample is related to the magnitude of the cultural priming effect.
Percentage females in a study did not have a significant effect on effect sizes overall, B = -.10,
t(24.8) = -0.60, p = .55. We also tested if the percentage of females mattered when only
considering comparisons between primed individualism and a control group or comparisons
between primed collectivism and a control group. Percentage of females did not moderate effect
sizes when comparing primed individualism with control, B = -.47, t(4.21) = -1.62, p = .18, or
when comparing primed collectivism with control, B = .13, t(7.11) = 0.54, p = .61.
Priming method. We excluded effect sizes from studies that used more than one priming
task (k = 78) or used a priming task with less than five effect sizes (gestures k = 4; subliminal
primes k = 4). First, we ran a moderator analysis to test if there are differences among effect sizes
across the three major types of priming tasks (icon, content and goal, cognitive style). Results
suggested that there was no difference, F(2, 57.7) = 1.21, p = .31. Icon primes (d = 0.36, 95% CI
= [0.26, 0.47], p <.001), content and goal primes (d = 0.42, 95% CI = [0.38, 0.47], p <.001), and
cognitive style primes (d = 0.45, 95% CI = [0.35, 0.54], p <.001) yielded highly similar effect
sizes.
We also tested if there are differences among effect sizes across specific priming tasks.
Effect sizes varied across priming tasks, F(16, 24.7) = 3.1, p = .006. For interested readers, Table
`

49
6 lists the average effect size by specific priming task.
We also tested if the effects of using each major type of priming task depended on
biculturalism. Biculturalism did not moderate the effect of icon primes, F(3, 12.2) = 2.03, p =
0.16, content and goal primes, F(3, 3.19) = 1.77, p = 0.32, or cognitive style primes, t(3.66) = -
0.39, p = -.72. Monocultural individuals were equally affected by each type of primes as
bicultural individuals.
Type of outcome. First, we examined whether the more general outcome domain was a
moderator of cultural priming effect sizes. Effect sizes did not vary across broader outcome
domains, F(7, 33.5) = 1.60, p = .17. Effects were medium in magnitude when outcomes were
categorized as self-concept (d = 0.56, 95% CI = [0.47, 0.65], p <.001). Effects were small-to-
medium in magnitude when outcomes were categorized as emotionality (d = 0.45, 95% CI =
[0.26, 0.65], p <.001), social attitudes (d = 0.39, 95% CI = [0.30, 0.47], p <.001), cognition (d =
0.39, 95% CI = [0.32, 0.46], p <.001), consumer attitudes (d = 0.39, 95% CI = [0.30, 0.49], p
<.001), values (d = 0.38, 95% CI = [0.26, 0.50], p <.001), personal characteristics (d = 0.29, 95%
CI = [0.11, 0.46], p = .003), and creativity (d = 0.22, 95% CI = [-0.74, 1.19], p = .42).
Next, we examined whether there was a significant difference in effect sizes across more
specific types of outcomes. Specific outcome type was a significant moderator of cultural
priming effect sizes, F(22, 64.5) = 2.48, p = .003. Cultural priming had the largest effect size
when outcome was self-other thought index (d = 0.94, 95% CI = [0.57, 1.31], p < .001). Table 6
lists the average effect size within each specific outcome type.
Because certain outcome types were mostly employed as manipulation checks rather than
main dependent variables and manipulation checks had larger effect sizes, we examined the
possibility that the effect of outcome type was driven by being manipulation checks by rerunning
`

50
the outcome type moderator analysis controlling for whether the outcome was a manipulation
check. The effect of outcome type reduced but was still significant after controlling for being a
manipulation check, F(22, 64.2) = 1.75, p = .04.
Prime and Outcome Domain. Test of bidirectionality requires simultaneous
consideration of prime and outcome domains. We examined how the effect sizes of priming each
domain (society, self, value, relationality, cognitive style) vary across different outcome
domains. For this analysis, we excluded levels of outcome domains with fewer than 5 effect sizes
within each prime domain, except for social and interpersonal attitudes (k = 4) using cognitive
style primes because including them allowed for the important test of bidirectionality.
Table 7 presents the results. Studies have used society-focused primes to test outcomes
categorized as self-concept, personal characteristics, values, social and personal attitudes, and
cognition, and its effect size did not depend on outcome domain, F(4, 21.5) = 1.80, p = .17.
Studies have used self-focused primes to test outcomes categorized as self-concept, personal
characteristics, social and interpersonal attitudes, values, consumer attitudes, emotionality, and
cognition. Effect sizes of using self-focused primes did not vary across outcome domains, F(6,
40) = 1.76, p = .13. Studies have used relationality-focused primes to test outcomes categorized
as self-concept, values, and social and interpersonal attitudes and its effect size did not depend
on outcome domain, F(2, 13.3) = 0.11, p = .90. Studies have used value-focused prime to test
outcomes categorized as self-concept, value, social and interpersonal attitudes, and consumer
attitudes, and its effect size did not depend on outcome domains, F(3, 1.32) = 0.49, p = .74.
Lastly, studies that used cognitive style primes have focused on cognition, consumer
attitudes, self-concept, and social and interpersonal attitudes. Importantly, the effect of cognitive
style primes on self-concept and social and interpersonal attitudes provides a test of bi-
`

51
directionality of cultural priming effects from non-social to social domains. Only seven studies
manipulated cognitive style to examine social outcomes. Among these studies, one assessed self-
concept (Arieli & Sagiv, 2018), three assessed sense of personal control (Zhou et al., 2012), and
three assessed social liberalism (Talhelm, 2018; Talhelm et al., 2015).
Because of the small number of effect sizes in self-concept (k = 5) and social and
interpersonal attitudes (k = 4), we combined these two domains into “social orientation” to
provide a more stable estimate of the effect of cognitive style primes on social outcomes. Table 7
presents effect sizes when these categories are considered separately. Effect size of cognitive
style primes did not vary as a function of outcome domain, F(2, 7.86)= 3.52, p = .08.
Importantly, cognitive style primes had a significant effect on influencing social orientations (d =
0.41, 95% CI = [0.19, 0.64], p = .005), providing initial evidence for a causal relationship from
non-social to social domains in cultural priming.
Moderator Analyses of Pragmatic Interest
Language. We excluded effect sizes from studies that used language to prime cultural
mindsets (k = 66), used more than one language (k = 7) or used a language with fewer than 5
effect sizes (Greek k = 3; Japanese k = 3; Norwegian k = 2; Portuguese k = 4; Spanish k = 1).
Study language did not have a significant main effect on cultural priming effect size, F(6, 12.3) =
2.09, p = .13. Studies conducted in German (d = 0.57, 95% CI = [0.31, 0.84], p = .003), Hebrew
(d = 0.53, 95% CI = [0.35, 0.71], p < .001) and French (d = 0.49, 95% CI = [-0.15, 1.13], p
= .09) yielded medium average effect sizes. Studies conducted in English (d = 0.40, 95% CI =
[0.36, 0.45], p < .001), Chinese (d = 0.40, 95% CI = [0.31, 0.50], p < .001), and Korean (d =
0.37, 95% CI = [-0.05, 0.79], p = .06) yielded small-to-medium effect sizes, whereas studies
conducted in Dutch yielded a small effect size (d = 0.15, 95% CI = [-0.04, 0.34], p = .10).
`

52
To further explore the possibility that language is an incidental cue of a cultural mindset
that influence cultural priming effects, we grouped languages into collectivism-triggering
(Japanese, Chinese, Korean) and individualism-triggering (English, German, French, Hebrew,
Norwegian, Dutch) languages. We did not find any effect size difference in studies using
collectivism-triggering languages (d = 0.41, 95% CI = [0.34, 0.49], p < .001) and studies using
individualism-triggering language (d = 0.41, 95% CI = [0.36, 0.45], p < .001), B = -.03, t(45.2) =
-0.69, p = .55.
Format. We tested whether effect sizes differed based on study format. Prior to the
moderator analysis, we excluded effect sizes from studies that used more than one format (k =
14) or used a format with fewer than 5 effect sizes (in lab neural scan k = 3; in lab interaction k =
4). Our results suggest that the difference in effect sizes based on study format was not
significant, F(4, 6.48) = 3.68, p = .07. Cultural priming effects were significant across different
formats. Specifically, effect sizes were small-to-medium when studies were delivered with
paper-and-pencil in lab (d = 0.37, 95% CI = [0.25, 0.49], p <.001), computerized in lab (d = 0.34,
95% CI = [0.28, 0.40], p <. 001), online (d = 0.34, 95% CI = [0.26, 0.41], p <.001), or offline
and outside of the lab (d = 0.36, 95% CI = [0.16, 0.56], p = .002). Effects were larger in size but
not significant when studies were delivered as face-to-face interview in lab (d = 0.78, 95% CI =
[-0.36, 1.91], p = .07).
Study design. Study design was not a significant moderator of cultural priming effects,
F(2, 4.29) = 0.20, p = 0.83. Effects obtained with between-subject (d = 0.42, 95% CI = [0.38,
0.45], p < .001), within-subject (d = 0.39, 95% CI = [0.16, 0.61], p = .002), and mixed factorial
designs (d = 0.37, 95% CI = [-0.19, 0.93], p = .10) were similar in size.
Whether effect sizes are from manipulation checks. Fifteen percent of effect sizes (k
`

53
= 73) in this meta-analysis were from manipulation checks of cultural priming effects. We tested
the possible effect size inflation for manipulation checks by comparing effect sizes for outcomes
described as a manipulation check versus as a dependent variable. Indeed, effect sizes were
larger for outcomes labeled as manipulation checks (d = 0.63, 95% CI = [0.52, 0.74], p < .001)
than for non-manipulation-check outcomes (d = 0.38, 95% CI = [0.34, 0.42], p < .001). This
difference was significant, B = 0.24, t(80.3) = 4.13, p < .001, 95% CI = [0.12, 0.36].
Whether the study contains a (successful) manipulation check. In this analysis, we
focused on effect sizes from non-manipulation-check study outcomes (k = 804). Thirty-one
percent of these effect sizes (k = 251) were from studies that reported a manipulation check.
Among this subset of effect sizes, only 3.6 % of them (k = 9) were from studies where
manipulation checks failed.
We examined whether effect sizes differ based on the presence and the success of
manipulation checks. Results show that effect sizes did not differ between studies that reported a
manipulation check (d = 0.42, 95% CI = [0.35, 0.49], p < .001) and studies that did not report
one (d = 0.36, 95% CI = [0.32, 0.41], p < .001), B = 0.05, t(151) = 1.31, p = .19, 95% CI = [-
0.03, 0.14]. Although cultural priming effects were medium in size and significant when the
manipulation checks succeeded (d = 0.43, 95% CI = [0.36, 0.50, p < .001) and were small and
non-significant when the manipulation checks failed (d = 0.18, 95% CI = [-0.20, 0.56], p = .17),
this difference did not reach significance, B = 0.24, t(2.11) = 2.49, p = .12, 95% CI = [-0.16,
0.63]. The implication is two-fold, failure to have a manipulation check does not undermine the
effect size and having a successful manipulation check does not inflate it.
Number and order of outcomes. To test the potential decay of the priming effect, we
tested whether effect sizes differ as a function of the number of outcomes and the order in which
`

54
the outcome appeared after the manipulation. Result provided no evidence that the magnitude of
cultural priming effects was influenced by the number of outcomes in the study, B = .01, t(37.7)
= 0.44, p = .66, 95% CI = [-0.04, 0.06], or the order it was measured, B = -.01, t(22.1) = -0.20, p
= 0.84, 95% CI = -0.08, 0.06]. At the same time, interpretation of these results requires caution
since it is possible that authors were not always fully clear about whether there had been other
variables measured but not reported in the analyses.
Moderator Analyses of Other Potential Sources of Heterogeneity
Publication year. Our moderator analysis finds no evidence that the magnitude of
cultural priming effect sizes change from 2008 to 2018, B = -0.01, t(157) = -1.75, p = .08, 95%
CI = [-0.02, 0.001].
Completeness of method reporting. We used completeness of method reporting as a
potential indicator of study quality and tested if effect sizes vary based on completeness of
method reporting. Results suggest that more complete method reporting was associated with
smaller effect sizes, B = -0.06, t(116.9) = -4.12, p < .001, 95% CI = [-0.09, -0.03]. We followed
up by considering that authors typically do not report details of sample and methods for pilot
tests of manipulations, hence this result may be an artifact of larger effect sizes from
manipulation checks. Indeed, follow-up analyses showed that when we excluded manipulation
checks, completeness of method report did not influence the magnitude of cultural priming
effects, B = -0.03, t(113.7) = -1.70, p = .09, 95% CI = [-0.06, 0.005].

Moderator Analysis Summary
Out of the moderators we examined, bicultural identity integration, specific priming task,
specific outcome type, and whether effect sizes were manipulation checks significantly
`

55
moderated the magnitude of cultural priming effects. People with integrated bicultural identities
assimilate to cultural mindset primes, whereas those with conflicted bicultural identities contrast
to them. As shown in Table 6, the magnitude of cultural priming effects may be larger or smaller
depending on the specific priming task employed or the specific type of outcomes examined.
Effect sizes were larger when they were from measures labeled as manipulation checks. In
addition, study location (whether examined as country and territory, world region, or East versus
West), biculturalism, percentage of female, and participant source did not moderate cultural
priming effect, supporting the idea that everyone is prime-able.
Results of Publication Bias Analyses
We followed our preregistered plan to examine potential publication bias in three steps.
First, we visually inspected the funnel plot distribution of aggregated effect sizes, used Egger’s
regression method (Sterne & Egger, 2005) to quantify the bias captured by the funnel plot, and
used the Trim-and-Fill method (Duval & Tweedie, 2000) to even out distribution. Second, we
used the RVE approach to perform PET-PEESE models (Stanley & Doucouliagos, 2014). Third,
we examined whether publication status moderates the magnitude of cultural priming effects and
conducted additional sensitivity analyses.
Funnel Plot
We used aggregated effect sizes to create a funnel plot. To do so, we aggregated the
dependent effect sizes within a study using a pre-registered correlation (default r = .50) between
the dependent effect sizes (Borenstein et al., 2009). A rightward asymmetry in a funnel plot
suggests that studies with low precision, negative or non-significant effects are missing, which is
interpreting as publication bias (Simmonds, 2015). As we show in Figure 4 Panel A, we found
some degree of asymmetry in the distribution. Egger’s regression test suggests that the
`

56
asymmetry was statistically significant, bias = 1.35, 95% CI = [0.87, 1.82], z = 5.57, p < .001.
Given the asymmetry, we used Duval and Tweedies's (2000) trim-and-fill method to remove
observations that caused the asymmetry and impute missing observations to achieve symmetry.
As we show in Figure 3 Panel B, the trim-and-fill method imputed 61 additional studies on the
left side, yielding an adjusted effect size estimate of d = 0.30, 95% CI = [0.26, 0.34], Z = 13.99, p
< .001.
PET-PEESE
We used the RVE approach to conduct the Precision-Effect Test (PET) and Precision-
Effect Estimate with Standard ERROR (PEESE) models which compute the relationship
between effect size and standard error. The PET model was significant, B = 1.08, t(95.5) = 4.47,
p < .001, suggesting that PEESE model should be used. The PEESE model was also significant,
B = 1.79, t(36.8) = 3.22, p =.003, suggesting the presence of publication bias. However, even
after correcting for publication bias, the estimate of overall cultural priming effects remained
significant, d = 0.30, t(82.3) = 8.65, p < .001, 95% CI = [0.23, 0.37].
Sensitivity Analyses
First, to address the possibility that significant results are more likely to be published, we
performed moderator analysis to test whether published studies, on average, yield larger effect
sizes than unpublished studies. We found evidence that published studies (d = .43, 95% CI =
[0.39, 0.46], p < .001) provided larger effect size than unpublished studies (d = 0.28, 95% CI =
[0.12, 0.44], p = .001), B = 0.17, t(29.7) = 2.19, p = .04, 95% CI = [0.01, 0.32]. However, the
overall cultural priming effect was smaller though still significant in unpublished studies.
Second, our results show that manipulation check outcomes yielded larger effect sizes than non-
manipulation-check outcomes, hence we tested the overall cultural priming effects excluding all
`

57
effect sizes coming from manipulation checks. The overall effect excluding manipulation check
measures was smaller in magnitude but highly significant (d = 0.38, 95% CI = [0.34, 0.42], p
< .001).
Summary
We found evidence for the existence of publication bias across analyses. Hence, the true
magnitude of cultural priming effects may be smaller than the original estimate (d = 0.40). At the
same time, the overall effect of cultural mindset priming remained significant when accounting
for publication bias corrections and in sensitivity analyses that excluded potentially biased
studies. Correcting for publication bias using the trim-and-fill method and PET-PEESE both
yielded an adjusted overall effect size of d = 0.30.
Discussion
Does priming a cultural mindset affect thinking, feeling, and doing? Our meta-analysis of
946 effect sizes from 307 cultural priming studies since 2008 suggests that priming a cultural
mindset (individualism, collectivism, or honor) can indeed produce effects on self-concepts,
personal characteristics, values, social attitudes, cognitive styles, consumer attitudes, and
emotionality. These effects are positive and significant, small-to-medium in size, and
heterogeneous. Importantly, priming every single cultural mindset – individualism, collectivism,
honor produces significant effects, suggesting that each of these mindsets may be universally
available for priming.
Importantly, the presence of cultural priming effects is patterned and in line with the
cross-cultural literature. The magnitude of cultural priming effects is similar to the size of
differences observed in between-society comparisons (Oyserman et al., 2002). In other words,
temporarily making a cultural mindset accessible produces comparable effects on judgment and
`

58
behavior as seen in cross-cultural differences. This finding supports the causal role of cultural
mindsets in inducing cross-cultural differences in self-concept, values, social judgment,
emotions, and cognitive procedures. It implies that between-society differences that are
attributed to culture should not be used to infer that culture is fixed. Instead, observed differences
across societies and groups can be explained by the differences in chronic accessibility of
cultural mindsets (what is typically on the mind), rather than the availability of cultural mindsets
(what knowledge exists in memory and can be primed).
The confidence intervals of overall effect size of cultural priming studies from the current
meta-analysis (d = 0.41, 95% CI = [0.38, 0.45]) overlap with the estimated effect size of priming
individualism and collectivism from the previous meta-analysis (d = 0.34, 95% CI = [0.29, 0.39]
by Oyserman and Lee (2008), though the confidence intervals overlap. Oyserman and Lee’s
meta-analysis has focused on the comparison between individualism and collectivism.
Integrating effect sizes of individualism-collectivism comparisons from the two meta-analyses
using a Bayesian approach yielded an average effect size of d = 0.37, 95% HPD [0.35, 0.40],
supporting the robustness of cultural priming effects over time. Importantly, though we found
some evidence for publication bias in cultural priming research based on various publication bias
analyses, adjusting for publication bias yielded a smaller but still significant estimate of cultural
priming effect (d = 0.30). This pattern of results suggests that though the average size estimate of
the cultural priming effect may be inflated, the existence of the cultural priming effect does not
appear to be driven by publication bias.
Our finding of a significant and stable cultural priming effect is of substantial
significance. Despite the ample use of priming methods in social psychological research, the
recent replication failures of priming effects have sparked a debate over the existence of priming
`

59
effects (e.g., Kahneman, 2012; Pashler et al., 2012; Shanks et al., 2013; Sherman & Rivers,
2021). The meta-analytical approach has been advocated as a useful tool to assess the robustness
and credibility of psychological effects (e.g., Fabrigar & Wegener, 2016). In addition to the
standard meta-analytical approach, we also utilized Bayesian meta-analyses to more thoroughly
quantify evidence for the true underlying effect of cultural priming and address the uncertainty
around effect estimates (Kruschke & Liddell, 2018). Specifically, we used the Bayesian
approach to incorporate existing knowledge about the field of psychology and effect size of
cultural priming studies before 2008 and used model-averaging to address uncertainty around
model selection. Synthesizing studies using the standard and the Bayesian approaches both
revealed substantial evidence for the effects of cultural priming on psychological processes,
lending confidence to the use of cultural priming methods to study culturally induce differences.
Addressing Theoretical Questions
Cultural mindsets can be triggered by content and goals, icons, and cognitive styles.
An important theoretical question in the cultural priming literature is what can activate a cultural
mindset. We found comparable effects of triggering content and goal, cultural icons, and
cognitive style on psychological processes. Content and goal primes focused on making a
particular self-concept (e.g., self as embedded in groups and relationships), goal (e.g., fitting in
with others), or value (e.g., collectivistic cultural values) accessible in the moment. Activating
content and goal within a cultural mindset shifts judgment and behavior in the social (e.g.,
interpersonal and social attitudes) and non-social (e.g., cognition) domains. Effects were the
strongest for self-concept and weakest for cognition. This result suggests that self-concept may
be the most affected outcome by cultural priming.
`

60
Consistent with the social-as-society perspective, results show that icon cues from a
society can activate the dominant cultural mindset in the society, producing consequences for
self-concept, interpersonal and social attitudes, values, and cognitive styles. That is, seeing icons
from America, speaking English, or considering one's identity as American increases the
likelihood that people act in an individualistic manner. This result suggests that a dominant
cultural mindset is indeed part of people's representation of a society and can be activated by
considering the society.
Studies that used cognitive style primes provided an opportunity to test the
bidirectionality of cultural priming effects. The social orientation hypothesis states that the
independent and interdependent ways of socializing within individualism and collectivism each
give rise to analytic and holistic cognitive styles (Kitayama et al., 2019; Varnum et al., 2010).
But research, to date, has not systematically addressed whether the priming effect is
bidirectional, such that activating a cognitive style can also affect how people process social
information (Varnum et al., 2010). In the current meta-analysis, we found support for
bidirectionality of cultural priming effects. Activating a cognitive style causally influences social
outcomes, providing support to the claim that social and cognitive tendencies mutually reinforce
each other in shaping cultural patterns (Nisbett et al., 2001). However, it is worth pointing out
that the majority of studies that manipulated cognitive styles examined cognitive outcomes or
cognitive-style-related consumer experiences. Only seven studies assessed the impact of
manipulating cognitive styles on social outcomes. Due to the heterogeneity in outcomes, we
cannot ascertain that this effect is stable. Nevertheless, the possibility of bidirectionality suggests
a promising direction for future research.
`

61
People across societies can be primed with cultural mindsets. A central question in the
cultural priming literature is: who can be primed? This question also addresses the theoretical
debate of whether each society or group socializes its members to use only one dominant cultural
mindset or different ones.
Different ways of considering culture yield divergent predictions. The culture-as-society
and culture-as-syndrome perspectives predict that a society socializes its members in one cultural
mindset and only individuals socialized in that cultural mindset can be primed with it. In
contrast, the culture-as-situated-cognition perspective suggests that individualism, collectivism,
and honor are universally available cultural mindsets and thus everyone can be primed. The
current meta-analysis assessed the evidence for each prediction, using region, gender, and
biculturalism as proxies for the likelihood that participants in each study had been socialized in a
particular cultural mindset.
We did not find evidence that being part of a particular cultural group is necessary for the
cultural priming effect to occur. The magnitude of cultural priming effects did not vary as a
function of world region, sample source, or percentage of females. Compared to a control group,
priming individualism shifted judgment and behavior whether the studies were conducted in the
West (North America, Northern and Western Europe) or in the East (East Asia) Priming
collectivism also shifted judgment and behavior whether the studies were conducted in the West
or the East. Because only six studies primed honor, we are unable to conduct subgroup analyses
for priming honor. Thus, we were unable to address whether effect sizes of priming honor vary
across regions and if honor is universally available to be primed. However, these small set of
studies (Novin & Oyserman, 2016; Shafa et al., 2015) demonstrated that priming honor shifted
responses in Netherland and the non-Southern U.S. – societies that are not typically
`

62
characterized as "honor cultures", lending support to the claim that honor mindset is available
outside "honor cultures" such as the Middle East and Latin America.
We also tested whether being exposed to more than one culture is a necessary condition
for cultural priming by organizing study samples into monocultural and four bicultural groups --
racial-ethnic minorities, sojourners, participants from a society with a colonial history, and
bilingual participants from a monocultural society. While effect sizes varied across bicultural
groups, we found comparable effects of priming a cultural mindset when studies used
monocultural participants and when studies used bicultural participants. Effects were also
positive and comparable for monocultural and bicultural groups when each type of primes
(content and goal, icon, cognitive style) was considered separately. Notably, icon primes were
able to activate cultural mindsets among monocultural and biculturals to the same extent. This
finding runs counter to the culture-as-society perspective that only individuals who have had a
chronic experience with a society can react to icon primes (Hong, 2009). Taken together, these
patterns of results suggest that people across societies are equally affected by cultural mindset
priming, lending support to the culture-as-situated-cognition theory. We interpret these findings
as suggesting that each of the cultural mindsets – individualism, collectivism, and honor, is
available across societies and groups and when cued, carries consequences for judgment and
behavior.
Notably, we did not find evidence of an additive or a ceiling effect due to the chronic
accessibility of a cultural mindset within a population. The effect of priming individualism or
collectivism did not vary as a function of world region (East vs. West) or the percentage of
females in a study. Contrary to other priming research (e.g., Bargh et al., 1986), our results did
not support the notion that temporary and chronic sources of accessibility can be combined in an
`

63
additive manner. This finding suggests that the temporary accessibility of a cultural mindset
increases the likelihood of the mindset being used regardless of whether it is chronically
accessible.
The culture-as-society perspective uniquely predicts that cultural primes would trigger
assimilative effects when cultural identities are congruent and contrastive effects when cultural
identities conflict. Results support this prediction. In the fourteen studies that reported separate
results for high and low BII groups, High BII individuals showed positive cultural priming
effects, whereas low BII individuals showed negative effects -- the opposite of what would be
predicted of the cultural mindset. One caveat is that all but one of these studies used icon primes
that were linked to a specific group or society. Hence, these results can be interpreted as
suggesting that cueing a society or group makes the associated cultural mindset accessible but
whether people use it or reject it depends on how they perceive their group identity. Thus, it is
unclear whether BII would moderate the effect of primes that elicit content, goal, and cognitive
style more generally (e.g., circling “I” vs. “we”).
Limitation and Future Direction
An important advance in cultural priming literature is the inclusion of a third cultural
mindset – honor. Research of honor has focused on comparing societies that are characterized as
honor cultures or not, suggesting that cultural psychologists have largely considered honor as a
culture-specific syndrome that is available to only certain populations (Nisbett & Cohen, 2018;
for a review see Uskul et al., 2019). The current meta-analysis found only six studies that used
experimental manipulation to test honor’s consequences. The small number of honor priming
studies rendered it impossible to conduct moderator analyses. Hence, we were unable to address
whether honor can be made accessible across populations or only to certain populations.
`

64
However, evidence reviewed in the meta-analysis showed that honor can be primed in societies
that are not characterized as honor cultures, suggesting that it is likely to be a universally
available mindset. However, a test of universality would require a larger pool of studies testing
the effect of priming honor in “honor” and “non-honor” societies.
In addition, studies that primed honor have focused on making a global concept of honor
accessible. In terms of priming tasks, two studies used honor-related words (IJzerman & Cohen,
2011) and four studies asked participants to fill out honor value scales before outcome measures
(Novin & Oyserman, 2016; Shafa et al., 2015). Thus, whether cueing only a component of honor
mindset (e.g., reputation concerns, willingness to retaliate) can also make honor mindset
accessible and produce similar effects remains an open question. In addition, honor priming
studies have assessed outcomes related to sensitivity to honor stimuli, use of potency visual cues,
regulatory strategy, and aggression, but have not examined its consequences for self-concept or
cognitive styles. These gaps in knowledge provide promising avenues for future research.
We did not find any evidence that cultural priming effects weaken as a function of
outcome order. However, the current set of cultural priming studies do not allow us to address
whether cultural priming effects last over obstacles and a longer time, because almost all studies
were conducted within a few minutes in one setting without temporal delay or filler task between
cultural primes and dependent variables. The broader literature on priming suggests that the
persistence or the decay of priming effects depends on what is being primed, with social
perception priming effects decrease, and goal priming effects persist or escalate, with a temporal
delay (Higgins, 1996; Weingarten et al., 2015). The current meta-analyses only speak to the
short-term consequences of cultural priming, with consequences over a longer time remaining
unknown.
`

65
Our findings are also limited by the lack of samples across different developmental
stages. Only two studies in the current dataset involved children and none of the studies focused
on older adults. Therefore, we are unable to address whether cultural priming effects occur
across the life span. The presence of cultural priming effects among children would suggest that
these cultural mindsets are learned early in life and maybe fundamental to social functioning.
Some evidence suggests that culture has less influence on cognitive processing style among
elderly adults (for a review, Park et al., 1999), thus it is plausible that cultural priming effects
would be weaker among elderly adults. Data from more diverse age groups are needed to
understand the role of individual differences in cultural priming effects.
We used available study indicators of theoretical moderators to conduct moderator
analyses on cultural priming effects. Our choice of operationalizations is constrained to a large
extent by what is reported in source studies. For example, we were unable to conduct subgroup
analyses based on gender because only four studies reported separate results for each gender. We
used region and percentage of females in a study to operationalize chronic accessibility of a
cultural mindset, though these operationalizations have limitations. The precision of these
measures can be low, given the within-region and within-gender variations in chronic
accessibility of a cultural mindset (e.g., Oyserman et al., 2002). Increased precision of
operationalizations would allow for more confidence in inferences about theoretical moderators.
Our categorization of the East and the West is also limited by the fact that studies have
disproportionately focused on participants from a few countries and territories (i.e., the U.S.,
Canada, mainland China, Hong Kong), making generalizability to regions that are less
industrialized or outside the preconceived realm of individualism and collectivism difficult. As
large-scale international collaborations in cultural psychological research become more common
`

66
(Vignoles et al., 2016), future research is encouraged to test cultural priming in more diverse
societies.
Conclusion
Culture has consequences. This claim has motivated decades of cultural psychological
research. The cumulative evidence thus far suggests that priming a cultural mindset –
individualism, collectivism, and honor, does indeed influence psychological processes. Making a
cultural mindset accessible causally influences how people make sense of themselves, engage
with others, make social judgments, experience feelings, and process information. The increase
in number and the diversification of study samples, manipulations, and outcomes over the past
decade facilitate a more nuanced discussion of how and when cultural priming effects emerge.
Our results suggest that between-society differences that are attributed to culture should not be
used to infer that culture is fixed. Instead, our results strongly suggest that people across groups
and societies can use different cultural mindsets when they are brought to mind. These results
can be achieved by using a wide range of cultural priming tasks, including by priming cognitive
style. It is our hope that the present review would shed some light on the outstanding theoretical
question in the cultural psychology literature and indicate useful directions for future research.

