Pers Soc Psychol Rev 2015 Parks Leduc 3 29

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/262224477
Personality traits and personal values: A meta-analysis
Article in Personality and Social Psychology Review · April 2014

DOI: 10.1177/1088868314538548
CITATIONS READS
137 12,460
3 authors:
Laura Parks-Leduc Gilad Feldman

James Madison University The University of Hong Kong
13 PUBLICATIONS 875 CITATIONS 38 PUBLICATIONS 283 CITATIONS
SEE PROFILE SEE PROFILE
Anat Bardi
Royal Holloway, University of London
41 PUBLICATIONS 3,629 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Impact of prior investment on decision-making: Pre-registered replication of the Arkes & Blumer’s (1985) sunk cost effect View project
Action-inaction biases View project
All content following this page was uploaded by Gilad Feldman on 22 October 2015.
The user has requested enhancement of the downloaded file.

538548
research-article2014
PSRXXX10.1177/1088868314538548Personality and Social Psychology ReviewParks-Leduc et al.
Article
Personality and Social Psychology Review
Personality Traits and Personal Values:

2015, Vol. 19(1) 3–29
© 2014 by the Society for Personality
and Social Psychology, Inc.
A Meta-Analysis Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1088868314538548
pspr.sagepub.com
Laura Parks-Leduc1, Gilad Feldman2, and Anat Bardi3
Abstract
Personality traits and personal values are important psychological characteristics, serving as important predictors of many
outcomes. Yet, they are frequently studied separately, leaving the field with a limited understanding of their relationships.
We review existing perspectives regarding the nature of the relationships between traits and values and provide a conceptual
underpinning for understanding the strength of these relationships. Using 60 studies, we present a meta-analysis of the
relationships between the Five-Factor Model (FFM) of personality traits and the Schwartz values, and demonstrate consistent
and theoretically meaningful relationships. However, these relationships were not generally large, demonstrating that traits
and values are distinct constructs. We find support for our premise that more cognitively based traits are more strongly
related to values and more emotionally based traits are less strongly related to values. Findings also suggest that controlling
for personal scale-use tendencies in values is advisable.
Keywords
personality traits, personal values, meta-analysis
Advancing an integrative view of the person is a major goal context of the most researched models of traits and values,
in current personality research (see Barenbaum & Winter, the Big Five (or Five-Factor) model and Schwartz’s (1992)
2008; Cervone, 2005; McAdams & Pals, 2006; Sheldon, Value Theory (respectively).
2004). Although personality traits have often been viewed as
central to the understanding of the person, the position of
Personality Traits and Personal Values
personal values has generally been more peripheral (see, for
example, Bilsky & Schwartz, 1994; Buss, 1989; Hofstee, Personality traits are typically defined as descriptions of
1994). Some personality scholars have suggested the inclu- people in terms of relatively stable patterns of behavior,
sion of values in an integrative model of characteristics of thoughts, and emotions (e.g., McCrae & Costa, 2003). The
the individual (McAdams, 1996; McClelland, 1996; Shoda Five-Factor Model (FFM) is the most researched taxonomy
& Mischel, 2006; Winter, John, Stewart, Klohnen, & Duncan, of traits worldwide (e.g., Allik, 2005; McCrae & Costa,
1998), yet, little theoretical or empirical work has been 1997); within this model, a large number of traits are com-
developed to accomplish this goal (Schwartz, 2011a). If bined into five broad trait dimensions that load onto orthogo-
traits and values are to be combined into a unified model, a nal factors. The factors and descriptive traits for each are
starting point is to examine empirical links between person- provided in Table 1.
ality traits and personal values. Personal values (e.g., achievement, security) are gener-
In this article, we review and clarify conceptual issues ally described as rather stable broad life goals that are impor-
regarding proposed models of relationships between person- tant to people in their lives and guide their perception,
ality traits and values, and use meta-analysis to summarize judgments, and behavior (e.g., Rokeach, 1973; Schwartz,
past findings regarding these relationships to advance a more 1992). Values are organized in personal hierarchies of
integrative understanding of the person. We make the follow-
ing contributions to the literature: First, we clarify defini- 1
tions and describe various views on the nature of the James Madison University, Harrisonburg, VA, USA
2
Hong Kong University of Science and Technology, Clearwater Bay, Hong
relationships between traits and values; second, we propose Kong
a conceptual underpinning for understanding which traits 3
Royal Holloway University of London, Surrey, UK
should have stronger relationships with values and why; and
Corresponding Author:
third, we offer empirical support for the distinction between Laura Parks-Leduc, Department of Management, James Madison
traits and values and explore their interrelations using meta- University, 800 S. Main St., MSC 0205, Harrisonburg, VA 22807, USA.
analysis. We position our discussion and analyses within the Email: leduclm@jmu.edu
Downloaded from psr.sagepub.com by guest on December 31, 2014

4 Personality and Social Psychology Review 19(1)
Table 1. Five-Factor Model of Personality.
Construct Description: The extent to which individuals tend to be . . .