`

67
Chapter 2: Seeing Meaning Even When None May Exist: Collectivism Increases Belief in
Empty Claims
Lin, Y., Zhang. Y.C. & Oyserman, D. (2021.) Seeing Meaning Even When None May Exist:
Collectivism Increases Belief in Empty Claim. Journal of Personality and Social Psychology.

Abstract
People often find truth and meaning in claims that have no regard for truth or empirical evidence.
We propose that one reason is that people value connecting and fitting in with others, motivating
them to seek the common ground of communication and generate explanations for how claims
might make sense. This increases the likelihood that people experience empty claims as truthful,
meaningful, or even profound. Seven studies (N > 16,000 from the U.S. and China) support our
prediction. People who score higher in collectivism (valuing connection and fitting in) are more
likely to find fake news meaningful and believe in pseudoscience (Studies 1 to 3). China-U.S.
cross-national comparisons show parallel effects. Relative to people from the U.S., Chinese
participants are more likely to see meaning in randomly generated vague claims (Study 4).
People higher in collectivism are more likely to engage in meaning-making, generating
explanations when faced with an empty claim, and having done so, are more likely to find
meaning (Study 5). People who momentarily experience themselves as more collectivistic are
more likely to see empty claims as meaningful (Study 6). People higher in collectivism are more
likely to engage in meaning-making unless there is no common ground to seek (Study 7). We
interpret our results as suggesting that conditions that trigger collectivism create fertile territory
for the spread of empty claims, including fake news and misinformation.
Keywords
Collectivism, Culture, Social Cognition, Communication, Misinformation

`

68
Four in ten Americans (42%) find astrology sort of or very scientific (General Social
Survey, 2018). One in three (33%) believe in reincarnation (Pew, 2018). People act on these
beliefs --as reflected in the $2.2 billion annual U.S. market for psychic services and astrology-
related goods (IBIS World, 2020). In these and other ways, Americans find meaning in empty
claims, claims produced with little or no concern for either truth or empirical evidence (Frankfurt
& Wilson, 2005; Risen, 2016). Claims can be empty (lack truth value) in two ways: They can be
unverifiable and irrefutable (e.g., believing in reincarnation), or mis- or dis-informative (e.g.,
believing fake news). People seem as likely to believe mis- or dis-informative claims as other
kinds of empty claims. For example, almost four in ten Americans believe that Joe Biden did not
legitimately win the 2020 U.S. Presidential election (36%, YouGov/Economist poll, February
2021). An equal proportion (39%) acted on risk-increasing COVID-19 misinformation, doing
things like gargling cleaning products (Centers for Disease Control, Gharpure, 2020). In this
paper, we predict that collectivism, the aspect of human culture that sensitizes people to connect
with others, may explain why people are vulnerable to such empty claims.
We build our prediction on evolutionary models of human culture. Human culture
structures human interactions (Heinrich, 2020; Oyserman, Kemmelmeier, & Coon, 2002; von
Hippel, von Hippel, & Suddendorf), is complex, tool-intensive, and cumulative (Mesoudi &
Thornton, 2018; Osiurak & Reynaud, 2020). A part of the evolutionary puzzle is that human
societies accumulate skills and knowledge that allow for technological development based on
trust (Mesoudi & Thornton, 2018; see also, Heinrich, 2020). People do not redevelop or
rediscover cultural knowledge at each generation. Instead, they acquire knowledge from others.
They assume that transmitted information is meaningful and proceed from there (Mesoudi &
Thornton, 2018; Osiurak & Reynaud, 2020). This tendency to accept in-group knowledge allows
`

69
cumulative culture and increasing cultural complexity. However, it also means that culturally
acquired ideas may not be optimal and can even be maladaptive (Mesoudi & Thornton, 2018).
Whereas acceptance may be passive, in the current paper, we suggest that people do not just
accept. They often actively make sense and meaning by generating rationales of their own for
acquired ideas.
Seeing Meaning in Empty Claims
Humans are meaning-makers, creating, both in their minds and in their social
interactions, a sense of what things are about and what they mean (Park & George, 2018) from
imperfect cues and culture-based structural formulations (Baumeister & Landau, 2018).
Meaning-making is culture-based in the sense that arbitrary symbols, like colors, are meaning-
signifiers only because of culture thus, red “means” stop, prosperity and good luck, or is the
color of love depending on culture (Oyserman, 2011). The idea that people actively construct
meaning rather than passively receive it and do so using culture-based tools is central to situated
accounts of human reasoning (Oyserman, 2015). Within a cultural context, people know what
things are likely to mean and this allows people to make predictions without investing higher-
level reasoning (Oyserman, Novin, Flinkenflögel, & Krabbendam, 2014). This constructive
process can lead people astray. After constructing meaning, people may come to believe things
are true when they are not because they have filled in the blanks for a claim that was itself
unverifiable and irrefutable, or mis- or dis-informative. While the literature provides two useful
accounts of how people might be led astray (an underthinking account and a motivated reasoning
account), as we detail next, neither account addresses the vulnerability built into the collectivistic
aspect of human culture.
`

70
Underthinking. Underthinking is the often-nonconscious reliance on quick, intuitive
thinking uncorrected by systematic reasoning (Kahneman & Frederick, 2005; Pennycook et al.,
2015; Risen, 2016). According to the underthinking account, deliberation reduces, and lack of
deliberation promotes, belief in empty claims. Supporting an underthinking account of seeing
meaning in empty claims, people who are low in cognitive ability have a higher tendency to
believe empty claims (e.g., superstition, fake news, Murphy, Loftus. Grady, Levine & Greene,
2019). The same is true for people who tend to think intuitively rather than analytically
(Pennycook et al., 2015). People are also more likely to believe empty claims when they are in
situations that trigger intuitive reasoning, such as when they have limited time (Bago et al., 2020)
or are in positive moods (Greifeneder, Jaffe, Newman, & Schwarz, 2020).
Motivated reasoning. Motivated reasoning is the often-nonconscious reliance on
reasoning approaches that support identities and pre-existing beliefs (Kahan, 2012). In the
context of empty claims, a motivated reasoning account implies that people are prone to see
meaning if the claims feel compatible with their identity or existing beliefs. Supporting a
motivated reasoning account of seeing meaning in empty claims, people are more receptive to
bogus personality feedback that implies that they have positive rather than negative traits
(Johnson et al. 1985). They are more receptive to fake news that supports rather than opposes
their political attitudes (Farago, Kende, Kreko, 2019; Murphy et al., 2019).
Gaps. Though valuable, neither of these current accounts fully explains the phenomena.
People are not fully protected from misinformation when they use strategies that engage slow
and deliberate reasoning (Majima, 2015; Kahan, 2012). They are not fully protected when they
are motivated to be accurate (Pennycook et al., 2020) or no longer need to self-affirm (Munro &
Stansbury, 2009).
`

71
Collectivism: A Focus on the Common Ground
Our cultural account can help fill the gaps in current explanations by highlighting an
additional route by which people may come to accept empty claims as meaningful and true.
Culture entails societal-level processes with implications for individual-level and society-level
outcomes (Oyserman & Uskul, 2008). Indeed, culture-as-situated-cognition (CSC) theory
proposes that culture affects multiple levels (universal, societal, situational, individual,
Oyserman, 2015, 2017). At the highest level, culture is a human universal. Societies create
culture as a 'good enough' solution to universal needs. For instance, humans cannot survive
alone; they need an entity (in-group) to sustain them (Baumeister & Leary, 1995; Von Hippel,
Von Hippel, & Suddendorf, in press). This universal need implies that there must be some
universal mechanism triggering people to band together to cooperate and share with others.
Cultural psychologists label this mechanism collectivism.
At the societal level, culture is a specific meaning-making framework, a mindset that
influences what is attended to, which goal or mental procedure is salient. Collectivistic mindsets
focus people's attention on goals and content relevant to fitting in and belonging, the mental
procedures relevant to connecting, and actions that facilitate attending to others (Oyserman,
2017). At the situational level, which aspect of culture-based knowledge is on the mind and
accessible for meaning-making depends on what seems relevant and apt at the moment, what has
recently been brought to mind, and what is chronically most relevant (Oyserman, 2016).
Situations that trigger a collectivistic mindset can be proximal and immediate or societal and
historical. What matters is that they require interdependence (for example, ecologies with high
pathogen risk, Fincher et al., 2008) or ethnocentrism (for example, hostile contexts requiring
group defense, Von Hippel et al., in press). In these societies and situations, insiders are trusted
`

72
more than outsiders, who are viewed with suspicion (Romano, Balliet, Yamagishi, & Liu, 2017;
Yamagishi, 2017). Comparing behaviors across societies reveals differences in chronic
propensity to focus on connecting, supporting the notion that people in some societies encounter
more situations that call for a collectivistic mindset than people in other societies (e.g., China
versus the U.S, Nisbett, 2004; Oyserman et al., 2002). Experimental manipulations of the
accessibility of a collectivistic mindset across societies document that situational cues can easily
trigger collectivism (Oyserman & Lee, 2008).
One way to facilitate a goal of fitting in and belonging is to use an indirect
communication style that reduces the chances of directly creating or confronting disagreements
by relying on contextual cues and inferences to carry meaning (Hall; 1976; Gudykunst et al.,
1996). Communication in societies that prioritize interdependence and relationships is more
likely to rely on context and receiver interpretations than on what the communicator says (Hall,
1976). Messages that are context- and interpretation-driven are intentionally less directive and
more ambiguous, putting the onus on the receiver to read between the lines and fill in the blanks
(Singelis & Brown, 1995). When communication is indirect, interpretation in context generates
meaning. Meaning does not exist separate from context-based interpretation. Receivers can only
figure out a communicator's intended meaning if they are attuned to the communicator’s
perspective (Haberstroh et al., 2002). The implication is that collectivism is associated with
increased sensitivity to what others are trying to say and a focus on making sense when the
message is ambiguous (Gudykunst et al., 1996).
Indeed, experimental evidence shows that when a collectivistic mindset is triggered,
people are more sensitive to other people’s perspectives (Haberstroh et al., 2002) and perform
better in judgment tasks that require perspective-taking (Wolgast & Oyserman, 2019). Similarly,
`

73
between-country comparisons suggest that, on average, people from collectivistic societies attend
more to indirect cues, such as vocal tone (Ishii, Reyes, & Kitayama, 2003) and communicator
intent (Haberstroh et al., 2002). They decipher ambiguous messages (Sanchez-Burks et al., 2003)
and other people's mental states (Wu & Keysar, 2007) better. Together, these results imply that
collectivism increases peoples’ sensitivity to the communicative intent of others. In the current
paper, we document that this sensitivity can be a double-edged sword. If the claims
communicators present are non-probative, recipients may still construct meaning.
Current Studies
We propose that collectivism increases people’s sense that they are responsible for
inferring what a communicator is trying to say. To do so, people process claims as if they were
asking implicitly, “How might this claim make sense?” This focus on making sense motivates
people to interpret, fill in the blanks, and construct meaning for empty claims. People are more
likely to experience claims as truthful, meaningful, even profound once they have filled in the
blanks that allow them to construct meaning.
We derived three hypotheses (H1, H2, H3) from our proposal, which we tested across
seven studies. H1: People who are higher in collectivism are more likely to believe empty claims
(operationalized as pseudoscience in Studies 1, 2; fake news in Studies 3a-c, 5b; and randomly
generated statements in Studies 4, 5a, 6). As correlates of H1, we expect that because belief is
based on self-generated reasons, it has downstream consequences, increasing people’s false
belief that they saw newly generated fake news before (Study 3a) and their willingness to share
fake news (Study 3b). In H2 and H3 we predict the process by which this occurs: people who are
higher in collectivism see meaning in empty claims because they generate meaning (construct
explanations of how the claims might be meaningful or truthful) in seeking common ground with
`

74
the communicator. SpecificallyHence, H2: Meaning construction should mediate the effect of
collectivism on seeing meaning (Study 5), and H3: The absence of a human communicator
should moderate the collectivism-seeing meaning relationship (Study 7).
Any test is a test of an operationalization, not a direct assessment of the theoretical
construct itself. We maximized our chance of testing collectivism and belief in empty claims
through our use of multiple common ways to operationalize each construct (as detailed in Table
1).
Any result can be attributed to multiple causes and no set of studies can rule out all
alternatives. In the current studies, we addressed four alternative accounts for why collectivism
may lead to belief in empty claims (credulity, reasoning style, bias toward in-group trust,
affective states).
The credulity account rests on the association between collectivism and agreeableness
(Burton et al., 2021) and between yea-saying in survey responses and collectivism (e.g., Smith et
al., 2016). We addressed these possibilities in Study 4 by assessing measures of each and asking
if collectivism matters once these constructs are taken into account.
The reasoning style account rests on the holistic reasoning style associated with
collectivism (e.g., Kühnen & Oyserman, 2002; Varnum, Grossmann, Kitayama, & Nisbett,
2010)
1
and underthinking associated with belief in empty claims. We addressed the former in
Study 4 and Pilot Study 1 by measuring holistic reasoning (Chiu, 1972) and the latter in our

1
The flip side of holistic reasoning is analytic reasoning. Though the term analytic appears in
both cultural psychology and dual-process perspective literature, it does not mean the same
thing. For cultural psychologists, analytic entails processing strategies with a focus on
contrasting and pulling-apart (Oyserman & Lee, 2008) or use of different kinds of rules (Chiu,
1972), while in the dual-process literature it entails deliberate, rule-based reasoning rather than
fast, gut- or gist-based reasoning (Frederick, 2005).
`

75
cross-cultural pilot studies (Supplemental Materials) by measuring accuracy on the cognitive
reflection test (Frederick, 2005; Toplak, West, & Stanovich, 2014) and need for cognition score
(Norris, Pacini, & Epstein, 1998). We did so even though we did not find any research indicating
a potential link between collectivism and lack of systematic, deliberative thinking.
The trust account rests on the finding that collectivism is associated with less general
trust (e.g., Romano, et al., 2017). Hence, if the mechanism is trust rather than generating
meaning, collectivism should not be associated with seeing meaning where none may exist. To
address this alternative, in Study 1, we showed that general trust is negatively associated with
collectivism but is also negatively related to belief in pseudoscientific claims, ruling out that
general trust is the mediator.
The affective state account rests not so much on collectivism as on prior research on
receptivity to fake news. These studies suggest that people might believe fake news more if they
experience a lack of control, negative emotions, low optimism, or feel isolated (Anthony &
Moulding, 2019; Whitson et al., 2015; Whitson & Galinsky, 2008). We explored these
alternatives in Study 3.
We determined the sample size based on available empirical evidence on plausible effect
size before we collected data collection for each study. We report how we determined sample
sizes (decision rules and a-priori power analyses), data exclusion criteria, manipulations and
measures, and the achieved power of main findings in each study. We used the Open Science
Framework to provide preregistrations, study materials, datasets, and codes (https://osf.io/jc9v6).
Study 1: Collectivism and Belief in Astrology
In Study 1, we tested H1 using available indicators of collectivism and belief in empty
claims (operationalized as believing that astrology is scientific) in a nationally representative
`

76
sample of Americans. Specifically, we predicted that people who endorse greater collectivistic
values are more receptive to the idea that astrology is scientific.
Participants
Participants were part of the General Social Survey (GSS) between 2006 and 2018 when
the GSS asked relevant questions (belief in astrology, collectivism). The GSS uses full-
probability sampling -- each household in the U.S. was equally likely to be selected. The dataset
included a nationally representative sample of American English-speaking adults (N = 5,114,
44% female, 75.0% White, 14.9% African American, 4.3% Latino/Hispanic, 3.0% Asian, 2.3%
other ethnicities, 0.5% no ethnicity information).
Measures
Independent Variable: Collectivistic Values. Respondents ranked the relative
importance from 1 (Least important) to 5 (Most important) of five things for a child to learn to
prepare him or her for life: “to obey”, “to think for himself or herself”, “to be well-liked or
popular”, “to help others when they need help”, and “to work hard”. We followed Hamamura
(2012) and operationalized collectivistic values as the mean of two negatively correlated items
(rs = .48, p < .001): the item related to collectivist values (“to obey”) and the item related to
individualist values (“to think for himself or herself” reverse coded) (Spearman-Brown
coefficient = .66).
Dependent Variable: Belief in Astrology. Respondents answered the question: “Would
you say that astrology is very scientific, sort of scientific, or not at all scientific?”
Control Variables. To rule out alternative explanations we included general trust
(suggested in the review process), gender, race, social class, and religiosity-spirituality as control
`

77
variables. General trust was assessed with a single-item measure “Generally speaking, would you
say that most people can be trusted or that you can't be too careful in dealing with people?”.
Results and Discussion
We regressed belief in the scientific merit of astrology on collectivistic values using
ordinal regression. The more people endorsed collectivistic values, the more they believed in the
scientific merit of astrology (Figure 1; OR = 1.27, 95% CI [1.21, 1.33], 𝜒 2
(1) =86.60, Cox-and-
Snell R
2
= .02, p < .001). Adding controls did not change this relationship (controlling for
general trust, gender, race, social class, and religiosity-spirituality, OR = 1.11, 95% CI [1.04,
1.17], p = .001; Table S1). General trust was negatively correlated with collectivism (r = -.20, p
< .001) and predicted lower belief in the scientific merit of astrology (OR = .69, p <.001; see
Supplemental Materials for more details). We present our results graphically on the left-hand
panel of Figure 1.
Our results provide initial support for our prediction that collectivism is associated with a
propensity to see meaning in empty claims. We examined lower general trust as an alternative
explanation, helpfully suggested by an anonymous reviewer. We did find this negative
association but general trust is not associated with finding astrology scientific. We take our
results to imply that collectivists do find astrology scientific not because they trust. They find it
scientific because they actively fill in the blanks.
Study 2: Collectivism and Belief in Pseudoscientific Claims
In Study 2 we tested H1 using available indicators of collectivism and belief in empty
claims in the Chinese Online Social Survey (COSS; Ma, 2017). The COSS yields two single-
item measures of belief in empty claims: We used belief in fortune-telling, palm-reading, Feng
Shui, and the extent to which people believed a pseudoscientific report about negative sperm-
`

78
count effects of radiation from WiFi. We predicted that collectivism would be associated with
higher belief in each case.
Participants
Our sample consisted of 9,638 Chinese internet users (37.51% female; 43.78% < 30 years
old, 47.65% 30 to 50 years old, 8.57% > 50 years old) who responded to the measure of
collectivisms and belief in pseudoscience in the 2014-2017 Chinese Online Social Survey
Measures
Independent Variable: Collectivistic Values. We operationalized collectivistic values
with four items in the COSS (𝛼 = .76): “One should follow parents’ requests, even if they were
unreasonable”, “it is natural that one should obey his or her boss or people with a higher status”,
“the most important thing for children to learn is to obey and respect authority”, and “one should
always subdue personal interests to pursue national interests if there is a conflict between the
two” (1 = Strongly disagree, 5 = Strongly agree). “Don’t know” responses, coded as missing
values, constituted <1% of responses.
Dependent Variable: Belief in Pseudoscientific claims. We used the two items
available in the COSS to measure belief in pseudoscience. First, respondents indicated their
agreement with the statement “fortune-telling, palm reading, and Feng Shui can explain a lot of
things and I believe in them.” (1 = Strongly disagree, 5 = Strongly agree). “Don’t know”
responses, coded as missing values, constituted 2.73% of responses. Second, a subset of 2,299
respondents read the following claim: “Here is a piece of news: Wi-Fi can unknowingly kill
sperm and induce sperm DNA damage. Radiation emitted from Wi-Fi sources is the cause of
sperm count reduction.” and answered the question “Do you think Wi-Fi kills sperm?” (No/Don’t
know or not sure/Yes). Respondents were randomly assigned to one of three groups. One group
`

79
read the claim and did not receive debunking information. The other two groups read the claim
then received debunking information from either a media source or a scientific source.
Respondents were also asked if they had heard of this news before.
Control variables. We included gender, age-group, educational-attainment, and family-
income as control variables.
Results
People who were higher in collectivism believed in fortune-telling, palm-reading, and
Feng Shui more (𝛽 = .30, F(1, 9636) = 951.27, p <.001, ∆R
2
= .09). They were more likely to
believe that Wi-Fi kills sperm (Figure 1 right-hand panel; OR = 1.46, 95% CI [1.31, 1.63], SE
= .08, p < .001). This relationship was not moderated by the condition they were in (see
Supplemental Materials for detailed analyses). Associations remained significant when we
controlled for age, gender, education, family income, and having previously heard the claim that
“Wi-Fi kills sperm” (Tables S2 and S3 in Supplemental Materials present full results).
Study 3: Collectivism and Belief in Fake News
In Study 3 we tested H1, operationalizing belief in empty claims as belief in fake news.
We also tested corollaries of H1: having false memories about having seen newly generated fake
news before (Study 3a, based on Murphy, et al., 2019) and willingness to share fake news (Study
3b). We tested our prediction using COVID-19 fake news (Studies 3a, 3b) and non-COVID-19
novel pseudoscientific news (Study 3c).
Study 3a
We conducted Study 3a during the peak of the COVID-19 outbreak in China (January to
February 2020). We preregistered H1 (that collectivism is associated with a higher likelihood of
believing existing and newly fabricated fake news about COVID-19) and a corollary of H1: a
`

80
higher likelihood of forming false memory of having seen the newly fabricated news. We
predicted that these associations should be robust to controlling for a low sense of control and
negative emotions that people are likely to experience during a pandemic.
Methods
We preregistered our prediction, sample size, and analyses on AsPredicted.org.
Participants. We preregistered to collect at least 193 based on a small-to-moderate
correlation effect size (r = .20) as suggested in Studies 1 and 2. We aimed to collect as many
responses as we could obtain during the COVID-19 outbreak in China to increase statistical
power. Participants were recruited from social media platforms and received 4 Yuan (the
equivalent of 0.60 USD) as compensation. Our final sample included 278 Chinese participants
(64% female; Mage = 26.34, SD = 9.17; excluding 3 participants who lived outside of China).
Materials. We used 9 news headlines about the COVID-19 outbreak (materials in
Supplemental Materials). To test the formation of false memory, we fabricated three critical
headlines that contained novel untrue information (e.g., Mass culling of wild animals in Wuhan)
that participants could not have seen before. We conducted online searches to ensure that
fabricated news stories of this kind did not already exist. We also included three fake news
headlines that have been officially debunked (e.g., Drinking strong liquor kills coronavirus), and
three real news headlines that contained truthful information (e.g., Coronavirus contagious in the
incubation stage). Each news headline was accompanied by a news image.
Procedures. We conducted our survey in Chinese online between January and February
2020. Participants saw 9 news headlines, one at a time, in a randomized order. We assessed false
memory following previous research (Murphy et al., 2019), asking participants to choose one
option for each headline (“I remember seeing or hearing this”, “I don’t remember seeing or
`

81
hearing this but I remembered it happening”, “I don’t remember seeing or hearing this” and “I
remember it differently”). The first two options were coded as “Remember” and the rest were
coded as “Don’t remember”. Each time a participant chose “I remember seeing or hearing this”
or “I remember it differently”, they were asked where (e.g., online news, social media,
newspaper, other people, don’t remember where they heard it). Then participants were asked if
they believed the reported news (e.g., “do you believe that drinking strong liquor kills
coronavirus?”). They completed a modified version of Oyserman’s (1993) 6-item collectivism
scale (α = .83). Finally, to address alternative explanations, we measured participants’ sense of
control and experience of negative emotions concerning the coronavirus (fear, worry, anger,
disgust) followed by gender, age, education, and the city that they currently lived in.
Results and Discussion
Belief and False Memories of Fabricated News. We tested our prediction using
regressions. People who scored higher in collectivism were more likely to believe newly
fabricated news about the coronavirus (𝛽 = .18, F(1, 276) = 8.93, p =.003, achieved power
= .86). This association was reduced when controlling for people’s sense of control, negative
emotions concerning coronavirus, gender, age, education, and the number of COVID-19 cases in
their province (𝛽 = .12, F(1, 257) = 3.61, p =.06; Table S4).
Collectivism increased the likelihood of forming false memories by affecting belief, as
indicated by the significant indirect effect of collectivism on false memory via belief in
fabricated news articles (indirect effect: ab = .01, SE = .005, 95% CI = [.003, .024], overall
model: F(1, 276) = 8.93, p = .003). To ensure that this process indeed reflected false memory
driven by people’s belief in fabricated news, we performed the same mediation analyses on
existing fake news and real news that people could have seen before. These mediation results
`

82
were not significant – collectivism did not increase people’s likelihood of reporting remembering
existing fake (ab = -.0002, SE = .001, 95% CI = [-.005, .001]) or real news (ab = .003, SE = .003,
95% CI = [-.002, .011]) by affecting their belief in these news articles.
Belief in Existing Fake News. Collectivism did not predict belief in existing fake news
about the coronavirus in China (𝛽 = .04, F(1, 276) = 0.54, p =.46). We followed up by testing the
moderating role of remembering seeing or hearing the fake news. We used a mixed-effect model
with belief in each news headline nested within participants as our dependent variable,
collectivism, remembering the news (yes/no), and their interaction as our predictor variables.
Whether participants reported remembering the news moderated the effect of collectivism (𝛽
= .22, p = .02). Subgroup analyses suggest that collectivism predicted belief in existing fake
news that participants remembered seeing or hearing (𝛽 = .16, p = .02). Collectivism did not
matter when participants reported not remembering seeing or hearing the fake news or reported
that they remembered the news differently (𝛽 = -.08, p = .28). One reason may be that China had
made a national effort to debunk COVID-19 misinformation; people who reported not
remembering the fake news or remembering it differently might attend to only trustworthy
sources or official debunking information, and thus had adjusted their belief accordingly.
Study 3b
While official news outlets in China featured debunking of COVID-19 misinformation, in
the U.S., debunked coronavirus claims were circulated on then President Trump’s Twitter
account. Hence, Americans were more exposed to fake news about COVID-19, even from a
seemingly credible source – a sitting president. This would seem to increase the likelihood that in
the U.S., people would not distinguish fake news from fact-based COVID-19 information. In
Study 3b we tested H1 (people higher in collectivism are more likely to believe fake news) and a
`

83
corollary (increased likelihood of sharing fake news).
Methods
Participants. Results in Study 3a suggested that the relationship between collectivism
and fake news is small-to-moderate (r = .20). Based on this assumption, we recruited 200
participants (53% female; Mage = 31.67, SD = 11.33; 72% European American, 9.5% Asian
American, 6% African American, 5% Latino American, 7.5% other ethnicities) from Prolific
and paid them 1.10 USD as compensation.
Materials. We used 10 news headlines about the COVID-19 outbreak (materials in
Supplemental Materials). Among them, 6 were fake news that contained information without
evidence (e.g., “Coronavirus survives and spreads faster in the snow”) and 4 were real news that
contained truthful information (e.g., “Coronavirus can survive on surfaces for days”).
Procedures. We collected the data in April 2020. Participants described their living
situation and physical distancing strategies during the COVID-19 outbreak, completed 5-item
social isolation (𝛼 = .90), 5-item negative emotion (fear, worry, anger, disgust, sadness, 𝛼 = .86)
and 2-item optimism (hope, optimism, 𝛼 = .89) regarding COVID-19 scales. Then we showed
them the 10 news headlines in randomized order, one at a time, asking them to rate each on
informativeness, truth (e.g., “do you think coronavirus survives and spreads faster in the
snow?”), and their likelihood of sharing it with others. Finally, they completed the same
collectivism scale as Study 3a (α = .75), reported their gender, age, highest education level, race-
ethnicity, and the state they currently lived in.
Results and Discussion
American participants who scored higher in collectivism were more likely to believe fake
news about the coronavirus in the U.S. (𝛽 = .20, F(1, 198) =8.21, p =.005, achieved power
`

84
= .81). This association remained significant when controlling for their feelings of social
isolation, negative emotions, and optimism regarding coronavirus, gender, age, education, and
the number of COVID-19 cases in their state (𝛽 = .15, F(1, 168) =4.39, p =.04). Collectivism
also predicted a higher likelihood of sharing fake news (𝛽 = .17, F(1, 198) = 5.81, p =.02).
Importantly, the effect of collectivism on sharing fake news was mediated by belief (indirect
effect: ab = .24, SE = .07, 95% CI = [.11, .40]).
Results from Studies 3a and 3b suggest that people who are higher in collectivism believe
COVID-19 fake news more. This misbelief, in turn, leads to false memories and sharing fake
news. Though high in external validity, our COVID-19 results might be due to political or other
social factors (e.g., conservatism, clarity, and uniformity of government COVID-19 messaging,
heightened threat). Therefore, in Study 3c we cross-validated our results with novel non-COVID-
19 pseudoscientific news.
Study 3c
In Study 3c we tested H1 with non-COVID-19 fake news. Specifically, we predicted that
collectivism is associated with a higher likelihood of believing randomly generated non-COVID-
19 fake news articles.
Methods
We preregistered our prediction, sample size, and analyses on AsPredicted.org.
Preregistration and participants. We preregistered our decision to collect data from
200 participants based on the assumption that the effect size would be small-to-moderate (r
= .20), consistent with the results we obtained in Studies 3a and 3b. We received slightly more
respondents (N = 202) from Amazon’s Mechanical Turk Services. Our final sample for analysis
included 141 American adults (55 females; 62.41% European American, 21.99% African
`

85
American, 7.80% Hispanic American, 4.26% Asian American, 3.55% other American) after
following our preregistered exclusion criteria (n = 56 duplicate geolocations indicative of
fraudulent overseas responses using Virtual Private Servers, n = 4 non-U.S. citizens, n = 1 non-
native English-speaker).
Materials. We randomly generated three correlation-based fake news articles (full
articles in Supplemental Materials) using Grover, an AI model that generates realistic-looking
fake news articles (Zellers et al., 2019). We used the Grover to randomly generate a list of
headlines. Empty claims typically take the form of a relationship between an action and an
outcome. Hence, we selected randomly generated news articles that had this form and were, to
our knowledge, both empty (not backed by any empirical evidence) and novel (not found in our
online search).
Procedures. We presented three randomly generated news articles (e.g., “eating pizza is
linked to financial security”), one at a time, to participants in an online survey. Participants rated
each article on informativeness (“How informative is the core message of this article?”),
meaningfulness (“How meaningful is the core message of this article?”), and belief (e.g., “Do
you think people who eat pizza are more financially secure than people who eat fast food”). We
averaged ratings across the three articles to form a score of belief in fake news (α = .94). Then
participants completed the collectivism scale we used in Study 3a (α = .87) and provided
demographic information (gender, ethnicity, U.S. citizen, first language, and highest education
attained).
Results and Discussion
Participants who scored higher on collectivism were more likely to believe randomly
generated non-coronavirus-related fake news (Study 3c: 𝛽 = .44, F(1, 139) = 33.50, p <.001,
`

86
achieved power > .99). This association remained significant when controlling for their gender,
race-ethnicity, and level of education (𝛽 = .36, F(1, 136) = 20.72, p <.001). We infer that the
association between collectivism and believing fake news may generalize across topic domains.
Study 4: Country Differences in Seeing Meaning Where None May Exist
In Study 4, we operationalized collectivism as a person’s country of residence, comparing
people in the U.S. and China. We compared the U.S. and China based on a large body of cross-
cultural research showing that China has a stronger cultural focus on collectivism than the U.S.
(for a meta-analysis, see Oyserman et al., 2002). We tested H1, preregistering two specific
predictions based on each operationalization of collectivism. First, people in China will see more
meaning in empty claims than people in the U.S. Second, in each country, people who score
higher in collectivism will find more meaning in empty claims. Alternative accounts posit that
people who endorse collectivism may report higher belief in empty claims because they reason
holistically or are highly agreeable (Burton et al., 2021) and agree to anything irrespective of
content (yea-saying; e.g., Johnson, Kulesa, Cho, & Shavitt, 2005). We tested these alternative
possibilities in this study by including measures of yea-saying, agreeableness (John &
Srivastava, 1999), and holistic thinking style (Chiu, 1972).
Methods
We preregistered our prediction, sample size, and analyses on AsPredicted.org.
Participants. We preregistered our plan to collect 240 responses (120 from each country)
based on the minimum size of the relationship we found between collectivism and belief in
vague statements in our pilot tests
2
(.18 < rs < .38). We preregistered to exclude people who

2
We conducted two exploratory pilot studies prior to Study 4. Supplemental Materials details
methods and results.
`

87
failed the attention check and, from our American sample, people who were not native English-
speaking or U.S. citizens. We recruited our American sample from the subject pool of the
University of Southern California. Each received course credit for participating. We excluded
non-U.S. citizens (n = 24), non-native English speakers (n = 31), and people who failed the
attention check (n = 4), yielding a final sample of 122 native English-speaking U.S. citizens (82
females; Mage = 19.69, SD = 1.59; 39.34% European American, 36.07% Asian American, 5.74%
African American, 18.86% other American). We recruited our Chinese sample from an online
student discussion board of Zhejiang University, offering 5 Chinese Yuan (the equivalent of 0.7
USD) for participating. We were caught by surprise at the swiftness of response, quickly
receiving 318 Chinese participants (135 females; Mage = 23.26, SD = 2.36; 49% undergraduates,
51% graduate students)
3
after excluding 3 participants who failed the attention check.
Procedure. Our survey was online in Chinese in China and English in the U.S. and
presented in the order we describe below (ending with age, gender, highest education attained,
first language, the state or province they grew up in, and for Americans, race-ethnicity).
Meaningfulness of randomly generated metaphors. We created 225 metaphor-like
sentences in each language of the form: “Love is a tree.” “Trust is sand.” Table S13 presents the
full list of 15 abstract and 15 concrete concepts that we randomly paired to create novel
metaphor-like sentences. Each participant saw five sentences that were randomly drawn from the
pool of 225 metaphor-like sentences. Each of the five sentences started with a different abstract
concept. We told participants that statements were from online sources and asked them to rate
how meaningful each was (1 = Completely meaningless to 7 = Very meaningful). Our method of

3
Supplemental Materials present similar results when only undergraduate students were
considered and when the first 120 responses from each country were considered.
`

88
having each participant rate a different set of metaphors ensures that differences are not
attributable to a particular metaphor.
The profundity of randomly generated vague statements. We used the Bullshit
Receptivity Scale by Pennycook et al. (2015). Specifically, we asked participants to rate how
profound (1 = Not at all profound to 7 = Very profound) they found each of 10 syntactically
correct statements randomly constructed from vague phrases (e.g., “Wholeness quiets infinite
phenomena”; American α = .82, Chinese α = .78).
Yea-saying. We asked participants how much they agreed or disagreed (1 = Strongly
disagree to 7 = Strongly agree) with five verifiable statements about mundane aspects of life
(e.g., “Most people enjoy some sort of music”) (Pennycook et al., 2015).
Belief in astrology. We asked participants how much they agreed or disagreed that
“Astrology has scientific truth” and “I think the horoscope can tell a person’s future” (1 =
Strongly disagree to 7 = Strongly agree; American α = .75, Chinese α = .85). Though originally
from the West, astrology is now a mainstream cultural trend in China (Qin, 2017).
Collectivistic values. We used Oyserman’s (1993) collectivism scale (American α = .68,
Chinese α = .69).
Agreeableness. We measured agreeableness using the trait agreeableness scale (John
&Srivastava, 1999; e.g., “I see myself as someone who is helpful and unselfish with others”; 1 =
Strongly disagree, 7 = Strongly agree; American α = .71, Chinese α = .71).
Thinking style. We measured holistic thinking with the 20-item triad task (Chiu, 1972),
which includes 8 critical triads and 12 filler triads. Participants saw a triad (e.g., doctor, teacher,
homework) and reported which two were most closely related. Holistic thinking was scored as
the proportion of relational responses in the eight critical trials in which items can be grouped
`

89
relationally (because they share a functional relationship) or categorically (because they share
category membership).
Results
Between-Country Differences. As shown in Figure 2, Chinese participants found more
meaning in randomly generated empty claims than Americans. Effects were consistent whether
the claim took the form of a metaphor-like sentence (Chinese M = 4.14, SD = 1.20 vs. Americans
M = 3.73, SD = 1.17, F(1, 438) = 10.70, p = .001, d = 0.35; Figure 2 left panel) or a non-
probative sentence formed from vague word-strings (Chinese M = 4.40, SD = 0.98 vs. Americans
M = 3.62, SD = 1.10, F(1, 438) = 51.98, p < .001, d = 0.75; Figure 2 middle panel). The right
panel of Figure 2 shows that though people in China (M = 2.81, SD = 1.45) believed more in
astrology than people in America (M = 2.66, SD = 1.52), this difference was not significant (F(1,
438) = .91, p = .34, d = 0.10).
Effects of Endorsing Collectivism. As shown in Figure 2, in each country, people who
endorsed collectivistic values were more likely to see meaning (except American participants’
belief in astrology). Indeed, even after taking into account between-country differences, people
who scored higher in collectivism found metaphor-like sentences more meaningful, (𝛽 = .25,
F(1, 437) = 26.99, p <.001, ∆R
2
= .06) and sentences randomly generated from vague word-
strings more profound (𝛽 = .26, F(1, 437) = 34.84, p <.001, ∆R
2
= .07), and believed in astrology
more (𝛽 = .16, F(1, 437) = 10.86, p =.001, ∆R
2
= .02). See Table S8 in SI for full model results
and Table S9 for similar within-country correlations.
Are Collectivism Effects Just Yea-Saying? If our results were simply due to yea-saying
(agreeing regardless of the content), the results should be the same for empty claims and
mundane, verifiable ones. That is not what we found. Agreeing with mundane and empty claims
`

90
did not follow the same country pattern. Americans agreed with mundane statements more than
Chinese participants, F(1, 438) = 12.39, p < .001, d = 0.37. Tables S8 and S9 in Supplemental
Materials detail further analyses on yea-saying.
Are Collectivism Effects Due to Agreeableness or Holistic Thinking?
The effect of collectivism cannot be explained by agreeableness or holistic thinking. As
detailed in Tables S10 and S11 in Supplemental Materials, between-country differences in
meaningfulness and profundity ratings and the associations between endorsing collectivism and
meaningfulness, profundity, and belief in astrology remained strong after controlling for
agreeableness and holistic thinking.
Study 5: Test of the Underlying Process: Meaning Making
In pre-registered Study 5, we tested H2 (our process prediction that collectivism increases
belief in empty claims by motivating people to actively generate ways to fill in the blanks and
create meaning). We operationalized empty claim as one of the metaphors we generated in Study
4. We asked participants to rate the metaphor’s meaningfulness and write down what came to
mind when they read it. We predicted that people higher in collectivism would find the metaphor
more meaningful and this relationship would be mediated by their likelihood of generating ways
in which the metaphor might make sense.
Methods
We preregistered our prediction, sample size, and analyses on AsPredicted.org.
Participants. We preregistered to recruit 250 participants (based on a small effect size of
r = .20). We recruited participants from Prolific. They received $0.64 for their participation. Our
sample consisted of 250 American participants (51.6% female; Mage = 31.24, SD = 11.56; 69%
European American, 12% Asian American, 8% African American, 6% Latino American, 7%
`

91
other ethnicities).
Procedures. Our empty claim was “love is a forest”, one of the metaphors we randomly
generated in Study 4. We told participants that the statement was taken from online sources and
that their task was to rate it for meaningfulness from 1 = Not meaningful at all, to 7 = Very
meaningful. Their response was carried forward to the next screen and embedded in the query:
“You rated it as a [the number they selected] of 7 on meaningfulness. Please write down below
what came to mind when you read the statement”. After this thought-listing task, participants
completed the collectivism scale used in Study 3 (α = .77), followed by demographics (gender,
age, whether they were U.S. citizens, race-ethnicity, and highest education).
Thought coding. Two research assistants who were blind to the study design and the first
author (blind to the collectivism data) independently coded each participant’s thought responses
into two variables. The first variable was any explanation of the meaning of the statement “love
is a forest” (1 = generated one or more explanations, 0 = did not generate any explanation). An
example response coded as 1 is “I imagined that it meant love is vast and expanding, easy to get
lost in”. An example response coded as 0 is “It sounds like an attempt to be poetic but comes
across as nonsensical”. The second variable was a continuous variable representing the number
of explanations generated for the meaning of the statement.
Our independent coding yielded excellent interrater reliability scores. Kappas ranged
from .84 to .86 for whether participants provided any explanation and our ICC = .95 for the
number of explanations. In cases of disagreement, three coders discussed until they reached an
agreement. Overall, about half (48.8%) of all participants provided one or more explanations for
why they thought love is like a forest. Participants who provided explanations provided on
average 1.80 explanations.
`

92
Results
Participants who scored higher on collectivism rated the randomly generated metaphor as
more meaningful (r(249) = .21, p < .001, achieved power = .92). To test our prediction that
collectivism is related to a higher likelihood of engaging in meaning-making, we fit a logistic
regression equation using collectivism scores to predict whether participants provided any
explanation. Supporting our prediction, participants who were higher in collectivism were more
likely to provide any explanation for why love can be like a forest (OR = 1.50, Z = 2.63, p
= .008, achieved power = .87). The pattern of results is consistent when we used the number of
explanations in an ordinal logistic regression
4
. People higher in collectivism provided more
explanations (OR = 1.47, t(249) = 2.56, p = .01). We used the number of words people wrote as a
way to rule out compliance as an alternative explanation. People’s collectivism score was
unrelated to the number of words they wrote (r(249) < .01, p = .99). This suggests that our
results were due to how people higher in collectivism engaged with the statement, not due to
their simply complying more by writing more.
Finally, we used the Medflex R package (Steen, Loeys, Moerkerke, & Vansteelandt,
2017) to test if the number of explanations mediates the effect of collectivism on the perceived
meaningfulness of a statement. We conducted mediation analysis using an imputation-based
approach to accommodate our ordinal mediator. Results suggest that the number of explanations
a person generated mediated the effect of their collectivism on how meaningful they found the

4
We pre-registered a continuous variable analysis but realized that this is an error and that the
number of explanations participants provided is an ordinal variable because successive units are
not equally spaced. An increase from 0 (no meaning-making) to 1 (a single meaning-making
explanation) is more substantial than an increase from 4 to 5 explanations. We present the similar
pattern of results if the measure is considered continuous in Supplemental Materials according to
our original preregistered analysis plan.
`

93
randomly generated metaphor (indirect effect = 0.22, SE = 0.09, p = 0.01, 95% CI = [0.06,
0.39]). Our results support our process prediction that collectivism enhances people’s tendency
to see meaning in empty claims by motivating them to consider how claims can be meaningful
and in doing so to construct meaning and even truth.
Study 6: A Test of Causality
In Study 6 we tested H1 as a causal claim. To do so, we momentarily induced people to
experience themselves as more collectivistic (vs. less collectivistic) using an experimental design
(pre-registered Study 6a and conceptual replication 6b). We predict that people led to experience
themselves as more collectivist have a stronger tendency to see meaning in empty claims.
Participants
In Study 6a, we preregistered the plan to recruit 300 participants based on the small-to-
medium effect size we obtained in a pilot study of our manipulation (See Supplemental
Materials). In both Study 6a and 6b we recruited American adults to complete a 3-minute study
on Amazon Mechanical Turk for $0.30. After exclusions (detailed next) our final samples were
American adult native speakers of English (Study 6a, n=288, Mage = 33.06, SD = 10.42; 44.1%
female, 68.4% White, 8.0% African American, 7.6% Latino or Hispanic, 12.1% Asian, 1.4%
Native American, 1.4% Mixed ethnicities, 1% Other American; Study 6bn=360, 44.0% female;
70.2% White, 14.8% African American, 7.5% Latino or Hispanic, 5.8% Asian, 1.1% Native
American, 0.6% Other American). We followed preregistered exclusion criteria, excluding
people whose first language was not English (n = 2) or had duplicate IP addresses (n = 10)
following recommendations for Mechanical Turk research that this may violate response-
independence (Berinsky et al., 2012). Similarly, in Study 6b, we excluded people who were not
U.S. citizens (n=4) and were not native speakers of English (n = 6).
`

94
Procedures
We told participants the study was about individual differences in judgment and opinion.
We used a force-agreement paradigm (Petrocelli et al., 2010), randomly assigning
participants to one of two groups: More Collectivistic (Study 5a: n = 148; Study 5b: n = 177) or
Less Collectivistic (Study 6a: n =140; Study 6b: n = 182). We asked the More Collectivistic
group to rate their agreement (1 = Slightly agree to 7 = Completely agree) with each of six
statements taken from the collectivism scale used in Study 3. We asked the Less Collectivistic
group to rate their disagreement (1 = Slightly disagree to 7 = Completely disagree) with the same
collectivism statements. Before proceeding, we pilot-tested the belief manipulation, finding that
it changed people’s momentary self-perception that they were collectivistic (see Supplemental
Materials for pilot results).
We included a different dependent measure in each study. In Study 6a, participants rated
the profundity of eight randomly generated vague sentences from Study 4 (Pennycook et al.,
2015; α = .86). In Study 6b, participants rated informativeness, meaningfulness, and belief of a
randomly generated fake news story from Study 3c (“texting decreases IQ”; α = .84). We placed
a manipulation check in Study 6a after the dependent measure, which included two items that
assessed collectivism (1 = Strongly disagree, 7 = Strongly agree; α = .72) and were not part of
the priming task. Finally, participants reported demographics (gender, first language, ethnicity,
highest education, whether they were U.S. citizens).
Results
Manipulation Check. Participants randomly assigned to the More Collectivistic
condition scored higher on the manipulation check collectivism measure (M = 5.04, SD = 1.05)
than participants randomly assigned to the Less Collectivistic condition (M = 4.71, SD = 1.23) in
`

95
Study 6a (F(1, 286) = 5.92, p = .02, d = .29).
Collectivism Increases Meaning-Making. Participants randomly assigned to the More
Collectivistic condition in Study 6a rated the vague word string sentences as more profound (M =
4.13, SD = 1.26) than those randomly assigned to the Less Collectivistic condition (M = 3.85, SD
= 1.33, F(1, 286) = 3.47, p = .06, d = .22. In Study 6b participants randomly assigned to the
More Collectivistic condition believed the randomly generated pseudoscientific news story
more(M = 4.39, SD = 1.53) than those randomly assigned to the Less Collectivistic condition (M
= 3.99, SD = 1.59, F(1, 357) = 5.68), p = .02, d = .26. Our meta-analytic synthesis of Studies 6a
and 6b (Z = 3.01, p = .003, d = 0.24, 95% CI [0.08, 0.39]) suggests that people see more meaning
after being led to consider themselves more collectivistic.
Study 7: The Absence of Communicator Moderates the Collectivism Effect
In Study 7 we tested H3. Specifically, we predicted that collectivism guides people to see
more meaning in empty claims because they are motivated to seek common ground with a
communicator. We tested this by manipulating whether a communicator is implied. If the
tendency to see meaning is driven by seeking common ground with a communicator then the
absence of a communicator should sever the collectivism-seeing meaning relationship.
Participants
We recruited 119 Chinese adults (38% female; Mage = 31.75, SD = 10.78) from a Chinese
crowdsourcing website (zbj.com) and paid them $0.70 for participation.
Procedures
The survey was online and in Chinese. We randomly assigned participants to one of two
conditions (Human Generated, Non-Human Generated). In the Human Generated condition, we
preserved the implicit assumption that content comes from another person. We asked participants
`

96
to choose a number to randomly select a statement in the form of “____ is ____” and then rate
the statement on meaningfulness. The unstated assumption was that the randomly selected
statement was created by a person (communicator). In the Non-Human Generated condition, we
made it clear that the content did not come from another person. We asked participants to choose
numbers to randomly generate the first and second parts of a statement ( “____ is ____”) and rate
its meaningfulness. Participants had no reason to infer the existence of a communicator since
they were told that they formed the statement by drawing two numbers. To make sure that people
all saw the same stimuli, we gave participants in both conditions the same five metaphor-like
sentences (e.g., “Time is air”) from our Study 4 pool of metaphors. After the rating task,
participants completed the collectivism scale we used in Study 3 (α = .69) and reported their
gender, age, the province they grew up in, their religiosity, and spirituality.
Results
Assignment to condition did not affect participants’ rating of meaning, ruling out the
possibility that the manipulation changed the meaning of statements (F(1, 117) = .27, p = .60).
Instead, we found a significant interaction between Condition and collectivism on meaning
ratings (F(1, 115) = 7.72, p = .006, ∆R
2
= .06). The achieved power (.80) was sufficient to detect
this interaction.
To understand this interaction, we tested the relationship between collectivism and
meaning ratings in the Human-Generated condition and the Non-Human Generated condition,
respectively. As we depict in Figure 3, what people likely assumed about the existence of a
communicator mattered. People who were randomly assigned to the Non-Human Generated
condition were explicitly told that the content they saw was not generated by a human. In
contrast, people who were randomly assigned to the Human-Generated condition did seem to
`

97
infer that the content they saw came from another person. Collectivism was positively related to
seeing meaning if a human communicator was assumed (blue line, r(56)= .42, p = .001) and was
unrelated to seeing meaning if a human communicator was explicitly excluded (red line, r(63)=
-.08, p = .54). People who were higher in collectivism were more likely to find meaning in empty
claims only if they were seeking common ground with an implied communicator.
Meta-analysis: Does Collectivism Increases Belief in Empty Claims across Studies?
We conducted a single-paper meta-analysis using a random-effects model across our ten
studies to test the overall effect of collectivism on belief in a variety of empty claims. Results
support our prediction that collectivism increases people’s belief in empty claims (d = 0.39, 95%
CI = [0.24, 0.43], Z = 5.21, p < .001). For interested readers, we present the forest plot in our
Supplemental Materials.
Discussion
We started with the observation that people commonly find meaning in empty claims.
Building on our synthesis of an evolutionary perspective on cultural accumulation (Mesoudi &
Thornton, 2018) and culture-as-situated cognition theory (Oyserman, 2017), we suggested that
an aspect of human culture, collectivism, could help explain why this is the case. People need
others and collectivism is that aspect of human culture that sensitizes people to the need to fit in.
Moving from this universal level to societal, situational, and individual levels, collectivism
motivates people to generate meaningful ways in which claims might make sense so that they
can attain a common ground with communicators. By generating meaning, people become
convinced, and may even see meaning when none may exist.
We documented an association between collectivism and belief in real-world empty
claims (e.g., astrology, superstitions) using two national samples. We showed that this
`

98
association can be generalized to a variety of empty claims, including newly created or widely
circulated fake news about COVID-19, AI-generated pseudoscientific news articles, and
sentences constructed with random components. We ruled out yea-saying, agreeableness, holistic
thinking, education, and religiosity-spirituality as alternative explanations. The collectivism
effect is causal: people induced to momentarily experience themselves as more collectivistic are
more likely to find meaning in empty claims. This occurs, in part, because collectivism motivates
people to generate reasons that a claim might be meaningful. Indeed, collectivism is only
associated with seeing meaning in empty claims that seem to come from another person. The
implication is that people make meaning because they are motivated to seek common ground
with a potential communicator.
Theoretical Implications
Our collectivism account builds on culture-as-situated cognition theory (Oyserman, 2017)
and expands prior cultural psychological research on sensitivity to the common ground of
communication. Prior studies show that priming collectivism increases sensitivity to the common
ground of communication (Haberstroh et al., 2002) and perspective-taking (Wolgast &
Oyserman, 2019). Our results suggest that people higher in collectivism are spontaneously
attuned to what others are trying to communicate, presuming that any claim they see is created
by another person and hence is supposed to have meaning. This inference potentially underlies
prior research documenting that people higher in collectivism can reconcile conflicting
perspectives (Peng & Nisbett, 1999) and agree with seemingly opposing survey statements
(Smith et al., 2016). Our results suggest that people higher in collectivism do so by considering
how each claim might be true.
`