Openness to Experience . . .curious, intellectual, imaginative, creative, innovative, and flexible (vs. closed-minded,
shallow, and simple)
Agreeableness . . .helpful, good-natured, cooperative, sympathetic, trusting, and forgiving (vs. rude,
selfish, hostile, uncooperative, and unkind)
Extraversion . . . sociable, talkative, optimistic, ambitious, assertive, reward-seeking, outgoing, and
energetic (vs. introverted, shy, reserved, quiet, and unadventurous)
Conscientiousness . . .organized, responsible, dependable, neat, efficient, and achievement-oriented (vs.
disorganized, lazy, irresponsible, careless, and sloppy)
Emotional Stability . . . calm, self-confident, stable, resilient, and well-adjusted (vs. neurotic, nervous,
insecure, fearful, and anxious)
Table 2. Schwartz Value Taxonomy.
Construct Description/Items: Individuals who value this believe in the importance of . . .

Power . . . being in charge of people and resources and having money (social power, wealth, authority)
Achievement . . . socially recognized successes (ambition, competence)
Hedonism . . .sensual pleasure (fun, enjoying life)
Stimulation . . .having stimulating experiences (daring, exciting life)
Self-direction . . .independence of thought and action (creativity, freedom, independent, curious)
Universalism . . . promoting the welfare of all people and nature (equality, social justice, protecting the environment)
Benevolence . . .promoting the welfare of people you are close to (helpfulness, loyalty, honesty, forgiving)
Conformity . . .controlling impulses to fulfill others’ expectations (self-discipline, obedience)
Tradition . . .maintaining traditions (moderation, respect for tradition, devout)
Security . . .safety and security of self, family, and nation (family security, social order, clean)
importance, so that different people consider some values as

more important than others. The most widely used model of
values is the Schwartz’s (1992) Value Theory that identifies
10 broad values based on the motivations underlying them.
Descriptions of these values are provided in Table 2.
According to Schwartz’s theory and ample empirical evi-
dence, values are structured in a circle based on their inter-
relationships, such that values that are more positively
correlated are closer to one another and are thought to be
based on compatible motivations (see Figure 1).For exam-
ple, both self-direction and stimulation values are based on
the motivation for novelty and are therefore positively cor-
related and adjacent to one another in the value circle. Values
that emanate from opposite sides of the circle are negatively
correlated and are thought to be based on conflicting motiva-
tions. For example, self-direction values stem from the moti-
vation for independent thought and action that conflict with
the motivation to fulfill others’ expectations, the latter under-
lying conformity values. The 10 values can be further
grouped into four higher order types of values organized on
two bipolar dimensions: self-enhancement versus self- Figure 1. The theoretical structure of values (Schwartz, 1992).
transcendence, and openness to change versus conservation
(see Figure 1). This values structure has been examined in
more than 75 countries worldwide and has been found to be Although traits and values are conceptually similar,
largely universal (Schwartz, 2011b). researchers describe several distinctions between the two