99
Our focus on culture as a driving mechanism of seeing meaning and truth in empty
claims is novel and notable. It is distinct from the two existing accounts which focus on
underthinking (Pennycook et al., 2015; Risen, 2016) and motivated reasoning (Kahan, 2012). We
interpret our results as suggesting that vulnerability to empty claims is not simply a consequence
of underthinking or motivated reasoning. It can stem from overthinking driven by a need to relate
with others and a focus on how claims might make sense.
As such, our collectivism account sheds light on otherwise inexplicable results. For
example, it reconciles the finding that Americans who engage in deliberate thinking are less
likely to believe empty claims (Bago et al., 2020; Pennycook et al., 2015), while the opposite is
true for Japanese (Majima, 2015). An underthinking account can explain the American but not
the Japanese results. Our account can reconcile conflicting findings by considering that people in
Japan are more likely to have a collectivistic focus (Kitayama & Imada, 2010). This motivates
them to seek common ground and fill in the blanks when claims are empty. Succeeding at filling
in the blanks is more likely when engaging in more thought, as shown by Majima (2015) among
Japanese participants. Our results suggest underthinking is not the only path to belief in empty
claims; this belief can also stem from overthinking with a focus on making sense.
Practical Implications
Situations that trigger collectivism are likely to increase people’s susceptibility to empty
claims. Because pathogen risk predicts higher collectivism (Fincher et al., 2008), our work
applies to public acceptance of misinformation surrounding COVD-19. Our results suggest that
people are vulnerable to misinformation not simply because they are too lazy to think, but
because they are motivated to find meaning in content provided by others. Our work implies that
to undo the effect of collectivism, people’s motivation to seek meaning needs to be disrupted.
`

100
This can be done, for example, by pointing out that the content comes from a nonhuman
communicator (e.g., Internet bots) or an untrustworthy source.
Alternative Explanations
We predicted that collectivism is associated with belief in empty claims because it
prompts people to seek common ground by filling in the blanks, and in so doing, generating
meaning where none may exist. Our process model builds on prior research documenting that
collectivism heightens context-sensitivity (Hall, 1976; Haberstroh et al., 2002) and perspective-
taking to see the common ground of communication (Wolgast & Oyserman, 2020). For example,
people primed with collectivism are more likely to provide disparate answers to two redundant
survey questions because they pay closer attention to the common ground to determine what the
questioner wants to know (Haberstroh et al., 2002). In this section, we consider three alternative
explanations for our results (credulity, reasoning style, bias toward in-group trust). As we detail
next, while each has an interesting association with culture, we do not find evidence that these
are sufficient alternative explanations to our process prediction of actively filling in the blanks in
pursuit of seeking common ground with a communicator.
First, consider credulity, the willingness to accept and believe claims without really
processing them. At least in some situations, people might be more willing to accept persuasive
arguments if the arguments are framed in culturally fluent terms (Oyserman, 2018) and linked to
their social identities (Oyserman & Dawson, 2020). In these situations, people may be credulous,
accepting arguments without really processing them. But in these cases, the mechanism is
cultural fluency, things unfolding as culturally expected, not collectivism. Indeed, our Study 5
results are incompatible with this alternative explanation that collectivism triggers acceptance of
claims without processing them. In Study 4 we did not find that yea-saying (agreeing regardless
`

101
of content) explained the effect of collectivism) and in Study 5 we found that people higher in
collectivism actively made connections and constructed meaning. They created something that
did not exist before. The more connections they made, the more meaning they found in the
randomly generated metaphor “love is a forest”. Consider the response a participant who rated
“love is a forest” as meaningful: “It makes sense to me - love is a forest because there's a lot to
explore and learn about someone you love, and love is a beautiful thing, but like a forest, it can
also be scary and new. It can be dangerous depending on what challenges you come across, but it
can be fun.” As this example highlights, what drives the high meaningfulness rating is not simply
accepting a claim but rather an active generation of meaning that is unique and personal beyond
the literal meaning of the words.
Next, consider reasoning style. Collectivism has been associated with more use of holistic
reasoning and less use of analytic reasoning (Varnum et al., 2010). In this body of work, holistic
reasoning is characterized by a focus on contextual information and relationships among objects,
whereas analytic reasoning is characterized by a focus on the main point and rule-based
categorizations of objects. We did not find any literature associating holistic reasoning with
belief in empty claims. Nonetheless, we tested the possibility that our collectivism effects are due
to holistic reasoning in two cross-cultural studies (Study 4 and Pilot Study 1) using a classic
paradigm -- the triad task. Compatible with other research on country-level differences in holistic
reasoning (e.g., Ji et al., 2014), Chinese participants were higher in holistic reasoning and lower
in analytic reasoning than American participants in both studies. However, we failed to find any
evidence that holistic reasoning was related to belief in empty claims. Between-country
differences in belief in empty claims and the effect of collectivistic values on belief in empty
claims remained unchanged when controlling for holistic reasoning. Hence, our data do not
`

102
support holistic reasoning as an alternative explanation for our results. At the same time, our data
are compatible with research showing that when primed with a collectivistic mindset, people are
better at solving less specified problems (Arieli & Sagiv, 2018). We believe that the
collectivism-induced, active generation of meaning may be the process behind this finding as
well. We look forward to future research exploring other ways in which this aspect of
collectivism affects human judgment.
Finally, consider potential bias toward ingroup trust. A feeling of belongingness to in-
groups is a common way of operationalizing collectivism (Brewer & Chen, 2007). One
implication is that collectivism triggers an in-group feeling, a sense of trust centered on the in-
group, not others (Romano, et al., 2017). To the extent that a human communicator is assumed to
be a member of the in-group, that might heighten trust. Our Study 7 results suggest that people
may spontaneously assume a human communicator is behind claims, though it is unclear whether
they are assuming that the communicator is an in-group member. That said, trust is not the same
as gullibility. Indeed, Yamagishi, Kikuchi, and Kosugi (1999) showed people are more sensitive
to signals of potential untrustworthiness if they are high in trust. Our Study 1 results also show a
negative association between general trust and acceptance of pseudoscience. General trust cannot
mediate the positive relationship between collectivism and accepting pseudoscience because it is
negatively associated with both.
Limitations and Future Directions
Our research sheds light on possible avenues for future research. Many of our tests
involved online participants from crowdsourcing platforms. These platforms engage non-college-
student adults who are relatively more diverse and less economically advantaged than college-
based samples (Paolacci & Chandler, 2014). Despite concerns about these samples, comparative
`

103
analyses suggest that these participants are at least as careful, if not more so than college students
(e.g., Hauser, Paolacci, & Chandler, 2019). Though, of course, these platforms do not provide
representative samples of national populations. We address this issue of ecological validity by
using a U.S.-based and a China-based national sample in Studies 1 and 2. In both countries,
within-society variations in collectivism correlate with belief in empty claims. We showed that
collectivism correlates with these beliefs even once we controlled for demographic factors
(people's age, religion, ethnicity, and socioeconomic status). That said, we focus on the kinds of
empty claims found in modern, industrialized societies. We cannot be certain about
generalizability to premodern societies. Though we tried to include a variety of empty claims, we
cannot fully delineate the population of such claims. Instead, we used a variety of
operationalizations of our core concepts (collectivism, empty claims) across studies to increase
confidence that our findings shed light on our core concepts rather than only on a particular
operationalization. Although some of our tests are correlational and cannot indicate causality, we
find converging evidence supporting the effect of collectivism by comparing countries and
experimentally manipulating collectivism.
Our research also points to other ways in which collectivism can lead to belief in empty
claims, providing promising avenues for future research. A fruitful next step would be to
examine other factors that increase motivation to seek common ground as these may also trigger
the same meaning-making processes we have identified for collectivism. For example,
collectivism and perspective-taking are more likely in situations in which people are interacting
with close others, people from their in-group, or people who have power over them (Boothby et
al., 2016; Galinsky et al., 2006). Though we did not find an association with holistic reasoning, it
is possible that such a relationship exists, given for example studies showing an association
`

104
between collectivism and generating solutions to loosely structured problems (Arieli & Sagiv,
2018).
Conclusion
To satisfy a human need to relate and fit in, people attempt to see what others see by
asking themselves “how might this claim make sense?” In doing so, people self-convince.
Collectivism increases seeing meaning where none may exist. This very human sensitivity to the
communicative intent of others is likely to be a reason why conspiracy theories, fake news, and
pseudoscience spread. A core implication of our work is that to reduce acceptance of
misinformation, people’s perspective-taking tendencies need to be disrupted. Considering the
cultural roots of the tendency to see meaning where none may exist may be one important step to
counter the spread of false information in the public sphere.

`

105
Chapter 3: Cultural Fluency Means All Is Okay, Cultural Disfluency Implies Otherwise
Lin, Y., Arieli, S., & Oyserman, D. (2019). Cultural fluency means all is okay, cultural
disfluency implies otherwise. Journal of Experimental Social Psychology, 84, 103822.

Abstract
Being part of a culture means knowing what to expect in most everyday situations --with the
implication that something may be awry if unfolding situation mismatches culture-based
expectation. We tested the prediction that culture-based mismatches challenge people’s sense
that current patterns (e.g. the color of money, the taste of toothpaste) represent a natural order,
calling into question whether social categories have stable essences. To do so, we asked people
in China, Israel, and the U.S. (N=1,803) to rate products (e.g., breakfast plates, wedding
photographs, Valentines) then complete unrelated scales, randomly assigning them to products
that matched or mismatched their respective cultural expectations. Exposure to mismatch
reduced psychological inherence --the feeling that existing patterns in the world reflect how
things ought to be in unrelated domains and this reduced cultural essentializing (the feeling that
cultures have fixed essences that cannot change). Effects were small-to-moderate-sized and
consistent across countries.

`

106
Culture is the set of practices that people in a time and place come to accept (Chiu et al.,
2015; Oyserman, 2017; Shteynberg, 2015; Triandis, 2007). Though cultures are not fixed and do
change over time, within each point in time, being part of a culture provides a culture-specific
vantage point or meaning-making organizing lens. People who are part of a culture know what to
expect, what “we” do and how “we” do it, and for two reasons, mostly experience situations that
seem to match these expectations. First, they have expertise about their culture in that time and
place. Second, like other people, they have a tendency to see what they expect to see (variously
termed confirmation bias, Wason, 1960; self-fulfilling prophecies; Merton, 1948; Snyder, 1984;
stereotype confirmation, Hamilton & Trolier, 1986). Yet, people do sometimes experience
violation of their culture-based expectations and in this paper we examine the downstream
psychological consequences of these experiences for people’s basic sense of the world as an
orderly place, a place in which the way things are is the way they ought to be. We build on
culture-as-situated-cognition theory (Oyserman, 2017) to make two original predictions. First,
when people experience a mismatch between what their cultural expertise leads them to expect
and what they actually observe, they experience a loss of what Cimpian (2015) describes as
psychological inherence --the sense that the way currently things are is the way they should be.
Second, as a result of this loss of inherence, people are less likely to essentialize social categories
(Gelman, 1999) --to experience cultures and themselves as having fixed essences. Figure 1
presents our theoretical process model. As we describe below, we also explore a number of
potential individual differences that might moderate this process.
Culture-as-situated-cognition
Culture-as-situated-cognition theory starts with the idea that people have available in
memory an array of culturally rooted associative knowledge networks (Oyserman, 2011;
`

107
Oyserman & Yan, 2018). These culturally rooted associative knowledge networks include
content, procedures, and goals related to everyday life (e.g., what breakfast entails, what playing
cards look like) and to overarching cultural themes (e.g., individualism, collectivism, and honor).
As they go about their day, people use the cultural-rooted associative knowledge networks that
are accessible to them at the moment of judgment to make automatic, often tacit, predictions
about what will happen next (Oyserman, 2017). Because predictions are tacit and automatic,
people may not notice that they are making predictions at all and instead simply experiences the
cognitive and psychological consequences of match and mismatch between prediction and
observation (Oyserman & Yan, 2018). Matches yield the sense that all is fine, that one can ‘go
with the flow’ and navigate one’s everyday lives intuitively with minimal cognitive resources;
mismatches yield the opposite sense that something might be awry, shifting people to
deliberative, systematic, rule-based reasoning strategies. Indeed, culture-as-situated-cognition
theory predicts that when culture-based predictions seem to match observation, thinking feels
easy (fluent), in contrast, when culture-based predictions seem to mismatch observation, thinking
feels difficult (disfluent). The terms cultural fluency and cultural disfluency were coined to
capture these effects of cultural knowledge on people’s experience and its downstream
consequences (for extended reviews, Oyserman & Yan, 2018; Oyserman, 2011).
An early example documenting that culture-based predictions affect processing fluency
comes from Bruner and Postman (1949). These researchers exposed American college students
to cards with diamonds or hearts and assessed how long it took these students to correctly
identify the image they were shown. The cards were playing cards, though students were not told
so. The researchers modified some cards so that they mismatched with culture-based
expectations as to what playing cards look like and divided students into four groups with each
`

108
group differing in the cards they were exposed to. One group only saw cards that matched
culture-based expectation about the link between shape (heart, diamond) and color (red, black) --
so hearts and diamonds were red. A second group saw only cards that mismatched culture-based
expectations about the link between shape and color -- so hearts were black. The third and fourth
groups saw different proportions of cards with color and shape combinations that matched and
mismatched culture-based expectations -- so sometimes hearts were red and other times hearts
were black. The effect was clear; students took longer to correctly identify shapes that
mismatched culture-based expectations of color and shape combinations for playing cards. Lag
in response was especially pronounced in two situations: on first trials if that first card was a
mismatch to culture-based expectation, and on the first mismatch to culture-based expectation
after a run of matches to culture-based expectation.
Though not labeled a study of cultural fluency and disfluency by the original authors, this
study does illuminate the basic idea that matches to cultural expectation make things easier (in
this case, faster) to process and that mismatches carry signal value as well. Note that we consider
the card study effects to be culture-based because the effect was not due to the ease of processing
a particular shape or color -- it is not that a specific shape was always easier to identify or that a
particular color always made a shape easier to process. Rather, it is that participants came into
the experiment with culture-based knowledge of how playing cards “ought” to look and applied
this knowledge automatically to the situation. They did so even though the experimenters never
labeled the cards as playing cards and never told them that the cards were supposed to match
their knowledge of playing cards to the situation. Time lag effects were largest at the first
instance of mismatch (first card, first after many matches) and diminished as participants came to
see that their automatically recruited “playing cards” cultural knowledge might not be
`

109
pragmatically relevant to the task at hand. Culture, of course, is dynamic, and that experiment
will only replicate among current American college students if playing cards are as common a
pursuit now as it seems to have been when the experiment was originally conducted in the 1940s.
As this example highlights, when something is culturally fluent or culturally disfluent, the
source of fluency and disfluency is not a feature of the target of judgment (e.g., the diamond)
alone or a general feature of the situation (e.g., the playing card, the lighting of the room, the
color contrast between the shape and the card background). Rather, it is the result of the match or
the mismatch between the target and culture-based expectation about the target. In the cultural
context of “playing cards” diamonds are supposed to be red, not black. The experience of
cultural fluency and disfluency is based in cultural knowledge, applied to situations in which it
seems relevant. Cultural knowledge sets up implicit expectations, which if met, make processing
easier, and if violated, make processing more difficult. Hence cultural fluency and cultural
disfluency require that cultural expectations are triggered and experienced as relevant, otherwise
culture-based predictions do not apply.
Cultural fluency and disfluency cues can come from central or peripheral features of
cultural knowledge. Central features have meaning on their own while the meaning of peripheral
features is more dependent on features of the context. Evidence for the downstream cognitive
consequences of cultural fluency and disfluency cues comes from seven experiments in the U.S.
and Hong Kong (Mourey, Lam & Oyserman, 2015). In these experiments, the researchers
randomly assigned participants to one of two groups, showing each group cultural products
including wedding photographs, obituaries, and patterned plates. One group saw a version of the
product that likely matched cultural expectations while the other group saw a version that likely
mismatched cultural expectations. The downstream effect of engaging with products that
`

110
matched or mismatched with likely culture-based expectation was tested in two ways: by
assessing consumption in appetitive contexts --willingness to buy or the actual amount of food
put on plates, or by assessing systematic reasoning on a formal reasoning task.
For example, to test effects in appetitive contexts, Mourey, Lam, and Oyserman (2015)
assessed willingness to buy an unrelated object (a shovel) after viewing wedding photographs in
one study and the amount of food people put on their plates at holiday events in other studies.
The wedding photographs (e.g., bride in white, groom in black vs. bride and groom in different
colors) and match or mismatch between the holiday and the plate design served as cultural
fluency and disfluency cues. Americans were more willing to buy a shovel after seeing the
culturally fluent rather than the culturally disfluent weddings. Americans put more food on their
plates at 4
th
of July and Memorial Day picnics if plates matched their culture-based expectation
of a patriotic theme (stairs and stripes) --compared to plates that mismatched culture-based
expectation and had a neutral (white only) or irrelevant theme (bats and pumpkins). Similarly,
Hong Kong Chinese participants put more food on their plates at a Chinese New Year buffet if
plate border color matched a peripheral cue, red, the color of Chinese New Year compared to
when it mismatched that cue (black border).
To document that these effects were due to cultural fluency rather than to features of the
product (e.g., maybe the color red is an appetite stimulant), the researchers designed follow-ups
that capitalized on the fact that in the Chinese New Year study, the cue was peripheral and so
dependent on other cues (e.g. time of year). In the first follow-up, red-bordered plate had no
effect on how much food Chinese participants put on their plates a month after Chinese New
Year when knowledge about Chinese New Year was no longer relevant. In the second follow-up,
the researchers documented that cultural fluency effects require cultural knowledge by using a
`

111
sample of Americans. In this sample, the red-bordered plate had no effect on the amount of food
participants put on their plates because they had no cultural expertise about Chinese New Year
(they were asked but could not identify when Chinese New Year occurs or what is done to
celebrate it). These results reinforce a cultural fluency and disfluency interpretation of the picnic
and buffet findings. That is, cultural fluency requires cultural knowledge and effects were due to
cultural fluency (the match between automatic culture-based expectation about the holiday and
designs) rather than to something about the design itself.
To test effects of cultural fluency and disfluency on gut-based versus rule-based
reasoning style, the researchers used Valentine’s Day and a peripheral cue, a Valentine’s color,
pink (Study 6), as well as central cues regarding weddings and funerals as detailed below. They
tested participant responses during and a week after Valentine’s Day, choosing participants from
the U.S. and Hong Kong (both countries celebrate Valentine’s Day in a similar way and on the
same day). Participants completed a test of spontaneous rule-based (systematic) reasoning. They
were randomly assigned to either take the test on a screen with a pink border or to take the test
on a screen with no border or a black-and-white border. One group of participants completed the
test on Valentine’s Day; the other half completed the test a week later. Participants in the
culturally fluent (pink border on Valentine’s Day) condition were more likely to use gut-based
reasoning than participants in the other conditions. Pink is a peripheral cue to Valentine’s Day,
effective on Valentine’s Day; a week after Valentine’s Day, it is just a color, neither culturally
fluent nor culturally disfluent. The effect of cultural fluency and disfluency on reasoning style is
quite stable; the researchers replicated the effect with central cultural cues using photographs of
wedding and texts from obituaries. In the wedding studies, participants rated the quality of
wedding photographs that were either of culturally fluent weddings (groom in black and a bride
`

112
in white) or of culturally disfluent weddings (groom in purple, bride in green). In the funeral
studies, participants were exposed to the texts of either culturally fluent obituaries (sadness, loss
of a loved one, extol virtues) or culturally disfluent obituaries (no virtues, no sadness, not loved).
Participants randomly assigned to the culturally fluent condition scored worse on the systematic
reasoning task than participants exposed to culturally fluent condition.
The researchers tested the possibility that cultural fluency and disfluency effects are
explained by positive and negative affect. They did not find that exposure to culturally fluent
products was associated with positive affect or that exposure to culturally disfluent products was
associated with negative affect. They also tested the possibility that cultural fluency and
disfluency effects are driven by product quality, attractiveness, and traditionality ratings but
again failed to find any mediation or moderation. The implication is that cultural fluency and
disfluency are basic cues, informative of whether one can ‘go with the flow’ consume what is
available, reason with one’s gut or if caution and systematic reasoning in necessary.
Inherence Underlies Psychological Essentialism and Categorical Reasoning
Prior studies are important, showing that cultural fluency affects processing speed,
appetitive consumption, and cognitive processing style and that results cannot be explained as
mood effects or as consequences of fluency effects on product quality and attractiveness ratings.
However, prior research and theorizing do not address how cultural fluency and disfluency might
shape worldview at a more basic level --by influencing one’s sense that existing patterns in the
world are ideal. This feeling that current patterns are the natural order; the way things ought to be
is termed psychological inherence (Cimpian, 2015). It is an important cognitive precursor of
category learning via its connection to psychological essentialism, the belief that categories are
stable, inevitable, and immutable (Salomon & Cimpian, 2014).
`

113
People who score higher in inherence are more likely to essentialize the world around
them (Salomon & Cimpian, 2014). Essentialist reasoning emerges at an early age, and is
universal, sticky, and consequential. That is, children infer value from unseen essences and adult
reasoning retains these patterns (Gelman & Echelbarger, 2019). They do so in part because
essentialist reasoning facilitates efficient learning and category-based prediction by implying that
categories are not haphazard but natural and include essences not visible to the naked eye
(Gelman, 2003; Gelman & Diesendruck, 1999; Medin & Ortony, 1989; Rhodes, 2013).
Essentialist reasoning about the self allows people to use self-knowledge to make predictions
about future preferences (Oyserman, 2019). However, essentialist reasoning also has negative
consequences. It is associated with acceptance of stereotypes (Bastian & Haslam, 2006) and
race-based inequalities (Williams & Eberhardt, 2008). If people essentialize social categories,
they are more likely to experience differences, including differences between cultures, as
immutable, with potentially negative consequences for engagement, trust, and cooperation
(Bastian & Haslam, 2006; Chiu, Dweck, Tong, & Fu, 1997). People are less likely to counter
argue persuasion attempts linked to categories experienced as natural and true, especially if these
categories, like culture, are experienced as self-relevant (about “me” or “us”, Oyserman, 2019).
What Else Do We Know About Fluency and Disfluency Effects
People can experience processing fluency and disfluency at the perceptual level -- a
smudged picture, text written with hard-to-read font, poor color contrast. Each is visually more
difficult to process than clear pictures, easy-to-read font, and sharp color contrast (see Schwarz,
2004; Reber, Schwarz, & Winkielman, 2004). These processing effects are perceptual rather than
culture-based -- sharp is easier to process than smudged due to the functioning of our visual
systems. Perceptual fluency affects subsequent ratings of attractiveness, quality, innovativeness
`

114
(Schwarz, 2015), and psychophysiological measures capture subtle effects on affect
(Winkielman, Huber, Kavanagh, & Schwarz, 2012; Winkielman, Schwarz, Fazendeiro, & Reber,
2003). People are sensitive to their metacognitive experiences of processing ease and difficulty,
but not to the source of these experiences and so may misattribute processing (dis)fluency that is
not inherent to the target of judgment as being due to the target of judgment (Schwarz, 2015).
For example, when presented information about a product with difficult-to-read font, people will
rate the product as less attractive, but more innovative.
People can also experience fluency and disfluency at a conceptual level, that is, people
are quicker to recognize a key after seeing pictures of locks (see Reber et al., 2004). Effects of
conceptual fluency parallel those of perceptual fluency (Schwarz, 2015). Though described as
being due to familiarity with concepts, conceptual fluency requires culture. That is, people
expect to see keys after seeing locks because they have culture-based knowledge about locks.
Cultural fluency and disfluency research builds on this basic, but unspoken, premise of
conceptual priming research.
As distinct from prior research, cultural fluency and disfluency research uses finely tuned
cultural products like weddings and Valentine‘s Day cards. Using a culture-as-situated cognition
perspective leads to questions not asked within the larger fluency literature in part because by
neglecting culture, these questions have not come to mind and in part because the priming tasks
used did not allow them to be addressed. For example, while seeing a key makes a lock easier
(more fluent) to process, there is no parallel disfluent condition as is the case in cultural fluency
and disfluency.
Until now, research on cultural fluency and disfluency focused on ecological validity,
using available products varying on whether or not they meet cultural-based expectations.
`

115
Participants have been incidentally exposed to cultural products in one setting and the effects of
exposure were tested either in that same setting or in a subsequent one. Care has been taken to
never explicitly forewarn participants that the cultural products may have unexpected
components. Yet the logic of priming, generally (Bargh, 1994) and in the context of culture
(Oyserman, 2016), suggests that people use what is on their mind at the moment of judgment if it
is experienced as relevant to the judgment at hand. What this implies is that people are likely to
experience the consequences of cultural fluency or disfluency whenever the intuitions that come
to mind due to culture-based expectations are experienced as relevant to the judgment task. This
should be the case whether or not people are explicitly forewarned that they will be examining
expected or unexpected product designs, usual or unusual product combinations because the
experience of fluency is inherent to cultural expertise and lack of fluency is a problem signal.
Are There Individual Differences in Responsivity to Cultural Fluency and Disfluency?
Until now, we have focused on the general process triggered by cultural expertise,
describing how experiences of cultural fluency and disfluency facilitate everyday life by
allowing people to know when something might be awry. Our prediction is that this will yield
downstream effects for people’s sense of inherence and hence for their essentializing. In this
section, we briefly consider some possible individual difference variables that might moderate
the effect of cultural fluency and disfluency. We focus on individual differences in cultural
values and in response to the unexpected because they may affect the extent to which disfluency
becomes cause for concern. Specifically, we consider individual differences in endorsement of
cultural values relevant to concern about fitting in versus appreciation of uniqueness as well as
individual differences in need for cognition, need for structure, and intolerance of uncertainty. In
Figure 2, we show these potential moderators of the effect of experienced disfluency on
`

116
experienced inherence (we thank our reviewers for suggesting the addition of need for structure
and intolerance of uncertainty as potential moderators).
We considered the possibility that endorsing the values of collectivism or the values of
individualism might moderate the downstream consequences of experiencing cultural fluency
and disfluency. With regards to collectivism, endorsing the values of collectivism implies one
attributes importance to fitting in, following the norm, and accepting tradition. Higher
endorsement of these values might enhance the rattling effect of cultural disfluency on one’s
sense that all's right with the world. With regards to individualism, endorsing the values of
individualism implies that one values uniqueness, difference, and distinction. Higher
endorsement of these values might dampen the rattling effect of cultural disfluency on one’s
sense that all’s right with the world. In exploratory analyses, we asked if experiencing cultural
disfluency may be more problematic for people who value fitting in and connecting to the group
and less problematic for people who enjoy being distinct, unique and sticking out. We assessed
both values of individualism and of collectivism given that people endorse each of these sets of
values to differing degrees (Markus & Oyserman, 1989; Oyserman, 1993).
Now consider individual differences in people’s response to the unexpected, ambiguous,
or unusual. Hofstede (2011) described differences in uncertainty avoidance as a cultural axis. At
the individual level, people differ in how they respond to unusual situations. Need for cognition
was initially described by Cohen, Stotland, and Wolfe (1955) as “a need to structure relevant
situations in meaningful, integrated ways. It is a need to understand and make reasonable the
experiential world” (p. 291). The assumption was that people higher in need for cognition would
find unstructured ambiguous situations more frustrating than people lower in need for cognition.
Cohen and colleagues provided only an example item in their paper so subsequent researchers
`

117
developed scales of need for cognition (NFC; Cacioppo & Petty, 1982, Cacioppo, Petty, & Kao,
1984) and personal need for structure (PNS, Neuberg & Newsom, 1993). NFC is operationalized
with items such as “I prefer complex to simple problems” and was designed to distinguish
individuals who dispositionally tend to engage in and enjoy effortful analytic activity from those
who do not. Our exploratory prediction was that people who are high in NFC may find culturally
disfluency particularly rattling, enhancing the effects of cultural disfluency on inherence.
PNS is operationalized with items such as “I become uncomfortable when the rules in a
situation are not clear” (Neuberg & Newson, 19933). In that sense, PNS may be closer to what
Cohen and colleagues meant by a need for cognition as it assesses desire for simple structure
with clear interpretation. Some researchers (e.g., Freund, Kruglanski, & Shpitzajzen, 1985)
suggest that need for structure is contextually cued, that in ambiguous contexts in which choices
must be made, people experience a momentary rise in need for structure, implying that a need for
structure functions as a mediator rather than a moderator of cultural disfluency effects. Our
exploratory prediction is that the experience of cultural disfluency may be particularly
problematic for people who prefer simple structures and prefer not to engage in effortful
thinking.
A related, though, less used, individual difference construct is intolerance of uncertainty
(IU, Carleton, Norton, & Asmundson, 2007). IU entails fear or worry about the possibility of a
negative occurrence and is operationalized by responses to items such as “Uncertainty makes me
uneasy, anxious, or stressed” (Carleton et al., 2007). Fear of the unknown might be relevant to
responses to cultural fluency and disfluency, yielding an exploratory prediction that people
higher in IU might be more rattled by experiencing cultural disfluency and this might enhance
`

118
the effect of cultural disfluency on their sense that all’s right with the world. Though individual
differences were not our main focus, we explored each of these possibilities.
Current Studies
Our primary prediction is that experiencing cultural disfluency undermines inherence
(Studies 1, 2, 3, 4, 5, 6, 7, 8). Our secondary predictions are that undermined inherence carries
forward to undermine people’s belief that cultures have fixed essences (Studies 2, 3, 4) and that
explicit forewarning does not undermine the effect of cultural disfluency on inherence (Studies 7,
8). We explored momentary affect (Studies 1, 2) as a possible mediator, and individualism,
collectivism, NFC (Studies 1, 2), PNS and IU (pre-registered Study 8) as possible individual
difference moderators. We addressed reviewer questions as to whether the effects of cultural
fluency and disfluency on inherence were due to effects on certainty about the world or if they
would also affect certainty about the self (Study 6). Study 5: https://aspredicted.org/4c8th.pdf
and Study 8: https://aspredicted.org/ir4bj.pdf) were preregistered for hypotheses, design, and
analyses. Primes, primary and secondary dependent variables are all in the supplemental
materials.
Prior cultural fluency and disfluency researchers used a mix of peripheral (pink border on
Valentine’s Day, red border on Chinese New Year) and central (wedding photographs) cues. We
used cues likely to be central (e.g., a Valentine’s Day card) to a cultural-based associative
knowledge network (e.g., Valentine’s Day). Central cues should more stably affect experienced
cultural fluency and disfluency. We validated this prediction of stability in Study 1 by repeating
the experiment on Valentine’s Day and a week after Valentine’s Day (as detailed in the
Supplemental Materials). As manipulation checks, we examined the extent that participants rated
`

119
the products they saw as similar to expectation, traditional or appropriate for the occasion and
their immediate fluency response as reflected in product quality and attractiveness ratings.
In each study, we screened for native speakers, including only native speakers in our
analytic samples for two reasons. First, and most importantly, native speakers are clearly
members of the culture being studied. Second, a large body of evidence suggests that using a
second language decreases processing fluency (e.g. Keysar, Hayakawa, & An, 2012; Hayakawa,
Costa, Foucat, & Keysar, 2016). Hence non-native language might reduce experienced inherence
through another route, reducing the clarity of our prediction and results.
In Amazon Mechanical Turk (Mturk) studies, we looked for participants with U.S. IP
addresses who had not participated in our previous cultural fluency studies. We did so for two
reasons. First, this insured that our participants had exposure to U.S. culture. Second, this insured
that participants in each study were independent and no person participated in more than one
study to the best of our knowledge.
Scales not already available in Hebrew or Chinese were translated and back-translated by
bilingual researchers and questions screened for meaning following the standards of the
American Association of Public Opinion Research for cross-cultural survey design (Harkness et
al., 2010). We provide our full experimental materials and dependent measures in Supplemental
Materials. All studies followed the same procedures as detailed next. To make the full set of
results easier to digest, we describe our samples, the full set of procedures, and all results
together. We present by-study information in the Supplemental Materials. To maintain a full
record, we also include an exploratory study and an MTurk study in which we have concerns
about data quality in Supplemental Materials though not in the main text.
Power and Stop Rules
`

120
We determined our target sample size prior to data collection using the combined effect
size (d=.47) of Mourey and colleagues’ (2015) two-condition cultural fluency studies. G*power
analyses yielded a target sample of 102 participants (51 per condition) We verified this assumed
effect size in an exploratory study (S1, detailed in Supplemental Materials) and found a medium
cultural (dis)fluency effect (d = .50) on inherence. We used this same stop rule until we
calculated our actual effects from Studies 1 to 4. This resulted in a downward adjustment of
expected effect size to d = .40 for pre-registered Study 5, which we used to determine an a priori
sample size to achieve three-condition power of .80 and p = .05, yielding a target sample of 246
(82 per condition). In Study 6 we used the Study 1 prime and effect size (d = .38), yielding a
target sample size of 440 (220 for each dependent variable order) to attain power of .80 and p
= .05. Our subject pool yielded a smaller sample size of 332 once non-native English speakers
and repeat responders were excluded. In Study 7 we used the average effect size of Studies 1 to 6
(d = .38), power of .80 and p = .05 to calculate that 270 (90 per condition) were needed for three
conditions. In Study 8 we repeated this process using the final average effect size (d =.34) and
the effect size found in Study 2 (same stimuli, d=.34) to calculate that 339 (113 per condition)
were needed for three conditions.
Sample
Participants were adults (total N = 1,803, Mage= 30.67, SD = 7.86). They were native
speakers of English from the U.S., or native speakers of Chinese from China, or native speakers
of Hebrew from Israel. Table 2 details demographics by study. In Studies 1, 3, 5, 7, participants
were American Mturk workers paid $.40. In Study 2, they were Jewish Israeli undergraduates at
an Israeli university who received course credit as compensation. In Study 4, they were Chinese
adults from a crowdsourcing website (zbj.com) paid the equivalent (¥3). In Study 6, they were
`

121
native undergraduates at an American university who received course credit as compensation. In
Study 8, participants were Jewish Israeli adults from a crowdsourcing website (Panel4all) and
paid about $2 (8.5 NIS).
Procedure
Study procedure. Table 1 summarizes each study design. Studies, programmed in
Qualtrics, took 5 to 7 minutes to complete. We randomly assigned participants to one of two
groups (Studies 1, 2, 3, 4, 6: Cultural Match, Cultural Mismatch) or to one of three groups:
(Study 5: Cultural Match, Cultural Mismatch, Control; Studies 7, 8: Cultural Match, Cultural
Mismatch, Explicit Mismatch). In each group, participants saw cultural products relevant to the
culture being studied and rated each product for quality and attractiveness.
Rating products served two functions. It provided participants with a psychologically
meaningful task that allowed us to expose them to either culturally fluent or culturally disfluent
stimuli. The ratings themselves allowed us document that we were manipulating fluency and
disfluency. As detailed in Table 3, the products were American Valentine’s Day cards (Studies 1,
6, 7), Israeli breakfasts (Study 2, 8), a set of wedding photographs of European American
couples (shown to European Americans, Study 3), a set of wedding photographs of a Han
Chinese couple (shown to Han Chinese, Study 4), and Labor Day shopping bags (Study 5).
Across studies, the Match group saw culturally fluent versions of the products and the Mismatch
group saw culturally disfluent versions of the products.
In Study 5, People randomly assigned to the Control group saw culturally neutral
versions of the products. In Studies 7 and 8, people randomly assigned to the Explicit group saw
the same products as the Mismatch group, and were explicitly told that they would be seeing
surprising or unexpected products that might lead them to question their beliefs. For Study 7, the
`

122
instructions read: “In this study, you will be shown a sampling of Valentine’s Day card design
images with non-traditional, unexpected features. We are interested in how being exposed to
unexpected designs has consequences for feeling that everything else in life is going well. For
each one that you see, rate design quality, attractiveness, and traditionality.” For Study 8
(English translation of the Hebrew): “In this study you will be presented with unexpected
breakfast dishes with unusual ingredients. Looking at this kind of photos makes people question
things they usually take for granted.”
After the rating task, participants continued to an ostensibly unrelated second part in
which they rated how much they agreed or disagreed to a series of statements. In all studies
except Study 6, participants first presented with our primary dependent variable, the inherence
scale. Then came (in this order) the secondary dependent variable, essentialism (obtained in
Studies 2, 3, 4) and then any potential moderator or mediator (obtained in Studies 1, 2, 8). In
Study 6, we randomized the order of presentation of the statements in our primary dependent
variable (inherence) and our secondary dependent variable (self-certainty). At the end of each
study, participants rated the extent products were similar to expectation, traditional, appropriate
and completed demographics.
Inherence. We operationalized inherence (1=strongly disagree, 7=strongly agree) with
Salomon and Cimpian’s (2014) 15-item Inherence Heuristic Scale (Study 1 α=.87; Study 2 α
=.76; Study 3 α=.81; Study 4 α=.75; Study 5 α=.86; Study 6 α=.75; Study 7 α=.85; Study 8 α
=.82). We adjusted culture-bound elements to be relevant to each culture. “There are good
reasons why dollar bills are green” was translated as “There are good reasons why currency is in
different colors” (Hebrew) or “There are good reasons why 100-Yuan bills are red” (Chinese). In
Study 3, we removed two items -- “It seems right to use white for wedding dresses” and “It
`

123
seems right that black is the color associated with funerals” -- because white and black were part
of the wedding study manipulation. We substituted an attention check: “Please choose “strongly
disagree” for this question so that we know you are paying attention” for the original culture-
bound catch items.
Essentialism. We operationalized essentialism (1=strongly disagree, 7=strongly agree;
Study 2 α=.79; Study 3 α=.75; Study 4 α=.77) with an 8-item scale based on Haslam, Rothschild,
and Ernst (2000) and Chiu, Dweck, Tong, and Fu (1997) essentialism scales. Example items are:
“Traits of a culture are stable over time. They do not change much.” “Though some phenomena
can be changed, it is unlikely that the core dispositions of the world can be altered.”
Self-certainty. We operationalized self-certainty (1=strongly disagree, 7=strongly agree;
Study 6 α=.90) with a 10-item scale based on Campbell et al. (1996)’s self-concept clarity scale.
Example items are: “I am certain about the kind of person I am.” “I lack a clear sense of my
skills” (reverse-coded).
Potential Moderators. Cacioppo, Petty, and Kao’s (1984) 18-item Need-for-Cognition
scale (1= extremely uncharacteristic of me, 5=extremely characteristic of me, Study 1 α =.94)
was used in Study 1. Oyserman’s (1993) Hebrew language 6-item individualism (e.g., “I
determine my own destiny” α=.51) and 6-item collectivism (e.g., “In general, I accept the
decisions made by my group” α=.70) scales (1=strongly disagree, 7=strongly agree) were used
in Study 2. Nine items from Neuberg and Newson’s (1993) Personal Need for Structure (α=.74)
and Carleton, Norton and Asmundson’s (2007) Intolerance of Uncertainty (α=.89, 12 items) were
used in Study 8.
`

124
Potential Mediators. Thompson’s (2007) 10-item Positive (Study 1 𝛼 =.86; Study 2
𝛼 =.74) and Negative (Study 1 𝛼 =.87; Study 2 𝛼 =.78) Affect Scales (PANAS, 1=very slightly or
not at all, 7= extremely) were used in Studies 1 and 2.
Manipulation checks. Participants rated quality and attractiveness of each presented
cultural product as they viewed them. At the end of the study, participants rated the products they
saw overall for their similarity to expectation, traditionality, and appropriateness for the
occasion, we created a mean similarity to expectation score from the mean of these three items.
Table 5 details individual items used in each study.
Attention checks. The inherence scale included an attention check item and as
recommended (Meade & Craig, 2012) we dropped participants who failed it. In Studies 1 to 5,
the cultural products were linked to holidays and we asked participants what the holiday
connected to the products they saw was and when that holiday is celebrated. Following Zayas,
Pandey, and Tabak (2017) we dropped participants without cultural exposure – those who could
not name the holiday or when it was celebrated.
`

125
Results
Preliminary Analyses and Manipulation Checks
We used confirmatory factor analyses (CFA) to test the factor structure of our dependent
variables -- inherence, essentialism, and self-certainty, finding good-to-moderate fit in each
study. We present details of our CFA in Supplemental Materials (Table S4). Moreover, as
detailed in Tables 4 and 5, our manipulation checks suggest that we succeeded manipulating
cultural fluency and disfluency. Thus, participants in the Match condition rated the products that
they saw as higher in attractiveness and quality and as more similar to what they expected, more
traditional, and more appropriate than participants in the Mismatch condition. This conclusion is
supported by our single paper meta-analysis of quality and attractiveness ratings (Studies 1 to 8)
and of similarity to expectation ratings (Studies 2 to 8), both detailed in Supplemental Materials.
`

126
Cultural Disfluency Undermines Inherence
Results, detailed in Table 6, and displayed graphically in Figure 3, support our primary
prediction that cultural fluency supports and cultural disfluency undermines inherence. Across
our eight studies, participants randomly assigned to the Match Condition experienced higher
psychological inherence than participants randomly assigned to the Mismatch Condition. At the
same time, though consistent, the effect of cultural fluency and disfluency on inherence was
small-to-moderate rather than the moderate-sized effect we had predicted. We followed up in
two ways. First, we ran study-by-study sensitivity power analyses to determine the minimum
detectable effect size with power of .80 and p = .05 for each study. The results of these analyses
suggest that our found effects are smaller than what our samples are powered for (minimal
detectable effects were: Study 1 d = .40, Study 2 d =.47, Study 3 d =.52, Study 4 d =.57, Study 5
d =.43, Study 6 d = .31, Study 7 d = .40, Study 8 d = .37). Second, we conducted a single-paper
meta-analysis to obtain a more stable estimate of the effect of Cultural Match and Cultural
Mismatch on psychological inherence. To do so, we used Review Manager 5.3 software. Our
meta-analysis revealed that being exposed to cultural disfluency had a small-to-moderate sized
effect on subsequent experience of inherence, d = 0.32 (95% CI: .22, .43, see Table 6 for details).
Moreover, the non-significant test of heterogeneity, 𝜒 2
= 1.85, df = 7, p = .97, I
2
= 0%, reveals
that effects were not dependent on country or a particular operationalization of cultural fluency
and cultural disfluency. Taken together, results suggest that there is a stable, small-to-moderate,
effect of cultural fluency and disfluency on inherence.
In pre-registered Study 5 (Labor Day shopping bags), a control group (non-themed
shopping bag) made sense, allowing us to test that cultural disfluency decreases inherence rather
than cultural fluency increasing it. We used Planned Contrast Linear regression (Control = 1,
`

127
Cultural Match = 1, Cultural Mismatch = -2) which revealed a significant effect of Condition on
inherence, F(1, 257) = 6.69, p = .01, R
2
= .03. Inherence was lower among participants randomly
assigned to the Cultural Mismatch Condition (M = 4.68, SD = .90) compared to participants
randomly assigned to the other conditions (Cultural Match M = 4.99, SD = .83, Control M =
4.97, SD = .90). Inherence did not differ among participants randomly assigned to the Cultural
Match or to the Control conditions F(1, 168) = .02, p = .90, 𝜂 2
<.001.
To test our secondary prediction that explicitness does not undo the effect of cultural
disfluency, we added an explicit instruction condition in Study 7 and pre-registered Study 8. A
one-way ANOVA (Cultural Mismatch, Explicit Mismatch) revealed a significant effect of
condition on inherence in Study 7, F(2, 303) = 3.58, p = .03, 𝜂 2
= .02, and the same directional
effect, though not significant at the .05 level in Study 8, F(2, 339) = 2.81, p = .06,𝜂 2
= .02. In
both studies, participants randomly assigned to the Explicit Mismatch condition (Study 7: M =
4.78, SD = .96; Study 8: M = 4.85, SD = .89) experienced lower inherence than those assigned to
the Cultural Match condition (Study 7: M = 5.08, SD = .78; Study 8: M = 5.09, SD = .79), and
this differences was significant in Study 7, F(1, 207) = 6.01, p = .02, d = .34, and in Study 8, F(1,
223) = 4.60, p = .03, d = .29. Inherence did not differ between participants in the Explicit
Mismatch condition and those in the Cultural Mismatch condition in Study 7, F(1, 200) = .02, p
= .90, or in Study 8, F(1, 229) = .09, p = .77.
Cultural Disfluency Reduces Tendencies to Essentialize via Inherence
In testing our mediation prediction, we followed Zhao, Lynch, and Chen (2010) who state
that the presence of a significant indirect effect establishes mediation and used Process Syntax
Model 4 (Hayes, 2013) with 1,000 bootstrapped samples. Results supported our prediction that
inherence mediates the effect of cultural fluency and disfluency on essentialism, as displayed
`

128
graphically in Figure 4 and detailed in Table 7. We followed up with a single-paper meta-
analysis using Comprehensive Meta-Analysis to provide a stable estimate of this effect.
Experiencing cultural disfluency has a stable significant indirect effect (ab) on essentializing via
inherence, ab = -.16 (95% CI: -.23, -.08). This effect does not vary by sample and country, as
revealed by a non-significant between-study Q statistic, Q = .05, df = 2, p = .97, I
2
=0%.
Cultural Disfluency Reduced Feelings of Self-Certainty via Inherence
To address a reviewer question as to whether the effect of cultural disfluency on
inherence (uncertainty about the world) is also found for uncertainty about the self, in Study 6
we added uncertainty about the self as a secondary dependent variable. We found no direct effect
of condition on feelings of self-certainty, F(1, 330) = .02, p = .91. Instead, we found that cultural
disfluency reduced feelings of self-certainty through its effect on inherence, ab = -.04, SE = .02,
95% CI = [-.10, -.003]. The overall model was significant, R
2
= .04, F(2, 329) = 6.00, p = .003.
The implication is that cultural disfluency induces a sense of uncertainty about the world and this
carries over to induce a lack of certainty about the self.
`

129
Analyses of Potential Moderators and Mediators and Exploratory Follow-up Analyses
We tested three potential mediators and five potential moderators of the effect of cultural
fluency and disfluency on inherence. These null findings are detailed in Tables S1 and S2
(Supplemental Materials). Condition effects were not moderated by Need for Cognition (Study
1), by Individualism or Collectivism (Study 2), or by Perceived Need for Structure or Intolerance
of uncertainty (Study 8). Neither Positive nor Negative Affect mediated the cultural fluency and
disfluency effect (Studies 1, 2).
As a final set of exploratory supplemental analyses, we also explored whether product
quality and attractiveness or product similarity to expectation ratings affected inherence and
whether this mediated the effect of cultural fluency and disfluency. These exploratory analyses,
detailed in Table S3, focus on a final post hoc set of alternative explanations for our results.
Taken as a whole, these analyses rule out alternative explanations for cultural fluency and
disfluency effects via the effect of fluency on quality, attractiveness, and similarity to
expectations. Specifically, product quality and attractiveness ratings did not mediate the effect of
condition on inherence in six of eight studies (Studies 1, 2, 4, 6, 7, 8). Product similarity-to-
expectation ratings mediated effect of condition on inherence only in one of eight studies (Study
7).
Discussion
We start with a culture-as-situated cognition perspective on culture, which suggests that
people’s culture-based expertise means that they have tacit knowledge about how the everyday
situations in their lives are likely to unfold and that this tacit knowledge matters. That is, people
have tacit knowledge about what breakfast, shopping bags, Valentine’s cards and weddings look
like. This tacit knowledge allows them to ‘nod along’ and use gut-based heuristic strategies when
`

130
observations match implicit predictions and shifts them to rule-based systematic strategies when
they do not. We predicted and showed that when things look as one might expect, matches
between tacit cultural expectation and observation reinforce people’s sense of inherence, the
feeling that the current order is the natural and legitimate one. In contrast, mismatches open
people to the possibility that alternatives are possible, with downstream effects on essentialist
reasoning and certainty about the self.
We used everyday cultural products (Valentines, breakfasts, wedding photographs, Labor
Day shopping bags) to documents these effects. Americans who were shown Valentine’s Day
cards with grey hearts made of skulls or Labor Day shopping bags with unexpected eco-themed
designs experienced lower inherence than Americans who were shown Valentine’s Day cards
with pink hearts (Studies 1, 6, 7) or Labor Day shopping bags with expected patriotic-themed
designs (Study 5). Similarly, Israelis who were shown unexpected breakfast ingredients for Israel
such as meats and fried food experienced lower inherence than Israelis who were shown
expected breakfast ingredients such as raw vegetables and fresh cheese (Studies 2 and 8).
Chinese (Study 4) and Americans (Study 3) who were shown weddings with unexpected
elements such as a black wedding dress also reported lower inherence than those who were
shown culturally expected weddings.
We showed that experiencing match and mismatch has small-to-moderate-sized effects
on inherence and that this effect is stable across manipulations and countries (China, Israel, and
the U.S.). We showed that the effect of our cultural fluency and disfluency carries over to
cultural essentializing and self-certainty across various cultural cues, suggesting that cultural
(dis)fluency effects are robust. Our manipulation checks confirmed that we manipulated cultural
fluency and disfluency. That is, people randomly assigned to the Cultural Mismatch Condition
`

131
(e.g. saw Valentines with grey hearts made of skulls) rated the products they saw as less
traditional and less similar to what they expected than people randomly assigned to the Cultural
Match Condition (e.g., saw Valentines with pink hearts). Products that mismatched cultural
expectations were processed less fluently than those that matched cultural expectations, as
revealed by the lower ratings of these products on quality and attractiveness. Psychological
consequences of cultural disfluency remained when people were explicitly warned, were not
mediated by product ratings, and were not a function of momentary affect or of individual
differences in cultural values (individualism, collectivism) or in need for cognition, perceived
need for structure, or intolerance of uncertainty. These null results suggest that cultural fluency
and disfluency effects are not a function of these processes or of individual differences but rather
are a function of people’s automatic tendency to draw on cultural expertise to make predictions
about how everyday life will unfold.
Theoretical Implications
We document that an as yet underappreciated aspect of culture, cultural fluency and
disfluency, is psychologically consequential. It supports people’s situated and pragmatic
reasoning, in part by preserving or disrupting their sense of inherence. Our results build on an
assumption made in culture-as-situated cognition theory which is that cultural expertise makes it
easier to navigate everyday life by providing a road map of how “we” act and what “we” do. We
subtly manipulated an immediate situation to support or undermine the seeming applicability of
people culture-based road map and documented that people were sensitive to the pragmatic
implications of situations at hand.
Our results have several important theoretical implications for understanding how people
respond to disruptions in their everyday expectations. First, our results enrich the growing body
`

132
of research on psychological inherence. We showed that cultural fluency and disfluency affects
people’s endorsement of inherence, the general sense that current patterns are natural and
legitimate, and through that sense, affects essentialist reasoning about cultures and self-certainty.
While some have argued that the relationship between psychological inherence and essentialism
is specific to individualistic cultures, arguing that individualistic (inductive and causal) reasoning
is required to move from inherence to essentialism (Baron, 2014). We document effects in China
and Israel, suggesting the relationship between inherence and essentialism is not limited to
individualistic cultures.
Second, our results are relevant to research on the relationship between inherence
(Cimpian, 2015) and essentialism, the sense that social categories have fixed essences (Gelman,
2003). Our results support the Salomon and Cimpian (2014) assertion that inherence affects
essentialism. We show that essentialism is disrupted by cultural disfluency and supported by
cultural fluency via the effects of cultural fluency and disfluency on inherence. These results
matter given that essentialism is a double edged-sword. It is necessary, serving as precursor to
categorical reasoning, which allows for predictions about the world and the self (Gelman, 2003),
but also increases the likelihood of stereotyping and prejudice (Bastian & Haslam, 2006).
Third, our results support the contention of culture-as-situated cognition theory that
people are sensitive to the pragmatic implications of situational support or violation of their tacit,
automatic culture-based predictions. We show that explicit warning does not undermine the
effect of cultural (dis)fluency, implying that people experience cultural fluency and disfluency as
pragmatically useful to their reasoning about the world. When things are not as expected and
processing is difficult, one experiences the world as a less certain place. Accessible information
`

133
is used in making judgments when it is experienced as relevant to the judgment task at hand, no
matter why it is on the mind.
Fourth, our results support adding the concept of culture to our understanding of what a
situated cognition “thinking is for doing” notion means. That is, thinking is both situated and
culture-based and pragmatic. Situated cognition approaches predict and show that people use
their metacognitive experiences of fluency and disfluency in making inferences about product
quality and attractiveness, for example, rating a key as more attractive after being exposed to a
lock (Schwarz, 2015). In our studies, we document that culture-based expertise matters, affecting
people’s automatic predictions, and that when these predictions are maintained, people rate
products rate as higher in quality and in attractiveness than when these predictions are violated.
However, in our studies quality and attractiveness ratings do not affect inherence; inherence was
directly affected by cultural fluency and disfluency. The implication we draw is that prior models
are insufficient to explain the process underlying cultural fluency and disfluency and that a
broader culture-as-situated cognition model is needed to understand how culture supports
pragmatic inference.
Fifth, our results support a broadened understanding of the interface between situated
cognition and other approaches to meaning making. Our situated approach predicts that people
make a pragmatic inference when their culture-based predictions are supported -- that things are
as they ought to be, paving the way for using essentialistic reasoning and the self as a predictive
anchor. The alternative, when culture-based predictions are violated --that things do not have to
be as they are now, paves the way to be open to new possibilities, requiring that essentialist
reasoning be put aside.
`

134
Thus, our cultural fluency and disfluency model posits and shows that people’s response to the
unexpected entails increased openness to the possibility that things are not as they had assumed
them to be. As one of our reviewers helpfully noted, the unexpected can be threatening, as
highlighted by terror management theory (TMT; Greenberg, Solomon, & Pyszcynski, 1997) and
the meaning maintenance model (MMM, Heine, Proulx, & Vohs, 2006; Proulx & Heine, 2008;
Proulx & Inzlicht, 2012). TMT focuses on existential threat to meaning (for example by
considering what death does to one’s body, Greenberg et al., 1997). MMM focuses on violation
of meaningfulness, for example by switching experimenter mid-study without explanation
(Proulx & Heine, 2008), evaluating surrealistic art or considering absurdist literature (Proulx,
Heine, & Vohs, 2010), or even trying to solve problems with no clear best solution (Grieve &
Hogg, 1999). These violations increase certainty in another domain. For example, solving
problems with no clear solution increased ingroup bias in a separate task (Grieve & Hogg, 1999).
Experiencing a switch in experimenter mid-study increased certainty in one’s moral beliefs
compared to not experiencing a switch. In contrast, cultural fluency and disfluency is about the
automatic predictions people make as everyday life unfolds. A breakfast plate, a wedding
photograph, a Valentine’s Day card can hardly be construed as implying that life may have no
meaning but do shift the pragmatic inferences people draw about whether to go with current
assumptions or be open to other possibilities.
Limitations
Like any set of studies, our studies have a number of limitations. First, our focus on
adults means that we cannot make inferences about developmental patterns, which may be
important in understanding the experience of inherence. Cimpian and Steinberg (2014) show that
children demonstrate stronger reliance on inherence to make sense of their daily experiences than
`

135
adults. It is possible that our effects would be stronger with children, but our studies cannot
address this possibility. Second, our focus on modern cultures exposed to Western-style
education means that we cannot address questions of whether our effects would have been found
in premodern cultures not exposed to Western-style education. Compared to premodern
societies, modern societies experience more rapid change, so it is possible that effects would
have been larger in premodern societies. Compared to traditional, religious-based education,
which focuses on the idea of a stable, revealed truth, Western-style education focuses on the idea
that truth is contingent, what seems to be true in the moment may later need to be revised. It is
possible that our effects would be larger if we included people from societies not exposed to
Western-style educational models. Third, our focus on documenting that cultural fluency and
disfluency effects could occur means that we did not attempt to enumerate the entire population
of cultures and cultural experiences and randomly draw from them. Each of these strategies
increases statistical power and the generalizability of found effect sizes (e.g., Westfall, Kenny, &
Judd, 2014). While theoretically desirable, enumerating the population of respondents (all
residents of all modern cultures) and of dependent variables (all everyday cultural experiences)
and randomly sampling from them is not practically feasible. To address issues of
generalizability, we used on-line panels to reduce the age and social class bias that student-only
samples entail and we used a variety of everyday cultural experiences to increase the everyday
realism of our stimuli. Having said that, lacking such sampling, we cannot be sure that the small-
to-moderate effect that we found would apply to all possible experiences of cultural fluency and
disfluency.
Conclusions
`

136
Based on the assumption that thinking is situated, pragmatic, and based in cultural
experience, we articulate how people’s culture-based expertise supports their thinking, allowing
them to filter expected from unexpected situations and guiding pragmatic reasoning about the
world. Our results provide support for this broader understanding of what cultural expertise is
and highlight some potential upsides to the kind of disruptions that may occur in heterogeneous
societies. When observed reality does not match culture-based expectations, cultural disfluency
disrupts people’s sense of inherence -- that the current pattern of everyday life is the natural way
for things to be. This, in turn, undermines people’s belief in stable essences, a potentially
positive turn of events, as it may increase tolerance for others since essentializing may have the
effect of bolstering out-group stereotyping.