Parks-Leduc et al. 5
constructs (for reviews, see Bilsky & Schwartz, 1994; Hitlin whether the term “personality” is a reference to temperament
& Piliavin, 2004; Parks & Guay, 2009). The most basic dif- only, or to both temperament (traits) and character (values).
ference between traits and values is that traits are descriptive Many researchers do not seem to distinguish between
variables whereas values are motivational variables. That is, traits and values; they adhere to the view that all meaningful
traits describe how individuals tend to feel, think, and aspects of personality are subsumed under traits (Buss,
behave. They are therefore summaries of an individual’s 1989), or that traits and values are different ways of measur-
responses and behaviors. Unlike traits, values express a per- ing the same thing. Others blur the distinction between traits
son’s motivations that may or may not be reflected in behav- and values. For example, the HEXACO personality inven-
ior (Roccas, Sagiv, Schwartz, & Knafo, 2002). For example, tory (K. Lee & Ashton, 2004) is presented as a six-factor
a creative person (trait) tends to engage in creative thinking model of personality traits. Yet it includes honesty, which
and in creative acts; otherwise, this person would not be most values researchers would consider to be a value rather
labeled as having the trait of creativity. But valuing creativity than a trait. In contrast, others view traits and values as
may or may not result in creative thinking or behavior. entirely separate constructs that exist at the same level of
Valuing creativity means that the person would like to be abstraction and prediction; these theorists tend to view psy-
creative and thinks that creativity is important, whether or chological needs as antecedents to both (e.g., Parks & Guay,
not he or she acts on this value. Hence, although it makes 2009; Roccas et al., 2002). They also tend to define personal-
sense to expect that most creative people will view creativity ity as the aggregate of traits only (not values).
as important in their lives, the trait and the value are not iden- Other researchers view traits and values as different com-
tical, and people can have different scores on a trait and a ponents of personality (e.g., Caprara, Alessandri, &
value that share similar content. Eisenberg, 2012; Saroglou & Munoz-Garcia, 2008), drawing
Many researchers (e.g., Olver & Mooradian, 2003) also on two integrative models of personality. The first model
propose the distinction that traits are more biologically based suggests three levels of personality components, differing in
(Goldberg, 1993; McCrae & Costa, 2008), whereas values their level of contextualization (McAdams, 1995; see also
are more of a product of a person’s environment, including Sheldon, 2004, for a broader variation of this model). In this
culture, education, parental upbringing, and life events model, traits are located in the first level as non-contextual-
(Rokeach, 1973). We believe that this theoretical distinction ized components of personality, whereas values are part of
has some merit, although it is most likely an oversimplifica- the second level of more contextualized elements of person-
tion, and additional research is needed to test the accuracy of ality (the third level has to do with one’s life narratives and
this claim. Although traits are known to be influenced by personal identity; McAdams, 1995).
genetics, they do vary somewhat by culture and are influ- The second integrative model suggests that traits are basic
enced by environmental variables in addition to genetics tendencies that have a biological basis and that traits influence
(e.g., Heine & Buchtel, 2009; Kandler, 2012). In addition, characteristic adaptations, which include values (McCrae &
research on heritability suggests that values have genetic ori- Costa, 2008). In this model, values are influenced both by
gins in addition to environmental ones (Knafo & Spinath, traits and by external influences, such as culture and life
2011; Schermer, Vernon, Maio, & Jang, 2011). Understanding events. Thus, values are influenced by traits but not solely
the relationships between traits and values has the potential determined by them. To illustrate, if an individual is naturally
to add clarity to continued research in this area. creative (trait), he or she might also value creativity as an
important life goal to pursue. But this relationship is not deter-
The Nature of the Relationships ministic—a person might value creativity even if he or she is
not creative, perhaps as a result of culture or upbringing.
Between Traits and Values Neither of the integrative models suggest any reason to
Researchers differ in the way they view the nature of the expect strong links between the levels of personality (i.e.,
relationships between traits and values. They also differ on McAdams, 1995; McCrae & Costa, 2008). Therefore, if
how they believe traits and values fit within the overall con- traits and values are related, the relationships are not likely to
ceptualization of characteristics of the individual (which is be particularly strong. These models also both view traits as
often broadly termed “personality,” even when it includes antecedent to values (see also Wijnen, Vermeir, & Van
characteristics beyond traits). Although both traits and val- Kenhove, 2007). Yet, values might also influence traits
ues share a common heritage in the lexical hypothesis (the (Roccas et al., 2002). Specifically, as values motivate behav-
idea that all important descriptors of an individual will be ior, if a value (e.g., benevolence) leads to recurrent behavior
encoded in language, and can therefore be culled from a dic- (e.g., caring for one’s younger siblings), this recurrent behav-
tionary), the two constructs have been examined separately ior will later become a trait, because traits include recurrent
since at least the 1930s, when Allport (1937) took pains to patterns of behaviors.
remove values items from his research studies on personal- To summarize, some researchers do not clearly distin-
ity. He referred to traits as temperament and values as char- guish between traits and values, some view them as distinct
acter (Allport, 1937); these descriptors resurface at times and separate constructs, and some view them as loosely
(see, for example, Cloninger, 1994), but it is not always clear related components at different levels of personality. This