`

137
General Discussion

In this dissertation, I started with a meta-analysis to provide a comprehensive overview of
causal consequences of cultural mindsets – individualism, collectivism, and honor. A decade of
cumulative evidence from cultural priming research supports the notion that cultural mindsets are
universal, situated, and psychologically consequential. People across groups and societies can
use different cultural mindsets when they are brought to mind. The cultural mindset accessible in
the moment guides how people make sense of themselves, engage with others, experience
feelings, make social judgments, and process cognitive information. In the next two chapters,
using the framework of culture-as-situated-cognition theory, I identified two unique
consequences of culture for reasoning in the moment. I showed that collectivism, a universal
cultural mindset that prioritizes fitting in and connecting with others, motivates people to
construct meaning and therefore find meaning even in places where meaning might not exist.
Shifting to everyday cultural knowledge, I documented that subtle matches and mismatches
between culture-based expectations and the immediate situations influence how people make
sense of the world. Matches with expectations preserve, whereas mismatches disrupt, people’s
sense that current patterns in the world are the natural order, and this has downstream
consequences for essentialist thinking and self-certainty.
Together, the meta-analysis and the two sets of studies from three chapters support and
broaden the understanding that human reasoning is situated and culture-based. The findings
challenge a view of culture as a static entity that represents inherent differences between groups
and societies. Instead, culture’s consequences depend on the situation. Situational cues that bring
a cultural mindset to mind or interact with culture-based knowledge powerfully influence how
we process information and perceive the world around us.
`

138
References
Anaby, D., Jarus, T., & Zumbo, B. D. (2010). Psychometric evaluation of the Hebrew language
version of the Satisfaction with Life Scale. Social Indicators Research, 96(2), 267-274.
Anthony, A., & Moulding, R. (2019). Breaking the news: Belief in fake news and conspiracist
beliefs. Australian Journal of Psychology, 71(2), 154–162.
Arieli, S., & Sagiv, L. (2018). Culture and problem-solving: Congruency between the cultural
mindset of individualism versus collectivism and problem type. Journal of Experimental
Psychology: General, 147(6), 789-814.
Arieli, S., & Sagiv, L. (2018). Culture and problem-solving: Congruency between the cultural
mindset of individualism versus collectivism and problem type. Journal of Experimental
Psychology: General, 147(6), 789.
Bago, B., Rand, D. G., & Pennycook, G. (2020). Fake news, fast and slow: Deliberation reduces
belief in false (but not true) news headlines. Journal of Experimental Psychology:
General.
Bargh, J. A. (1994). The four horsemen of automaticity: Awareness, intention, efficiency, and
control in social cognition. Handbook of social cognition, 1, 1-40.
Bargh, J. A., Bond, R. N., Lombard, W. J., & Tota, M. E. (1986). The Additive Nature of
Chronic and Temporary Sources of Construct Accessibility. 10.
Baron, A. S. (2014). Is the inherence heuristic simply WEIRD?. Behavioral and Brain
Sciences, 37(05), 481-481.
Barsalou, L. W. (2016). Situated conceptualization offers a theoretical account of social priming.
Current Opinion in Psychology, 12, 6–11. https://doi.org/10.1016/j.copsyc.2016.04.009
`

139
Bastian, B., & Haslam, N. (2006). Psychological essentialism and stereotype
endorsement. Journal of Experimental Social Psychology, 42(2), 228-235.
Baumeister & Landau, 2018
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal
attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497.
Benet-Martínez, V., Leu, J., Lee, F., & Morris, M. W. (2002). Negotiating biculturalism:
Cultural frame switching in biculturals with oppositional versus compatible cultural
identities. Journal of Cross-Cultural Psychology, 33(5), 492–516.
Berry, J. W. (2003). Conceptual approaches to acculturation. American Psychological
Association.
Boothby, E. J., Smith, L. K., Clark, M. S., & Bargh, J. A. (2016). Psychological distance
moderates the amplification of shared experience. Personality and Social Psychology
Bulletin, 42(10), 1431–1444.
Borenstein, M., Cooper, H., Hedges, L., & Valentine, J. (2009). Effect sizes for continuous data.
The Handbook of Research Synthesis and Meta-Analysis, 2, 221–235.
Borenstein, M., Hedges, L. V., Higgins, J. P., & Rothstein, H. R. (2021). Introduction to meta-
analysis. John Wiley & Sons.
Boroditsky, L., Fuhrman, O., & McCormick, K. (2011). Do English and Mandarin speakers think
about time differently? Cognition, 118(1), 123–129.
Boski, P., & Youssef, K. I. (2012). Consequences of linguistic frame switching: Cognitive and
motivational shifts in bilingual Tunisians. Psychology of Language and Communication,
16(2), 143.
`

140
Brewer, M. B., & Chen, Y. R. (2007). Where (who) are collectives in collectivism? Toward
conceptual clarification of individualism and collectivism. Psychological Review, 114(1),
133-252.
Bruner, J. S., & Postman, L. (1949). On the perception of incongruity: A paradigm. Journal of
personality, 18(2), 206-223.
Burton, L., Delvecchio, E., Germani, A., & Mazzeschi, C. (2021). Individualism/collectivism
and personality in Italian and American Groups. Current Psychology, 40, 29–34.
Cacioppo, J. T., & Petty, R. E. (1982). The need for cognition. Journal of Personality and Social
Psychology, 42(1), 116-131.
Cacioppo, J. T., Petty, R. E., & Kao, C. F. (1984). The efficient assessment of need for cognition.
Journal of Personality Assessment, 48, 306–307.
Campbell, J. D., Trapnell, P. D., Heine, S. J., Katz, I. M., Lavallee, L. F., & Lehman, D. R.
(1996). Self-concept clarity: Measurement, personality correlates, and cultural
boundaries. Journal of personality and social psychology, 70(1), 141 - 156.
Carleton, R. N., Norton, M. P. J., & Asmundson, G. J. (2007). Fearing the unknown: A short
version of the Intolerance of Uncertainty Scale. Journal of anxiety disorders, 21(1), 105-
117.
Chen, S. X., & Bond, M. H. (2010). Two languages, two personalities? Examining language
effects on the expression of personality in a bilingual context. Personality and Social
Psychology Bulletin, 36(11), 1514–1528.
Chiu, C. Y., Dweck, C. S., Tong, J. Y. Y., & Fu, J. H. (1997). Implicit theories and conceptions
of morality. Journal of Personality and Social Psychology, 73, 923-940.
`

141
Chiu, C. Y., Gelfand, M. J., Harrington, J. R., Leung, A. K. Y., Liu, Z., Morris, M. W., Mu, Y.,
Shteynberg, G., Tam, K., Wan, C., & Zou, X. (2015). A conclusion, yet an opening to
enriching the normative approach of culture. Journal of Cross-Cultural
Psychology, 46(10), 1361-1371.
Chiu, C.-y., & Hong, Y.-y. (2007). Cultural processes: Basic principles. In A. W. Kruglanski &
E. T. Higgins (Eds.), Social psychology: Handbook of basic principles (p. 785–804). The
Guilford Press.
Chiu, C.-y., Leung, A. K-y., & Kwan, L. (2007). Language, cognition, and culture: Beyond the
Whorfian hypothesis. In S. Kitayama & D. Cohen (Eds.), Handbook of cultural
psychology (p. 668–688). The Guilford Press.
Chiu, L.-H. (1972). A cross-cultural comparison of cognitive styles in Chinese and American
children. International Journal of Psychology, 7(4), 235–242.
Chiu, L.-H. (1972). A cross-cultural comparison of cognitive styles in Chinese and American
children. International Journal of Psychology, 7(4), 235–242.
Cimpian, A. (2015). The inherence heuristic: generating everyday explanations. Emerging
Trends in the Social and Behavioral Sciences: An Interdisciplinary, Searchable, and
Linkable Resource. (pp. 1–15). Hoboken, NJ: John Wiley and Sons.
Cimpian, A., & Salomon, E. (2014). The inherence heuristic: an intuitive means of making sense
of the world, and a potential precursor to psychological essentialism. Behavioral and
Brain Sciences, 37(05), 461-480.
Cimpian, A., & Steinberg, O. D. (2014). The inherence heuristic across development: Systematic
differences between children’s and adults’ explanations for everyday facts. Cognitive
psychology, 75, 130-154.
`

142
Cohen, A., Stotland, E., & Wolfe, D. (1955). An experimental investigation of need for
cognition. Journal of Abnormal and Social Psychology, 51, 291-294.
Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences. Hillsdle. Erlbaum.
Conner, BE (1988).
Cross, S. E., & Madson, L. (1997). Models of the self: Self-construals and gender. Psychological
Bulletin, 122(1), 5.
Duval, S., & Tweedie, R. (2000). A nonparametric “trim and fill” method of accounting for
publication bias in meta-analysis. Journal of the American Statistical Association,
95(449), 89–98.
E. Nisbett, R., & Norenzayan, A. (2002). Culture and cognition. Stevens’ Handbook of
Experimental Psychology.
Fabrigar, L. R., & Wegener, D. T. (2016). Conceptualizing and evaluating the replication of
research results. Journal of Experimental Social Psychology, 66, 68–80.
Faragó, L., Kende, A., & Krekó, P. (2019). We only believe in news that we doctored ourselves:
The connection between partisanship and political fake news. Social Psychology, 51, 77–
90.
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(1640), 1279–1285.
Fisher, Z., & Tipton, E. (2015). robumeta: An R-package for robust variance estimation in meta-
analysis. ArXiv Preprint ArXiv:1503.02220.
`

143
Förster, J., & Liberman, N. (2007). Knowledge activation. In A. W. Kruglanski & E. T. Higgins
(Eds.), Social psychology: Handbook of basic principles (p. 201–231). The Guilford
Press.
Francis, G. (2012). Too good to be true: Publication bias in two prominent studies from
experimental psychology. Psychonomic Bulletin & Review, 19(2), 151–156.
Frankfurt, H. G., & Wilson, G. (2005). On bullshit (Vol. 8). Princeton University Press
Princeton, NJ.
Frederick, S. (2005). Cognitive reflection and decision making. Journal of Economic
Perspectives, 19(4), 25-42.
Freund, T., Kruglanski, A. W., & Shpitzajzen, A. (1985). The freezing and unfreezing of
impressional primacy: Effects of the need for structure and the fear of invalidity.
Personality and Social Psychology Bulletin, 11(4), 479-487.
Fu, J. H.-Y., Morris, M. W., & Hong, Y. Y. (2015). A transformative taste of home: Home
culture primes foster expatriates’ adjustment through bolstering relational security.
Journal of Experimental Social Psychology, 59, 24–31.
Galinsky, A. D., Magee, J. C., Inesi, M. E., & Gruenfeld, D. H. (2006). Power and perspectives
not taken. Psychological Science, 17(12), 1068–1074.
Geertz, C. (1984). Culture and social change: The Indonesian case. Man, 511–532.
Gelman, A. (2006). Prior distributions for variance parameters in hierarchical models (comment
on article by Browne and Draper). Bayesian Analysis, 1(3), 515–534.
Gelman, S. A. (1999). The role of essentialism in children's concepts. Advances in child
development and behavior, 27, 55-98.
`

144
Gelman, S. A. (2003). The essential child: origins of essentialism in everyday thought. New
York, NY: Oxford University Press.
Gelman, S. A., & Diesendruck, G. (1999). What’s in a concept? Context, variability, and
psychological essentialism. In I. E. Sigel (Ed.), Development of mental representation:
theories and applications (pp. 87-111). Mahwah, NJ: Erlbaum.
Gelman, S. A., & Echelbarger, M. E. (2019). Children, Object Value, and Persuasion. Journal of
Consumer Psychology, 29(2), 309-327.
Gharpure, R. (2020). Knowledge and Practices Regarding Safe Household Cleaning and
Disinfection for COVID-19 Prevention—United States, May 2020. MMWR. Morbidity
and Mortality Weekly Report, 69.
Greenberg, J., Solomon, S., & Pyszczynski, T. (1997). Terror management theory of self-esteem
and cultural worldviews: Empirical assessments and conceptual refinements. Advances in
Experimental Social Psychology, 29, 61-139.
Greifeneder, R., Jaffe, M., Newman, E., & Schwarz, N. (Eds.). (2020). The psychology of fake
news: Accepting, sharing, and correcting misinformation. Routledge.
Grieve, P. G., & Hogg, M. A. (1999). Subjective uncertainty and intergroup discrimination in the
minimal group situation. Personality and Social Psychology Bulletin, 25(8), 926-940.
Gronau, Q. F., Van Erp, S., Heck, D. W., Cesario, J., Jonas, K. J., & Wagenmakers, E.-J. (2017).
A Bayesian model-averaged meta-analysis of the power pose effect with informed and
default priors: The case of felt power. Comprehensive Results in Social Psychology, 2(1),
123–138. https://doi.org/10.1080/23743603.2017.1326760
`

145
Guan, M., Lee, F., & Cole, E. R. (2012). Complexity of culture: The role of identity and context
in bicultural individuals’ body ideals. Cultural Diversity and Ethnic Minority Psychology,
18(3), 247.
Gudykunst, W. B., Matsumoto, Y., Ting-Toomey, S., Nishida, T., Kim, K., & Heyman, S.
(1996). The influence of cultural individualism-collectivism, self construals, and
individual values on communication styles across cultures. Human Communication
Research, 22(4), 510–543.
Haberstroh, S., Oyserman, D., Schwarz, N., Kühnen, U., & Ji, L.-J. (2002). Is the interdependent
self more sensitive to question context than the independent self? Self-construal and the
observation of conversational norms. Journal of Experimental Social Psychology, 38(3),
323–329.
Hall, E. T. (1976). Beyond culture. Garden city. NY: Anchor.
Hamamura, T. (2012). Are cultures becoming individualistic? A cross-temporal comparison of
individualism–collectivism in the United States and Japan. Personality and Social
Psychology Review, 16(1), 3–24.
Hamilton, D. L., & Trolier, T. K. 1986. Stereotypes and stereotyping: An overview of the
cognitive approach. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination,
and racism (pp. 127-158). Orlando, FL: Academic Press.
Harkness, J. A., Braun, M., Edwards, B., Johnson, T. P., & Lyberg, L. E. (2010). Survey
methods in multicultural, multinational, and multiregional contexts. Hoboken, New
Jersey: John Wiley & Sons.
Haslam, N., Rothschild, L., & Ernst, D. (2000). Essentialist beliefs about social
categories. British Journal of Social Psychology, 39(1), 113-127.
`

146
Hauser, D., Paolacci, G., & Chandler, J. (2019). Common Concerns with MTurk as a Participant
Pool: Evidence and Solutions, Chapter 17 in F. R. Kardes, P. M. Herr, and N. Schwarz
(Eds.) Handbook of Research Methods in Consumer Psychology. Routledge.
Hayakawa, S., Costa, A., Foucart, A., & Keysar, B. (2016). Using a foreign language changes
our choices. Trends in Cognitive Sciences, 20(11), 791–793.
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: a
regression-based approach. New York, NY: Guilford Press.
Hedges, L. V., Tipton, E., & Johnson, M. C. (2010). Robust variance estimation in meta-
regression with dependent effect size estimates. Research Synthesis Methods, 1(1), 39–
65.
Heine, S. J., Proulx, T., & Vohs, K. D. (2006). The meaning maintenance model: On the
coherence of social motivations. Personality and social psychology review, 10(2), 88-
110.
Henrich, J. (2020). The weirdest people in the world: How the west became psychologically
peculiar and particularly prosperous. Farrar, Straus, and Giroux.
Higgins, E. T. (1996). Activation: Accessibility, and salience. Social Psychology: Handbook of
Basic Principles, 133–168.
Hofstede, G. (1983). National cultures in four dimensions: A research-based theory of cultural
differences among nations. International Studies of Management & Organization, 13(1–
2), 46–74.
Hofstede, G. (2011). Dimensionalizing Cultures: The Hofstede Model in Context. Online
Readings in Psychology and Culture, 2(1). https://doi.org/10.9707/2307-0919.1014
Hong, Y. (2016). Multicultural identity processes. Current Opinion in Psychology, 6.
`

147
Hong, Y.-y. (2009). A dynamic constructivist approach to culture: Moving from describing
culture to explaining culture. In R. S. Wyer, C.-y. Chiu, & Y.-y. Hong
(Eds.), Understanding culture: Theory, research, and application (p. 3–23). Psychology
Press.
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), 1-55.
IBIS World. (2020). Psychic Services Industry in the US - Market Research Report.
https://www.ibisworld.com/united-states/market-research-reports/psychic-services-
industry/
IJzerman, H., & Cohen, D. (2011). Grounding cultural syndromes: Body comportment and
values in honor and dignity cultures. European Journal of Social Psychology, 41(4), 456–
467.
Ishii, K., Reyes, J. A., & Kitayama, S. (2003). Spontaneous attention to word content versus
emotional tone: Differences among three cultures. Psychological science, 14(1), 39-46.
John, O. P., & Srivastava, S. (1999). The Big Five trait taxonomy: History, measurement, and
theoretical perspectives. Handbook of Personality: Theory and Research, 2(1999), 102–
138.
Johnson, J. T., Cain, L. M., Falke, T. L., Hayman, J., & Perillo, E. (1985). The" Barnum effect"
revisited: Cognitive and motivational factors in the acceptance of personality
descriptions. Journal of Personality and Social Psychology, 49(5), 1378.
Kahan, D. M. (2012). Ideology, motivated reasoning, and cognitive reflection: An experimental
study. Judgment and Decision Making, 8, 407–24.
`

148
Kahneman, D., & Frederick, S. (2005). A model of heuristic judgment. KJ Holyoak & RG
Morrison (Eds.), The Cambridge handbook of thinking and reasoning (pp. 267-293).
New York, NY: Cambridge University Press.
Keysar, B., Hayakawa, S. L., & An, S. G. (2012). The foreign-language effect: Thinking in a
foreign tongue reduces decision biases. Psychological science, 23(6), 661-668.
Kim, U. E., & Berry, J. W. (1993). Indigenous psychologies: Research and experience in
cultural context. Sage Publications, Inc.
Kitayama, S., & Imada, T. (2010). Implicit independence and interdependence. In B. Mesquita,
L.F. Barrett, & E.R. Smith (Eds), The mind in context (pp. 174-200). New York: Guilford
Press.
Kitayama, S., Karasawa, M., Grossmann, I., Na, J., Varnum, M. E. W., & Nisbett, R. (2019).
East-West Differences in Cognitive Style and Social Orientation: Are They Real?
[Preprint]. PsyArXiv. https://doi.org/10.31234/osf.io/c57ep
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(2), 236.
Kruschke, J. K., & Liddell, T. M. (2018). The Bayesian New Statistics: Hypothesis testing,
estimation, meta-analysis, and power analysis from a Bayesian perspective. Psychonomic
Bulletin & Review, 25(1), 178–206. https://doi.org/10.3758/s13423-016-1221-4
Kuhn, M. H., & McPartland, T. S. (1954). An empirical investigation of self-attitudes. American
Sociological Review, 19(1), 68–76.
`

149
Kühnen, U., & Oyserman, D. (2002). Thinking about the self influences thinking in general:
Cognitive consequences of salient self-concept. Journal of Experimental Social
Psychology, 38(5), 492-499.
Kühnen, U., & Oyserman, D. (2002). Thinking about the self influences thinking in general:
Cognitive consequences of salient self-concept. Journal of Experimental Social
Psychology, 38(5), 492–499.
Lee, A. Y., Aaker, J. L., & Gardner, W. L. (2000). The pleasures and pains of distinct self-
construals: The role of interdependence in regulatory focus. Journal of Personality and
Social Psychology, 78(6), 1122.
Lehman, D. R., Chiu, C., & Schaller, M. (2004). Psychology and culture. Annu. Rev. Psychol.,
55, 689–714.
Leung, A. K.-Y., & Cohen, D. (2011). Within- and between-culture variation: Individual
differences and the cultural logics of honor, face, and dignity cultures. Journal of
Personality and Social Psychology, 100(3), 507–526. https://doi.org/10.1037/a0022151
Ma, D. (2017). Data from Chinese Online Social Survey. Available at
http://www.cnsda.org/index.php?r=projects/view&id=69084413.
Majima, Y. (2015). Belief in pseudoscience, cognitive style and science literacy. Applied
Cognitive Psychology, 29(4), 552–559.
Markus, H. R., & Kitayama, S. (1994). A collective fear of the collective: Implications for selves
and theories of selves. Personality and Social Psychology Bulletin, 20(5), 568-579.
Markus, H., & Oyserman, D. (1989). Gender and thought: The role of the self-concept. In
Gender and thought: Psychological perspectives (pp. 100–127). Springer.
`

150
Markus, H., & Oyserman, D. (1989). Gender and thought: The role of the self-concept. In M.
Crawford and M. Gentry (Eds.) Gender and thought: Psychological perspectives (pp.
100-127). Springer, New York, NY.
Markus, Hazel R., & Kitayama, S. (1991). Culture and the self: Implications for cognition,
emotion, and motivation. Psychological Review, 98(2), 224.
Markus, Hazel Rose, & Kitayama, S. (2010). Cultures and Selves: A Cycle of Mutual
Constitution. Perspectives on Psychological Science, 5(4), 420–430.
https://doi.org/10.1177/1745691610375557
Markus, Hazel Rose, Kitayama, S., & Heiman, R. J. (1996). Culture and" basic" psychological
principles.
Meade, A. W., & Craig, S. B. (2012). Identifying careless responses in survey data.
Psychological Methods, 17(3), 437–455.
Medin, D. L., & Ortony, A. (1989). Psychological essentialism. Similarity and Analogical
Reasoning, 179, 195.
Merton, R. K. (1948). The self-fulfilling prophecy. The Antioch Review, 8(2), 193-210.
Mesoudi, A., & Thornton, A. (2018). What is cumulative cultural evolution? Proceedings of the
Royal Society B: Biological Sciences, 285(1880), 20180712.
Mok, A., & Morris, M. W. (2009). Cultural chameleons and iconoclasts: Assimilation and
reactance to cultural cues in biculturals’ expressed personalities as a function of identity
conflict. Journal of Experimental Social Psychology, 45(4), 884–889.
Monga, A. B., & John, D. R. (2007). Cultural differences in brand extension evaluation: The
influence of analytic versus holistic thinking. Journal of Consumer Research, 33(4), 529–
536.
`

151
Morey, R. D., Rouder, J. N., Jamil, T., & Morey, M. R. D. (2015). Package ‘bayesfactor.’ URLh
Http://Cran/r-Projectorg/Web/Packages/BayesFactor/BayesFactor Pdf i (Accessed 1006
15).
Morris, M. W., & Peng, K. (1994). Culture and cause: American and Chinese attributions for
social and physical events. Journal of Personality and Social Psychology, 67(6), 949.
Mosquera, P. M. R., Manstead, A. S., & Fischer, A. H. (2002). Honor in the Mediterranean and
northern Europe. Journal of Cross-Cultural Psychology, 33(1), 16–36.
Mourey, J. A., Lam, B. C., & Oyserman, D. (2015). Consequences of cultural fluency. Social
Cognition, 33(4), 308-xx.
Munro, G. D., & Stansbury, J. A. (2009). The dark side of self-affirmation: Confirmation bias
and illusory correlation in response to threatening information. Personality and Social
Psychology Bulletin, 35(9), 1143–1153.
Murphy, G., Loftus, E. F., Grady, R. H., Levine, L. J., & Greene, C. M. (2019). False Memories
for Fake News During Ireland’s Abortion Referendum. Psychological Science, 30(10),
1449–1459.
Neuberg, S. L., & Newsom, J. T. (1993). Personal need for structure: Individual differences in
the desire for simpler structure. Journal of Personality and Social Psychology, 65(1),
113-131.
Ng, T. K., Ng, S. H., & Ye, S. (2016). Assimilation and contrast effects of culture priming
among Hong Kong Chinese: The moderating roles of dual cultural selves. Journal of
Cross-Cultural Psychology, 47(4), 540–557.
Nisbett, R. (2004). The geography of thought: How Asians and Westerners think differently...
and why. Simon and Schuster.
`

152
Nisbett, R. E. (2018). Culture of honor: The psychology of violence in the South. Routledge.
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.
Osiurak, F., & Reynaud, E. (2020). The elephant in the room: What matters cognitively in
cumulative technological culture. Behavioral and Brain Sciences, 1–57.
Oyserman, D. (1993). The lens of personhood: Viewing the self and others in a multicultural
society. Journal of Personality and Social Psychology, 65(5), 993-xx.
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. (2015). Culture as situated cognition. Emerging Trends in the Social and
Behavioral Sciences: An Interdisciplinary, Searchable, and Linkable Resource, 1–20.
Oyserman, D. (2016). What does a priming perspective reveal about culture: Culture-as-situated
cognition. Current Opinion in Psychology, 12, 94-99.
Oyserman, D. (2017). Culture Three Ways: culture and subcultures within countries. Annual
Review of Psychology, 68, 435-463.
Oyserman, D. (2019). Cultural fluency, mindlessness, and gullibility. In J.P. Forgas & R. F.
Baumeister (Eds.), The social psychology of gullibility: Fake news, conspiracy theories,
and irrational beliefs (Sydney Symposium of Social Psychology Series). pp. 255-278.
Routledge/ Psychology Press.
Oyserman, D. (2019). The essentialized self: Implications for motivation and self-regulation.
Journal of Consumer Psychology.
`

153
Oyserman, D. & Dawson, A. (2020). Your fake news, our facts: Identity-based motivation
shapes what we believe, share, and accept. Chapter 10 in Greifeneder, R., Jaffé, M.,
Newman, E.J., & Schwarz, N. (Eds.). The psychology of fake news: Accepting, sharing,
and correcting misinformation. Psychology Press.
Oyserman, D. & Yan. V. X. (2018). Making meaning: a culture-as-situated cognition approach
to the consequences of cultural fluency and disfluency. In S. Kitayama & D. Cohen
(Eds.), Handbook of Cultural Psychology. Pp 536-565. NY: Guilford Press.
Oyserman, D., & Lee, S. W. (2008). Does culture influence what and how we think? Effects of
priming individualism and collectivism. Psychological Bulletin, 134(2), 311.
Oyserman, D., & Uskul, A. K. (2008). Individualism and collectivism: Societal-level processes
with implications for individual-level and society-level outcomes. In F. van de Vijver, D.
van Hemert, and Y. Poortinga (Eds.) Multilevel analysis of individuals and cultures (pp.
145-173). Erlbaum.
Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and
collectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological
Bulletin, 128(1), 3-72.
Oyserman, D., Novin, S., Flinkenflögel, N., & Krabbendam, L. (2014). Integrating culture-as-
situated-cognition and neuroscience prediction models. Culture and Brain, 2(1), 1-26.
Paolacci, G., & Chandler, J. (2014). Inside the Turk: Understanding Mechanical Turk as a
participant pool. Current Directions in Psychological Science, 23(3), 184-188.
Park, C. L., & George, L. S. (2018). Lab-and field-based approaches to meaning threats and
restoration: Convergences and divergences. Review of General Psychology, 22(1), 73-84.
`

154
Pashler, H., Coburn, N., & Harris, C. R. (2012). Priming of social distance? Failure to replicate
effects on social and food judgments. PloS One, 7(8), e42510.
Pennycook, G., Cheyne, J. A., Barr, N., Koehler, D. J., & Fugelsang, J. A. (2015). On the
reception and detection of pseudo-profound bullshit. Judgment and Decision Making,
10(6), 549–563.
Pennycook, G., McPhetres, J., Zhang, Y., Lu, J. G., & Rand, D. G. (2020). Fighting COVID-19
Misinformation on Social Media: Experimental Evidence for a Scalable Accuracy-Nudge
Intervention. Psychological Science, 0956797620939054.
Petrocelli, J. V., Martin, J. L., & Li, W. Y. (2010). Shaping behavior through malleable self-
perceptions: A test of the forced-agreement scale effect (FASE). Journal of Research in
Personality, 44(2), 213–221.
Pew Research Center (2018). FactTank News in the Numbers downloaded 2/20/2021 from
https://www.pewresearch.org/fact-tank/2018/10/01/new-age-beliefs-common-among-
both-religious-and-nonreligious-americans/
Protzko, J., & Schooler, J. W. (2017). Decline effects: Types, mechanisms, and personal
reflections. In S. O. Lilienfeld & I. D. Waldman (Eds.), Psychological science under
scrutiny: Recent challenges and proposed solutions (p. 85–107). Wiley
Blackwell. https://doi.org/10.1002/9781119095910.ch6
Proulx, T., & Heine, S. J. (2008). The case of the transmogrifying experimenter: Affirmation of a
moral schema following implicit change detection. Psychological Science, 19(12), 1294-
1300.
Proulx, T., & Inzlicht, M. (2012). The five “A” s of meaning maintenance: Finding meaning in
the theories of sense-making. Psychological Inquiry, 23(4), 317-335.
`

155
Proulx, T., Heine, S. J., & Vohs, K. D. (2010). When is the unfamiliar the uncanny? Meaning
affirmation after exposure to absurdist literature, humor, and art. Personality and Social
Psychology Bulletin, 36(6), 817-829.
Pustejovsky, J. (2016). clubSandwich: Cluster-Robust (Sandwich) Variance Estimators with
Small-Sample Corrections. R package version 0.2. 1.9000 [Computer software].
Qin, A. (2017, July 22). When Young Chinese Ask, ‘What’s Your Sign?’ They Don’t Mean
Dragon or Rat. The New York Times.
https://www.nytimes.com/2017/07/22/world/asia/china-western-astrology.html
Reber, R., Schwarz, N., & Winkielman, P. (2004). Processing fluency and aesthetic pleasure: Is
beauty in the perceiver's processing experience? Personality and Social Psychology
Review, 8(4), 364-382.
Rhodes, M. (2013). How two intuitive theories shape the development of social
categorization. Child Development Perspectives, 7(1), 12-16.
Risen, J. L. (2016). Believing what we do not believe: Acquiescence to superstitious beliefs and
other powerful intuitions. Psychological Review, 123(2), 182.
Romano, A., Balliet, D., Yamagishi, T., & Liu, J. H. (2017). Parochial trust and cooperation
across 17 societies. Proceedings of the National Academy of Sciences, 114(48), 12702-
12707.
Rosenthal, R. (1995). Writing meta-analytic reviews. Psychological Bulletin, 118(2), 183.
Salomon, E., & Cimpian, A. (2014). The inherence heuristic as a source of essentialist
thought. Personality and Social Psychology Bulletin, 40(10), 1297-1315.
`

156
San Martin, A., Sinaceur, M., Madi, A., Tompson, S., Maddux, W. W., & Kitayama, S. (2018).
Self-assertive interdependence in Arab culture. Nature Human Behaviour, 2(11), 830–
837.
Sanchez-Burks, J., Lee, F., Choi, I., Nisbett, R., Zhao, S., & Koo, J. (2003). Conversing across
cultures: East-West communication styles in work and nonwork contexts. Journal of
Personality and Social Psychology, 85(2), 363.
Scandura, E., & Williams, E. (2007) The politics of being fair. In Neider, L. L., & Schriesheim,
C. A. (Eds.). Research in Management International Perspectives (pp. 127 – 149).
Charlotte, NC: IAP
Scheibehenne, B., Gronau, Q. F., Jamil, T., & Wagenmakers, E.-J. (2017). Fixed or random? A
resolution through model averaging: Reply to Carlsson, Schimmack, Williams, and
Bürkner (2017). Psychological Science, 28(11), 1698–1701.
Schwarz, N (2015). Metacognition. In M. Mikulincer, P.R. Shaver, E. Borgida, & Bargh, J. A.
(Eds.), APA Handbook of Personality and Social Psychology: Attitudes and Social
Cognition (pp. 203-229). Washington, DC: APA
Schwarz, N. (2004). Meta-cognitive experiences in consumer judgment and decision making.
Journal of Consumer Psychology, 14, 332–348.
Shafa, S., Harinck, F., Ellemers, N., & Beersma, B. (2015). Regulating honor in the face of
insults. International Journal of Intercultural Relations, 47, 158–174.
Shanks, D. R., Newell, B. R., Lee, E. H., Balakrishnan, D., Ekelund, L., Cenac, Z., Kavvadia, F.,
& Moore, C. (2013). Priming intelligent behavior: An elusive phenomenon. PloS One,
8(4), e56515.
`

157
Shteynberg, G. (2015). Shared Attention at the Origin: On the Psychological Power of
Descriptive Norms. Journal of Cross-Cultural Psychology, 46(10), 1245-1251.
Simmonds, M. (2015). Quantifying the risk of error when interpreting funnel plots. Systematic
Reviews, 4(1), 1–7.
Singelis, T. M. (1994). The measurement of independent and interdependent self-construals.
Personality and Social Psychology Bulletin, 20(5), 580–591.
Singelis, T. M., & Brown, W. J. (1995). Culture, self, and collectivist communication: Linking
culture to individual behavior. Human Communication Research, 21(3), 354–389.
Smith, P. B., Vignoles, V. L., Becker, M., Owe, E., Easterbrook, M. J., Brown, R., Bourguignon,
D., Garðarsdóttir, R. B., Kreuzbauer, R., & Cendales Ayala, B. (2016). Individual and
culture-level components of survey response styles: A multi-level analysis using cultural
models of selfhood. International Journal of Psychology, 51(6), 453–463.
Snyder, M. (1984). When belief creates reality. Advances in Experimental Social Psychology, 18,
247- 305.
Spencer-Rodgers, J., & Peng, K. (Eds.). (2017). The psychological and cultural foundations of
East Asian cognition: Contradiction, change, and holism. Oxford University Press.
Spencer, S. J., Zanna, M. P., & Fong, G. T. (2005). Establishing a causal chain: Why
experiments are often more effective than mediational analyses in examining
psychological processes. Journal of Personality and Social Psychology, 89(6), 845.
Stanley, T. D., & Doucouliagos, H. (2014). Meta-regression approximations to reduce
publication selection bias. Research Synthesis Methods, 5(1), 60–78.
`

158
Steen, J., Loeys, T., Moerkerke, B., & Vansteelandt, S. (2017). Medflex: an R package for
flexible mediation analysis using natural effect models. Journal of Statistical
Software, 76(11).
Sterne, J. A., & Egger, M. (2005). Regression methods to detect publication and other bias in
meta-analysis. Publication Bias in Meta-Analysis: Prevention, Assessment and
Adjustments, 99–110.
Talhelm, T. (2018). Hong Kong liberals are WEIRD: Analytic thought increases support for
liberal policies. Personality and Social Psychology Bulletin, 44(5), 717–728.
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(2), 250–267.
Tanner-Smith, E. E., & Tipton, E. (2014). Robust variance estimation with dependent effect
sizes: Practical considerations including a software tutorial in Stata and spss: Robust
variance estimation. Research Synthesis Methods, 5(1), 13–30.
https://doi.org/10.1002/jrsm.1091
Tanner-Smith, E. E., Tipton, E., & Polanin, J. R. (2016). Handling Complex Meta-analytic Data
Structures Using Robust Variance Estimates: A Tutorial in R. Journal of Developmental
and Life-Course Criminology, 2(1), 85–112. https://doi.org/10.1007/s40865-016-0026-5
The Economist/YouGov Poll (February 13 - 16, 2021). Retrieved 2/20/2021 from
https://docs.cdn.yougov.com/pjjwlxqij4/econToplines.pdf The New York Times. (2017,
July 22). When Young Chinese Ask, ‘What’s Your Sign?’ They Don’t Mean Dragon or
Rat. https://www.nytimes.com/2017/07/22/world/asia/china-western-astrology.html
`

159
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(2), 227-242.
Tipton, E. (2015). Small sample adjustments for robust variance estimation with meta-
regression. Psychological Methods, 20(3), 375.
Toplak, M. E., West, R. F., & Stanovich, K. E. (2014). Assessing miserly information
processing: An expansion of the Cognitive Reflection Test. Thinking & Reasoning, 20(2),
147-168.
Triandis, H. C. (1964). Cultural influences upon cognitive processes. Advances in Experimental
Social Psychology, 1, 1–48.
Triandis, H. C. (1995). Individualism And Collectivism. 276. New York: Westview Press.
Triandis, H. C. (2007). Culture and psychology: A history of the study of their relationship. In S.
Kitayama & D. Cohen (Eds.), Handbook of Cultural Psychology (pp. 59-76). New York,
NY: Guilford Press.
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.
Uskul, A. K., Cross, S. E., Günsoy, C., & Gul, P. (2019). Cultures of honor. In D. Cohen & S.
Kitayama (Eds.), Handbook of cultural psychology (p. 793–821). The Guilford Press.
Van Erp, S., Verhagen, J., Grasman, R. P. P. P., & Wagenmakers, E.-J. (2017). Estimates of
Between-Study Heterogeneity for 705 Meta-Analyses Reported in Psychological Bulletin
From 1990–2013. Journal of Open Psychology Data, 5(1), 4.
https://doi.org/10.5334/jopd.33
`

160
Varnum, M. E. W., Grossmann, I., Kitayama, S., & Nisbett, R. E. (2010). The Origin of Cultural
Differences in Cognition: The Social Orientation Hypothesis. Current Directions in
Psychological Science, 19(1), 9–13. https://doi.org/10.1177/0963721409359301
Varnum, M. E., Grossmann, I., Kitayama, S., & Nisbett, R. E. (2010). The origin of cultural
differences in cognition: The social orientation hypothesis. Current directions in
psychological science, 19(1), 9-13.
Vignoles, V. L., Owe, E., Becker, M., Smith, P. B., Easterbrook, M. J., Brown, R., González, R.,
Didier, N., Carrasco, D., Cadena, M. P., Lay, S., Schwartz, S. J., Des Rosiers, S. E.,
Villamar, J. A., Gavreliuc, A., Zinkeng, M., Kreuzbauer, R., Baguma, P., Martin, M., …
Bond, M. H. (2016). Beyond the ‘east–west’ dichotomy: Global variation in cultural
models of selfhood. Journal of Experimental Psychology: General, 145(8), 966–1000.
https://doi.org/10.1037/xge0000175
von Hippel, W. W. von Hippel, F. A. & Suddendorf, T. (in press) Human Evolution. Chapter 1 in
Van Lange, Kruglanski and Higgins, Social Psychology: Basic Principles. New York:
Guilford.
Wason, P. C. (1960). On the failure to eliminate hypotheses in a conceptual task. Quarterly
Journal of Experimental Psychology, 12, 129–140.
Weingarten, E., Chen, Q., McAdams, M., Yi, J., Hepler, J., & Albarracín, D. (2016). From
primed concepts to action: A meta-analysis of the behavioral effects of incidentally
presented words. Psychological bulletin, 142(5), 472.
Westfall, J., Kenny, D. A., & Judd, C. M. (2014). Statistical power and optimal design in
experiments in which samples of participants respond to samples of stimuli. Journal of
Experimental Psychology: General, 143(5), 2020-2045.
`

161
Whitson, J. A., & Galinsky, A. D. (2008). Lacking control increases illusory pattern perception.
Science, 322(5898), 115–117.
Whitson, J. A., Galinsky, A. D., & Kay, A. (2015). The emotional roots of conspiratorial
perceptions, system justification, and belief in the paranormal. Journal of Experimental
Social Psychology, 56, 89–95.
Williams, M. J., & Eberhardt, J. L. (2008). Biological conceptions of race and the motivation to
cross racial boundaries. Journal of Personality and Social Psychology, 94(6), 1033-xx.
Winkielman, P., Huber, D. E., Kavanagh, L., & Schwarz, N. (2012). Fluency of consistency:
When thoughts fit nicely and flow smoothly. In B. Gawronski & F. Strack (Eds.),
Cognitive Consistency: A Fundamental Principle in Social Cognition (pp. 89–111). New
York, NY: Guilford Press
Winkielman, P., Schwarz, N., Fazendeiro, T, & Reber, R. (2003). The hedonic marking of
processing fluency: Implications for evaluative judgment. In J. Musch & K.C. Klauer
(eds.), The psychology of evaluation: Affective processes in cognition and emotion (pp.
189-217). Mahwah, NJ: Erlbaum.
Wolgast, A., & Oyserman, D. (2019). Seeing what other people see: Accessible cultural mindset
affects perspective-taking. Culture and Brain, 1–20.
Wu, S., & Keysar, B. (2007). The effect of culture on perspective taking. Psychological Science,
18(7), 600–606.
Yamagishi, T. (2017). Individualism-collectivism, the rule of law, and general trust. In P. A. M.
Van Lange, B. Rockenbach, & T. Yamagishi (Eds.), Series in human cooperation. Trust
in social dilemmas (p. 197–214). Oxford University Press.
`

162
Yamagishi, T., Kikuchi, M., & Kosugi, M. (1999). Trust, gullibility, and social intelligence.
Asian Journal of Social Psychology, 2(1), 145-161.Zellers, R., Holtzman, A., Rashkin,
H., Bisk, Y., Farhadi, A., Roesner, F., & Choi, Y. (2019). Defending against neural fake
news. Advances in Neural Information Processing Systems, 9051–9062.
Zayas, V., Pandey, G., & Tabak, J. (2017). Red roses and gift chocolates are judged more
positively in the U.S. near Valentine’s Day: evidence of naturally occurring cultural
priming. Frontiers in Psychology, 8:355.
Zhao, X., Lynch Jr, J. G., & Chen, Q. (2010). Reconsidering Baron and Kenny: Myths and truths
about mediation analysis. Journal of Consumer Research, 37(2), 197-206.
Zhou, X., He, L., Yang, Q., Lao, J., & Baumeister, R. F. (2012). Control deprivation and styles
of thinking. Journal of Personality and Social Psychology, 102(3), 460.

`

163
Appendices
Appendix A: Chapter 1 Tables and Figures

Table 1. Predictions for Cultural Priming Effects following Three Ways of Conceptualizing
Cultures
Question Prediction following the conceptualization of …
Culture-as-society Culture-as-syndromes Culture-as-situated-cognition
What activates a
cultural mindset?
Societal icons A component of
cultural mindset
Probabilistically determined
by co-occurrence of elements
Is everyone
prime-able?
No No Yes
Who can be
primed?
People from societies
with this cultural
mindset; Biculturals
Cultural syndrome
groups
Everyone

`

164
Table 2. General Predictions of Individualism versus Collectivism Effects
Domain Predicted effect of priming…
Individualism Collectivism
Self-concept Self as bounded and separate
from others, independence, focus
on self, standing out
Self as interconnected and embedded in
relationships, interdependence,
attending to others, fitting in
Values, goals,
and attitudes
Individualistic values, equity,
self-enhancement, internal locus
of control, promotion,
prioritization of personal
achievement and personal goals
Collectivistic values, equality,
prosocial attitudes and behavior, group
harmony, external locus of control,
prevention, prioritization of obligations
and group goals
Attribution A focus on internal factors (e.g.,
trait) and dispositional features
A focus on external factors (e.g.,
circumstances) and situational forces
Cognitive style Narrow focus in visual attention,
field independence, contrasting
mental procedures, greater
temporal distance, analytic
reasoning
Focus on relational and contextual
information in visual attention, field
dependence, connecting mental
procedures, less temporal distance,
holistic reasoning, dialectical thinking
Emotionality Socially disengaging emotion,
More weight on personal affect
and achievement
Socially engaging emotion, More
weight on relationships and others’
outcomes

`

165
Table 3. Moderator levels and Criteria
Moderator Level Criteria
Experimental
comparison
Individualism vs. collectivism The conditions from which the effect
size is computed Individualism vs. control
Collectivism vs. control
Honor vs. control
Region North America: U.S.A., Canada; Where the experiment was
conducted
Northern & Western Europe:
U.K., Austria, France, Germany,
Netherlands, Norway;

Southern & Eastern: Greece,
Poland;

Asia: Mainland China, Hong Kong,
Taiwan, Japan, Korea, Singapore

Latin America: Brazil, Mexico
Middle East: Israel
Africa: Tunisia

Language Arab, Chinese, Dutch, English,
Chinese, French, German, Greek,
Hebrew, Japanese, Korean,
Norwegian, Portuguese, Spanish
The language used to conduct the
study
Bicultural
group
Racial or ethnic minorities People classified as members of
minority racial-ethnic groups in the
country of testing (e.g., African
Americans in the U.S.)
Sojourners International students, expats, and
others living in a country other than
their country of origin at the time of
testing
Citizens from societies with a
colonial history
People from regions with a colonial
history (e.g., Hong Kong, Israel,
Tunisia)
Citizens from bi- or multi-lingual
societies
People from societies in which more
than one language is both officially
and commonly used (e.g., Belgium,
Switzerland)
Bi- or multi-linguals in monolingual
society
Bi- or multi-lingual people in
monolingual societies (e.g., English-
speaking Chinese in China)
Monoculturals People who do not fit any of the
above categories
Bicultural
Identity
Integration
High BII Participants who scored high on the
BII scale
Low BII Participants who scored high on the
BII scale
`

166
Gender % Female Calculated using study-reported
gender. We coded results for males
and females separately if these data
were available.
Participant
Source
College students Participants were college students
Crowdsourced adults Adults recruited from crowdsourcing
platform(s)
Non-crowdsourced, non-student
adults
Adults who were not college
students or crowdsourced
Children Children, participants <18 years old
Priming task Icon Primes:
Language, Cultural Icons, Ethnic
identity cues, Partner nationality or
ethnicity, Cultural context
Priming tasks that utilize cues
specific to a specific region or
cultural group to activate the cultural
mindset associated with the region
or the group
Content and Goal Primes:
Pronoun circling task, SDFF,
Sumerian warrior story, Scrambled
sentence, Biased scale, Subliminal
prime, Value promotion, Word
presentation, group image with or
without accompanying text, Writing
about self vs. with others, Recall self
vs. social experiences, Individual vs.
group imagination, Individual vs.
group framing
Priming tasks meant to activate
general content and goals core to the
cultural mindset
Cognitive Style Primes Priming tasks that focus on
manipulating analytic and holistic
cognitive styles
Outcome type Self-concept: Twenty-statement task
and variants, Self- vs. others-focused
thought index, Singelis self-
construals, Private self-concept,
Relational self-concept, Collective
self-concept, Relational & collective
self-concept
Measure aspects of self-concept
Values: Individualism-collectivism
values, Non-individualism-
collectivism values
Measure endorsement of values
Personal characteristics:
Personality traits, Self-enhancement
Measure personal attributes (e.g.,
traits, self-enhancing tendencies)
Social attitudes: Relationality,
Social judgments
Measure interpersonal or social
attitudes and judgment
Attribution: Measure attribution to internal
(personal) or external (situational)
factors
`

167
Fish attribution task, Trait vs.
situation inferences, Other attribution
tasks
Cognitive procedures: Analytic-
holistic thinking, Temporal
connection, Other separation-
connection procedures
Measure relevant cognitive
procedures (e.g., holistic vs. analytic
reasoning, connection vs. separation
in space or time)
Personal characteristics:
Personality traits, Self-enhancement
Measure personal attributes (e.g.,
traits, self-enhancing tendencies)
Emotionality Measure emotions or well-being
Consumer attitudes Measure attitudes or behavior
related to consumer products
Creativity Measures divergent or novel
thinking
Format Online computerized, In lab
computerized, In lab pen-paper, In
lab hybrid, In lab interaction, In lab
neural scan, Offline outside of lab
Format in which the study was
delivered
Study design Between-subject, Within-subject,
Mixed factorial
Whether the effect size was
estimated from between-subject,
within-subject, or mixed factorial
comparisons
Manipulation
check measure
Measure is a manipulation check,
Not also a manipulation check
Whether the effect size was
computed from a measure described
as a manipulation check
Manipulation
check
presence
Manipulation check is present (yes,
no)
Whether the study has conducted a
manipulation check for the priming
task
Manipulation
check success
Success, Failure For studies that included a
manipulation check, whether the
priming task had a significant effect
in the predicted direction on the
manipulation check items or not
Number of
outcomes
Number of culture-relevant
outcomes measured after the
priming task, excluding
manipulation check measures (if
any)
Outcome order The order in which this outcome
measure appeared in the study,
excluding manipulation check and
outcomes irrelevant to culture
Publication
year
Year in which the article first
appeared online
Publication
status
Published, Unpublished Whether the article was published in
a peer reviewed journal or was
`

168

unpublished (theses, dissertations,
conference presentations,
unpublished data)
Completeness
of reporting
Possible total score ranges from 0 to
6. One point to each report in the
study of how the sample was
recruited (source), country or
language in which the study was
conducted (country or language),
study format (format), mean age
(age), gender make-up (% female),
and race-ethnic make-up
(ethnicity %).
`

169
Table 4. Summary of Study Characteristics
Characteristics Level %
Publication status Published 88.9
Unpublished 11.1
Experimental comparison Individualism-Collectivism 78.9
Individualism-Control 10.8
Collectivism-Control 8.9
Honor – Control 1.5
Region English-speaking countries 54.7
Asia 21.1
Northern and Western Europe 10.6
Southern and Eastern Europe 0.8
Middle East and Africa 5.2
Latin America 0.5
Multiple regions 1.3
Information missing 5.8
Biculturalism Bicultural 63.1
Monocultural 32.7
Information missing 4.2
Participant source College students 84.9
Crowdsourced adults 6.9
Non-crowdsourced, non-college-student
adults
2.5
Children 0.4
Multiple sources 2.3
Information missing 3
Study design Between-subject 94.7
Within-subject 4.5
Mixed factorial 0.7
Note. % = Percentage of effect sizes

`

170
Table 5. Estimates of Overall Effect Size
k s d 95% CI
1
p
Overall effect
Random-effects with RVE 946 307 0.41 0.38, 0.45 <.001
Bayesian random-effects 946 307 0.40 0.36, 0.44 --
Bayesian fixed-effects 946 307 0.35 0.33, 0.37 --
Bayesian model-averaging 946 307 0.40 0.36, 0.44 --
Publication bias-corrected overall effect
Trim-and-fill 946 307 0.30 0.26, 0.34 <.001
PET-PEESE 946 307 0.30 0.23, 0.37 <.001
Excluding manipulation checks 804 273 0.38 0.34, 0.42 <.001
Overall effect by experimental comparison
Ind vs. Col, random-effects with RVE 746 285 0.44 0.40, 0.48 <.001
Ind vs. Col, Bayesian random-effects with
prior from Oyserman & Lee (2008)
746 285 0.37 0.35, 0.49 --
Ind vs. control, random-effects with RVE 84 35 0.22 0.10, 0.34 <.001
Col vs. control, random-effects with RVE 102 43 0.35 0.27, 0.44 <.001
Honor vs. control, random-effects with RVE 14 6 0.29 0.08, 0.50 .02
Note.
1
95% CI represents 95% confidence interval, except in the cases of Bayesian models, 95%
highest posterior density intervals are used. Ind = Individualism-primed. Col = Collectivism-
primed. Bayesian models do not provide p-values.

`

171
Table 6. Moderator Analyses of Prime Task and Outcome Type
Moderator (Bolded)
group and level
k s d F 95% CI p
Priming task 861 273 3.1 .006
Icon primes 180 60 0.36

0.26, 0.47 <.001
Cultural context 6 3 -0.08 -0.44, 0.28 .45
Cultural icon 89 30 0.51 0.32, 0.69 <.001
Language 66 20 0.27 0.13, 0.42 <.001
Ethnic identity cue 13 4 0.38 -0.17, 0.94 0.11
Individual vs group imagination 11 4 0.76

0.36, 1.16 .01
Partner nationality or ethnicity 6 3 0.20 -0.85, 1.26 .49
Content and goal primes 618 198 0.42 0.38,0.47 <.001
Pronoun circling 288 93 0.37

0.31, 0.43 <.001
SDFF 100 33 0.42 0.32, 0.52 <.001
Sumerian warrior 77 23 0.42

0.28, 0.56 <.001
Group relational vs individual framing 39 7 0.37 0.06, 0.67 .03
Writing about self vs with others 23 7 0.69 0.4, 0.98 .001
Value promotion 23 6 0.27 -0.2, 0.73 .20
Group image 21 5 0.54 0.41, 0.68 <.001
Scrambled sentence 18 11 0.65

0.4, 0.9 <.001
Word presentation with or without image 12 4 0.14 -0.68, 0.95 .63
Biased scale 7 6 0.42 0.12, 0.73 .02
Cognitive style primes 64 20 0.44 0.32, 0.55 <.001
Outcome type 943 304 2.48 .002
Self-Concept 183 91 0.56

0.47, 0.65 <.001
Singelis’s self-construal scale 41 20 0.41 0.19, 0.64 0.001
Ten/twenty statement task 23 18 0.59 0.45, 0.73 <.001
Self-other thought index 23 14 0.94 0.57, 1.31 <.001
Private self-concept 39 19 0.43 0.27, 0.59 <.001
Relational self-concept 19 13 0.41

0.25, 0.58 <.001
Collective self-concept 12 9 0.69 0.37, 1.00 .001
Relational and collective self-concept 8 6 0.34 -0.11, 0.8 .11
Other self-concept 16 10 0.30 0.12, 0.48 .004
Values 114 34 0.38 0.26, 0.50 <.001
Individualism-collectivism values 46 11 0.52

0.22, 0.83 .004
Other values 68 23 0.32 0.21, 0.42 <.001
Relational and Social Attitudes 227 34 0.39 0.30, 0.47 <.001
Relationality 147 55 0.38 0.26, 0.49 <.001
Social judgment 79 28 0.39 0.27, 0.51 <.001
Cognition 177 79 0.39

0.32, 0.46 <.001
Underwater animation task 10 5 0.82

-0.44, 2.07 .14
Trait situation inference 19 7 0.23 -0.11, 0.58 .15
Other attribution tasks 24 9 0.17 -0.13, 0.47 .23
`

172
Analytic holistic thinking 53 21 0.46 0.34, 0.59 <.001
Temporal connection 22 8 0.50

0.36, 0.64 <.001
Other separation-connection procedures 49 32 0.36 0.28, 0.44 <.001
Personal Characteristics 48 20 0.29 0.11, 0.46 .003
Personality 16 6 0.17 -0.19, 0.54 .26
Self enhancement 32 14 0.33 0.11, 0.56 .01
Consumer attitudes 149 46 0.39

0.3, 0.49 <.001
Emotionality 32 15 0.45 0.26, 0.65 <.001
Creativity 16 3 0.22 -0.74, 1.19 .42
Note. s = number of studies; k = number of effect sizes; d = Cohen’s standardized d; F = F-value
from Wald-test of overall moderation effect of specific priming task and outcome type; 95% CI
= 95% confidence interval; p = p-value.

`

173
Table 7. Cohen’s d Effect Size (Number of Effect Sizes) by Prime and Outcome Domain
Prime Domain
Outcome Domain Society Self Relationality Value Cognitive
style
Self-concept 0.45 (25) 0.57 (122) 0.55 (13) 0.75 (11) 0.53 (4)
1

Social attitudes 0.38 (25) 0.44 (153) 0.10 (11) 0.10 (6) 0.30 (5)
Personal
characteristics
0.11 (24) 0.50 (18) -- -- --
Values 0.42
+
(43) 0.42 (31) 0.44 (32) -0.07 (8) --
Cognition 0.46 (57) 0.33 (69) -- -- 0.52 (37)
Emotionality -- 0.61 (14) -- -- --
Consumer attitudes -- 0.30 (107) -- 0.22 (9) 0.31 (22)
Note. Each cell presents the mean weighted effect size and the number of effect sizes from which
the effect size was calculated from in parentheses. Effect sizes with 95% confidence intervals
that include zero are presented in grey.
1
Effect size was computed based on k = 4 hence may not
be stable.

`

174
Figure 1
PRISMA Flow chart documenting literature search and article inclusion and exclusion.

Screened Articles
(n = 26,312)

Irrelevant article (n =
26,035)
Full-text articles
assessed for eligibility
(n = 295)

Articles identified by
searching databases
(n = 28,607)
Unique articles after removing duplicates
(n = 26,312)

Articles identified via other
sources
(n = 24)
Non-experiments
(n = 4)

Non-cultural
primes (n = 48)

Biculturalism
primes (n = 19)

Article unavailable or
redundant (n = 2)

Neural outcomes only
(n = 7)

Society-specific
outcomes (n = 19)

Unclear a priori
predictions
(n = 11)

Lack of statistics to
compute effect size
(n = 13)

Articles included in
meta-analysis
(n = 174)

`

175
Figure 2
Top panel (a): the weakly informed prior (gray line) and posterior (blue line) distributions of
overall cultural priming effect sizes. Bottom panel (b): the informed prior distribution (gray line)
from Oyserman and Lee (2008) and posterior distribution (blue line) of effect sizes for priming
individualism and collectivism.

a
b
`

176

Figure 3
a) Overall funnel plot of cultural priming effects and b) funnel plot of cultural priming effects
after trim-and-fill method

a
b
`

177
Figure 4
Magnitudes of cultural priming effects by country or territory. Error bars represent 95%
confidence intervals.

`

178
Figure 5
Average effect sizes of priming individualism or priming collectivism did not vary across the
West, the East, and the rest of the world

`

179
Appendix B: Chapter 2 Tables and Figures

Table 1. Overview of Study Methods and Operationalizations of Collectivism and Empty Claims
Study Method and sample Operationalization of
collectivism
Operationalization of empty
claims
1 A correlational study, U.S.
national sample (GSS)
Collectivistic child-rearing
values
Astrology
2 A correlational study, China
national sample (COSS)
Collectivistic child-rearing
and group values
Superstition.
Pseudoscientific claims
3a-c Correlational studies, online
samples in U.S. and China
Collectivism scale
a
COVID-19 fake news;
Randomly generated
pseudoscientific news
4 A between-country
comparison (U.S., China,
college students)
Country and
Collectivism scale
Randomly generated
sentences (metaphors,
sentences formed from vague
word-strings
b
);
Astrology
5 A correlational study, U.S.
online samples
Collectivism scale A randomly generated
metaphor from Study 4
6a-b Experimental studies, U.S.
online samples
Manipulation of
momentary experience of
oneself being collectivistic
Randomly generated
sentences
b
;
Randomly generated
pseudoscientific news
7 An experimental study,
online sample in China
Collectivism scale Study 4 randomly generated
metaphors
Note. GSS = General Social Survey; COSS = Chinese Online Social Survey;
a
We used
Oyserman’s (1993) collectivism scale;
b
We used Pennycook and colleague’s (2015) bullshit-
receptivity scale with non-probative sentences randomly generated from ambiguous word-
strings.

`

180
Figure 1
People who endorse collectivistic values more (y-axis) are more likely to find astrology scientific
(Study 1, left panel) and think Wi-Fi kills sperm (Study 2, right panel)

Note. Error bars = 95% confidence intervals.

`

181
Figure 2
Study 4: Collectivism is related to seeing meaning in randomly formed metaphors (left),
randomly generated vague statements (middle), and astrology (right) in Study 4

Note. The blue line represents China; the red line represents the U.S.

`

182
Figure 3
Study 7: The relationship between collectivism and seeing meaning is a function of motivation to
establish common ground

Note. The blue line represents the positive relationship in the Human-Generated condition; red
represents the null relationship in the Non-Human Generated condition.

`

183

Appendix C: Chapter 3 Tables and Figures
Table 1. Overview of Studies 1 to 8
Study Language Conditions Potential
Mediators
Potential Moderators Primary
DV
Secondary
DV
1 English Match, Mismatch Momentary
affect
NFC, individualism,
collectivism
Inherence
2 Hebrew Match, Mismatch Momentary
affect
NFC, individualism,
collectivism
Inherence Essentialism
3 English Match, Mismatch Inherence Essentialism
4 Chinese Match, Mismatch Inherence Essentialism
5 English Match, Mismatch, Control Inherence
6 English Match, Mismatch Inherence Self-certainty
7 English Match, Mismatch, Explicit
Mismatch
Inherence
8 Hebrew Match, Mismatch, Explicit
Mismatch
PNS, IU Inherence
Note. DV = dependent variable, NFC= Need for Cognition, PNS=Personal Need for Structure, IU=Intolerance of Uncertainty.

`

184
Table 2. Studies 1 to 8: Study Location, Sample Demographics, and Size and Reasons for Exclusions
Study, Sample, and Demographics Sample Size
Study Country Gender Age Race-
Ethnicity
N Total
Excluded
Rationale for Exclusion (n per reason)

%
Women
M SD % Majority

Failed
Attention
Check n
Other Criteria n
1 U.S. 48% 36.16 12.93 EA 77.7%
a
197
b
18 14 4 identified wrong holiday
2 Israel 48% 29.13 6.79 Jewish
100%
147 8 8 --
3 U.S. 62% 36.13 10.73 EA 100%
c
120 40 8 35 not EA
4 China 48% 28.11 5.46 Han 100%
d
100 8 5 3 not Han
5 U.S. 51% 34.53 10.22 EA
74.71%
e

259 9 7 2 skipped dependent variable
6 U.S. 72% 19.98 2.61 EA
34.94%
f

332 359 49 105 non-native English speaker; 205
repeat responders
7 U.S. 53% 34.92 11.02 EA 69.3%
g
306 29 21 8 non-native English speaker
8 Israel 73% 30.64 5.59 Jewish
100%
342 12 10 1 vegetarian, 1 non-native Hebrew
speaker
Note: N = Sample for analysis after excluding people who failed the manipulation check (Studies 1 to 6), could not name the holiday
and/or date it occurs (Studies 1, 5, 6), whose first language is not English (Studies 1, 6, 7), repeat responders who took part in a pilot
test to establish scale validity (Study 6) and/or whose racial-ethnic heritage was not the same as the stimuli materials (Studies 3, 4).
a
EA= European American, others were African American (6.09%), Asian American (8.12%), Latino American (5.08)%, Native
American (1.02%), Middle Eastern or other American (2.03%). 98.98% were native speakers of English.
b
N = 102 on February 13 or
14, the day before or day of Valentine’s Day and N = 95 five weeks after, March 23
rd
.
c
EA= European American, all native speakers
of English.
d
Han Chinese are the main ethnic group (92% of Chinese population) in China.

EA= European American, others were
African American (7.78%), Asian American (8.56%), Latino American (5.06%), Native American (2.33%), Middle Eastern or other
American (2.33%).
f
EA= European American, others were African American (6.62%), Asian American (33.43%), Latino American
(9.34%), Native American (6.02%), Middle Eastern (3.61%), and other American (11.45%).
g
EA= European American, others were
African American (8.50%), Asian American (8.50%), Latino American (5.23%), Asian American (11.76%), Native American
(1.96%), Middle Eastern and other American (2.61%).
`

185
Table 3. Studies 1 - 8: Manipulation of Cultural Match, Mismatch, and Control
Study Verbatim instructions and how many
products were rated in each condition
Condition
Number of
products
Match (n) Mismatch (n) Control
a
(n)
1, 6,
7
b
“In collaboration with local
paper and package
companies, you will be asked
to rate the quality and
attractiveness of Valentine’s
Day card designs.”
4 Pink Heart
Valentine’s Day
Cards (n=97 Study 1;
n = 164 Study 6;
n=104 Study 7)
Heart-shaped skull
Valentine’s Day
Cards (n=100 Study
1; n = 168 Study 6; n
= 97 Study 7)
--
2, 8
b
“In collaboration with local
restaurants (Study 2)/a
website that offers different
breakfast dishes (Study 8),
you will be asked to rate the
quality and attractiveness of
plated breakfasts.”
8 Breakfasts of raw
vegetables, fresh
cheeses, fresh bread,
fried eggs (n=70
Study 2, n=111 Study
8)
Breakfasts of cooked
vegetables, meats,
fried breads,
croissants, Eggs
Benedict (n=77 Study
2, n=117 Study 8)
--
3 “In collaboration with local
wedding photographers, you
will be asked to rate the
quality and attractiveness of
wedding photographs.”
4 Bride in a white
gown, groom in a
black tuxedo, a white
fondant iced floral
tiered wedding cake
(n=62)
Bride in a black
gown, groom in an
off-white tuxedo, a
black fondant tiered
wedding cake with
the word “love”
(n=58)
--
4 “In collaboration with local
wedding photographers, you
will be asked to rate the
quality and attractiveness of
wedding photographs.”
5 Bride in a white
gown, groom in a
sharply tailored suit,
wedding car
festooned with
flowers (n=54)
Same bride, groom
and car, but bride in a
black sparkling gown,
wedding car
festooned with fruit
(n=46)
--
`

186
5 “In collaboration with
department stores, you will
be asked to rate the quality
and attractiveness of
shopping bag designs.”
4 Patriotic-themed
(fireworks, stars,
American flag) and
colored (red, blue)
shopping bags.
“Happy Labor Day”
logos (n=85)
Environmental-
themed (trees,
animals) and colored
(green, reddish
brown) shopping
bags. “Happy Labor
Day” logos (n=89)
Environmentally-
themed (trees, animals)
and colored (green,
reddish brown)
shopping bags.
“Shopping Bag” logos
(n=85)

Note. Products were rated for Quality (1= very poor, 7= excellent) and Attractiveness (1= extremely unattractive, 7= extremely
attractive). Wedding photographs were judged only on quality, not on attractiveness to avoid confusion with attractiveness of the
couple. Study 1 consisted of two subsamples that were recruited on Valentine’s Day (n = 102) and five weeks after Valentine’s Day (n
= 95). The Supplemental Materials show results separately. Effects did not differ by data collection date, thus they are combined.
a
Control Group: Only Study 5 had a Neutral Control group.
b
In Studies 7 and 8 the participant randomly assigned to the Explicit
Match condition (Study 7 n = 105, Study 8 n = 114) rated the same products as participants in the Mismatch condition, what differed
was the forewarning.

`

187
Table 4. Quality and Attractiveness Product Ratings by Condition

Note.
a
We took the product ratings of quality and attractiveness and created a mean composite score, which yields a mean fluency
score. For clarity, we report the total number of ratings and the Cronbach alpha (α) reliability (across conditions) of the fluency
composite score. In Studies 1, 2, 5 to 8, each product was rated for quality and attractiveness. In studies 3, 4, and 5 Pilot, each product
was only rated for quality. This yielded a total of 8 ratings in Studies 1, 5, 6 and 7, a total of 4 ratings in Studies 3 and 5 Pilot, a total
of 5 ratings in Study 4, and a total of 16 ratings is Studies 2 and 8. Study 5 was the only study with a Control group and Studies 7 and
8 were the only studies with an Explicit Mismatch condition, people randomly assigned to this group saw the same products as people
randomly assigned to the Mismatch condition saw, what different is that they were forewarned (Mean Quality and Attractiveness
Composite did not differ between the Mismatch and the Explicit Mismatch conditions (Study 7: t(203) = .95, p = .34; Study 8: t(229)
= 1.42, p = .16)). d = Cohen’s d, which reflects the magnitude of the difference between Cultural Match and Cultural Mismatch
conditions.

Study Fluency (Mean of Quality and
Attractiveness) Score Reliability
a
Fluency Score (Mean of Quality and Attractiveness) by Condition
Number of ratings
forming the score
α Match M
(SD)
Mismatch M
(SD)
Control

M
(SD)

Explicit M
(SD)
df F-Test p d
1 8 .90 4.76 (1.00) 3.34 (1.42) -- -- 1,193 65.13 <.001 1.16
2 16 .79 3.55 (.55) 3.12 (.57) -- -- 1,145 21.22 <.001 0.77
3 4 .76 5.51 (.87) 5.01 (1.00) -- -- 1,118 8.46 .004 0.53
4 5 .58 4.43 (.95) 4.34 (.81) -- -- 1,97 0.24 .63 0.10
5
Pilot
4 .90 4.81 (1.11) 3.47 (1.57) 4.70 (.89) -- 1,81 10.21 <.001 0.99
5 8 .92 4.22 (1.20) 3.29 (1.06) 4.12
(1.13)
-- 2,254 17.85 <.001 0.82
6 8 .92 4.56 (1.05) 2.56 (1.12) -- -- 1,329 281.67 <.001 1.84
7 8 .84 5.17 (.76) 3.77 (1.51) -- 3.57
(1.70)
2,303 41.31 <.001 1.17
8

16 .87 3.91 (.62) 3.36 (.74) -- 3.23 (.65) 2,339 32.03 <.001 0.81
`

188
Table 5. Similarity to Expectation Product Ratings by Condition
Study Similarity to Expectations Scale and reliability Mean Similarity to Expectation by condition
Specific Questions and Response Scale α Match
M (SD)
Mismatch
M (SD)
Explicit
M (SD)
df F-test p d
1
a
-- -- -- -- -- -- -- -- --
2 1. “As a whole, how traditional were the images you
saw previously?” (1=not at all traditional, 7= very
traditional).
2. “As a whole, how similar were the images of the
breakfasts you saw previously to typical Israeli
breakfasts?” (1=very dissimilar, 7=very similar)
.74 5.63
(0.94)
2.92
(1.25)
-- 1,145 216.51 <.001 2.45
3

1. “As a whole, how traditional was the wedding
you viewed?” (1=not at all traditional, 7= very
traditional).
2. “As a whole, when you think of weddings, how
similar were the images of the wedding you saw
previously to the weddings you imagine?” (1=very
dissimilar, 7=very similar).
.89 6.35
(0.69)
4.23
(1.61)
-- 1,118 89.59 <.001 1.71
4

1. “As a whole, did the images of the wedding you
viewed fit your impression of weddings?” (1=not at
all, 7 = very)
2. “As a whole, how appropriate were the images of
the wedding you viewed?” (1=not at all
appropriate, 7=very appropriate)
3. “As a whole, how similar were the images of the
wedding you saw previously to the weddings you
see in real life?” (1=very dissimilar, 7=very similar).
.70 5.31
(1.06)
3.04
(0.94)
-- 1,98 1.82 .18 2.27
5 1. “As a whole, how traditional were the shopping
bag designs that you viewed?” (1=not at all
traditional, 7= very traditional)
2. “As a whole, are the shopping bag designs you
viewed similar to the ones that you expect to see
.62 5.52
(0.87)
3.34
(1.73)
-- 1,172 110.05 <.001 1.59
`

189
Note.
a
In Study 1, manipulation check questions were omitted due to researcher error.α =Cronbach’s alpha. d = Cohen’s d, which
reflects the magnitude of the difference between Cultural Match and Cultural Mismatch conditions.
c
In Study 6, manipulation check
questions were omitted for 47% of the sample due to researcher error. In Studies 7 and 8, participants rated the same products in
Mismatch and Explicit Mismatch conditions; hence data were combined for ratings of cultural products. Products were rated less
similar to expectations in Explicit Mismatch condition than in the Mismatch condition in Study 7, t(203) = 3.65, p <.001, but were
rated similarly in Study 8, t(229) = .06, p = .96 .
during Labor Day?” (1=very dissimilar, 7=very
similar).
6
c
1.“As a whole, how traditional were the Valentine’s
Day cards that you viewed?” (1=not at all
traditional, 7= very traditional)
2.“As a whole, how similar were the Valentine’s day
cards to the ones that you expect to see?” (1=very
dissimilar, 7=very similar).
.90 5.49
(1.01)
1.82
(0.88)
-- 1,175 651.97 <.001 3.87
7 1.“As a whole, how traditional were the Valentine’s
Day cards that you viewed?” (1=not at all
traditional, 7= very traditional)
2.“As a whole, how similar were the Valentine’s day
cards to the ones that you expect to see?” (1=very
dissimilar, 7=very similar).
.82 5.31
(1.08)
2.71
(1.72)
1.94
(1.34)
2,303 166.15 <.001 1.81
8 1. "As a whole, how traditionally Israeli were the
breakfast dishes you viewed?" (1=not al all,
5=very much)
2. “As a whole, how similar were the breakfasts
you saw to typical Israeli breakfasts?” (1=not al
all, 5=very much)
3. “As a whole, how typical were the ingredients
for Israeli breakfasts?” (1=not al all, 5=very
much)

.93 4.50
(0.61)
2.35
(0.90)
2.35
(0.79)
2,339 284.60 <.001 2.80
`

190
Table 6. Cultural Fluency and Disfluency Effects on Inherence by Study and Meta-analytical Summary
Study
# Cultural Product Location Match
M (SD)
Mismatch
M (SD)
df F p Cohen’s d 95% CI
1 Valentines’ Day Cards U.S. 4.97 (.96) 4.61 (.93) 1,195 6.94 .01 0.38 0.10, 0.66
2 Plated Breakfasts Israel 4.89 (.66) 4.63 (.83) 1,145 4.46 .04 0.34 0.02, 0.67
3 Wedding Photographs U.S. 4.92 (.87) 4.57 (.86) 1,118 5.01 .03 0.40 0.04, 0.76
4 Wedding Photographs China 5.16 (.51) 4.89 (.61) 1, 98 6.07 .02 0.48 0.08, 0.88
5 Labor Day Shopping Bags U.S. 4.99 (.83) 4.68 (.90) 1,172

5.40 .02 0.36 0.06, 0.66
6 Valentines’ Day Cards U.S. 4.79 (.67) 4.62 (.78) 1,330 4.63 .03 0.23 0.02, 0.33
7 Valentines’ Day Cards U.S. 5.08 (.78) 4.80 (.94) 1,199 5.32 .02 0.32 0.04, 0.52
8 Plated Breakfasts Israel 5.09 (.79) 4.88 (.79) 1,226 3.95 .05 0.26 0.01, 0.53
Weighted average effect size 0.32 0.22, 0.43
Heterogeneity statistic Chi-square= 1.85, df = 7, p = .97, I
2
= 0%

`

191
Table 7. Effects of Cultural Fluency and Disfluency on Essentializing
Study Direct Effect Mediation Analyses
# Location Event Match
M (SD)
Mismatch
M (SD)
df F-test p ab 95% CI SE df F-test p R
2

2 Israel Breakfast 4.21 (.89) 4.31 (1.02) 1,145 0.33 .57 -.16 -.33 -.03 .08 2,144 19.65 <.001 21%
3 U.S. Wedding 4.33 (.86) 4.24 (.85) 1,118 0.36 .55 -.14 -.28, -.03 .06 2,117 12.13 <.001 17%
4 China Wedding 5.14 (.76) 4.88 (.73) 1, 98 2.94 .09 -.17 -.33, -.05 .07 2, 97 14.16 <.001 23%
Note. Essentializing was not assessed in Studies 1 and 5 to 8.
`

192
Figure 1
Cultural fluency and disfluency theoretical process model: How match and mismatch between
cultural-based expectation and observation affects experienced inherence and, via inherence,
essentializing.

`

193

Figure 2
Individual Differences May Moderate The Effect of Cultural Fluency and Cultural Disfluency on
Inherence

`

194
Figure 3
Experienced Mismatch with Cultural Expectation Reduces Inherence Relative to Experienced
Match and Neutral
Note. Studies 1 to 8: Mean inherence scores of participants randomly assigned to view products
that Matched cultural expectation (culturally fluent, grey bar), Mismatched cultural expectation
(culturally disfluent, white bar), were neutral (control, hatched bar), or Explicitly mismatched
cultural expectation (explicit disfluency, white bar with dots). Error bars indicate 95%
confidence intervals.

195

195
Figure 4
The Effect of Cultural Disfluency on Essentializing via Inherence (Studies 2-4)

Note: The meta-analytic synthesized effect is shown in bold, study specific effects are labeled by
study. * p < .05, ** p <.01.

Abstract (if available)

Abstract Much of cultural psychological research is rooted in identifying differences across cultures. This approach has inevitably led to the perception that culture has stable influences. Moving away from a static view of culture, this three-paper dissertation examines culture’s consequences for how we act and what we believe as processes situated in context. In Chapter 1, I synthesize the wide-ranging consequences of momentary cues that make a cultural mindset accessible using a meta-analytical approach. A decade of cumulative evidence from cultural priming research suggests that cultural mindsets?individualism, collectivism, and honor, are situated, universal, and psychologically consequential. In Chapters 2 and 3, I identify two unique consequences of culture for reasoning. Across seven studies in Chapter 2, I show that collectivism, a universal cultural mindset that prioritizes fitting in and connecting with others, motivates people to construct meaning and therefore find meaning even in places where meaning might not exist. This meaning-making process has far-reaching implications for the consumption of misinformation. In Chapter 3, I present eight experiments focusing on an important, yet underappreciated aspect of culture?having cultural expertise. I show that having cultural expertise allows people to make pragmatic inferences based on the immediate situation. Subtle mismatches between culture-based expectations and situations disrupt people’s sense that current patterns represent a natural order, calling into question whether social categories have stable essences. Collectively, empirical findings across three chapters highlight a situated account of culture’s consequences. Situational features that bring a cultural mindset to mind or interact with culture-based knowledge powerfully influence how we act and what we believe.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

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

PDF

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

PDF

Socio-ecological psychology of moral values

PDF

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

PDF

The antecedents and consequences of believing that difficulties are character-building

PDF

Expressing values and group identity through behavior and language

PDF

Classrooms are game settings: learning through and with play

PDF

Metacognitive experiences in judgments of truth and risk

PDF

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

PDF

Can I make the time or is time running out? That depends in part on how I think about difficulty

PDF

Thinking Of Trump and Weinstein: the impact of prominent cases of sexual harassment on perceptions of sexual harassment across countries

PDF

Difficulty-as-importance and difficulty-as-impossibility: unpacking the context-sensitivity and consequences of identity-based inferences from difficulty

PDF

People can change when you want them to: changes in identity-based motivation affect student and teacher Pathways experience

PDF

Using classification and regression trees (CART) and random forests to address missing data

PDF

#BLM or #ALM: accessible perspective shapes downstream judgment even among people high in social dominance

PDF

When photos backfire: truthiness and falsiness effects in comparative judgements

PDF

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

PDF

The cost of missing objectives in multiattribute decision modeling

PDF

Are jokes funnier when they’re easier to process?

Asset Metadata

Creator Lin, Ying (author)

Core Title Culture's consequences: a situated account

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Psychology

Degree Conferral Date 2021-08

Publication Date 07/28/2021

Defense Date 05/12/2021

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

Tag culture,Judgment,meta-analysis,OAI-PMH Harvest,Psychology,social cognition

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Oyserman, Daphna (committee chair), Dehghani, Morteza (committee member), Schwarz, Norbert (committee member), Townsend, Sarah (committee member)

Creator Email ying.lin2049@gmail.com,yinglin@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC15658933

Unique identifier UC15658933

Legacy Identifier etd-LinYing-9920

Document Type Dissertation

Format application/pdf (imt)

Rights Lin, Ying

Type texts

Source 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