myriad of views, and the confusion it creates in the literature, terms, so we do not expect it to be similar to emotional stabil-
needs to be acknowledged. As meta-analysis relies on corre- ity in its relationships with values.
lations, we do not attempt to provide explanations regarding Additional support for this premise comes from research
the direction of relationships or superiority of one model that examines the neuroscience of personality. Cloninger
over the others. However, before integrating traits and values (1994) developed a personality taxonomy that includes seven
into a comprehensive understanding of the individual, estab- major personality traits based on the different neurobiologi-
lishing the patterns and magnitudes of these relationships is cal processes that occur in the brain during trait expression.
an important first step. He retained Allport’s terms of temperament and character as
separate components of personality and defined tempera-
ment as “automatic associative responses to emotional stim-
Two Sources for the Strength of the uli that determine habits and moods, whereas character refers
Relationships Between Traits and to the self-aware concepts that influence our voluntary inten-
Values tions” (p. 266). The temperament traits relate to emotional
Although the links between traits and values are not expected and automatic processes—activities that are primarily asso-
to be strong, some of the traits of the FFM may be more ciated with the mid-brain. The character traits involve the
closely related to values than are others, as found in previous frontal lobe (frontal cortex) to a greater degree than the tem-
research. Although researchers have developed hypotheses perament traits, suggesting that the character traits are linked
for these relationships based strictly on content similarity, we to higher levels of cognitive processing. Neurobiological
propose that the strength of the relationships between traits research therefore supports the premise that traits vary in
and values may be based on two sources of similarities— terms of the extent to which they are affectively or cogni-
similarities in the nature of particular traits and values and tively based. In a subsequent study, De Fruyt, Van De Wiele,
similarities in the content of particular traits and values. and Van Heeringen (2000) correlated Cloninger’s traits with
the Big Five. They found that emotional stability (neuroti-
cism in their study) was strongly correlated with one of
Similarities in the Nature of Traits and Values Cloninger’s temperament traits (mid-brain), supporting our
All values are inherently cognitive (see, for example, Schwartz expectations for a weak correlation for emotional stability
& Bilsky, 1987). Yet unlike values, traits may vary in the extent with values. Openness to experience, extraversion, and con-
to which they are based on cognition (recall that traits are scientiousness all exhibited moderate positive correlations
described as recurrent patterns of thought, behavior, and affect; with both temperament and character traits, whereas agree-
McCrae & Costa, 2003). Supporting the stronger cognitive ableness showed a strong positive correlation only with a
nature of values compared with traits, Roccas et al. (2002) character trait (frontal lobe), leading us to expect relatively
found that values predicted a cognitively based outcome better strong links for agreeableness with values.
than traits, and traits predicted an affectively based outcome In summary, the FFM traits vary in the extent to which
better than values. We expect that traits that are more cognitive they are cognitively oriented, based on both research on the
in nature will tend to have stronger relationships with values, item-level content of Big Five traits and research on the neu-
because values are cognitive in nature. In contrast, values are robiological processes involved in personality expression.
not emotional variables—although they can elicit negative Taken together, we expect that openness to experience should
emotions when they are violated, or positive emotions when have the strongest links with values, followed by agreeable-
fulfilled (Locke, 1997; Schwartz, 1992; Sheldon & Elliot, ness. Emotional stability should have the weakest links with
1999). We therefore expect weaker correlations with values for values, and conscientiousness and extraversion should fall
traits that have a large emotional component. somewhere in between.
Which traits in the FFM are the most cognitive and which
ones are the most affective? Using judges’ ratings of items
Similarities in the Content of Traits and Values
from multiple Big Five inventories, Pytlik Zillig, Hemenover,
and Dienstbier (2002) found that openness to experience had As research has demonstrated, the strength of relations
a consistently strong cognitive component, and emotional between traits and values should also be somewhat determined
stability had a strong affective component. The remaining by content similarity when comparing each trait with each
three traits were all described primarily by behavioral items value. We briefly review previous hypotheses for expected
(defined as overt, directly observable actions). This finding links between traits and values, focusing on links that have
suggests that openness to experience should have the stron- been hypothesized by at least two of these previous articles.
gest relationships with relevant values, whereas emotional
stability should have the weakest. The remaining three traits Openness to experience. As stated above, we expect openness
should fall in between. Note that although extraversion is to experience to have the strongest and most coherent pat-
often defined as an affective trait, the typical measurement of terns of relations with values, as compared with the other
extraversion is primarily in behavioral rather than affective traits in the FFM. The content of this trait dimension is quite

similar to the bipolar higher order value dimension of open- goal-directed behavior (e.g., Fiske, 1994; Hogan & Ones,
ness to change versus conservation (see Figure 1), which 1997; John & Srivastava, 1999). McCrae and John (1992)
contrasts openness to new ideas and experiences with a pref- suggested that conscientiousness has two major components,
erence for rigid rules of actions and thoughts (Rohan, 2000). each compatible with different values. The first is a proactive
Hence, individuals who score highly on openness to experi- aspect of conscientiousness, which is related to the motiva-
ence are likely to value stimulation and self-direction and to tion for success according to social standards (Costa &
ascribe low importance to conformity, tradition, and security McCrae, 1988). This motivation is also expressed in achieve-
values (Luk & Bond, 1993; Olver & Mooradian, 2003; Roc- ment values (Luk & Bond, 1993; Roccas et al., 2002). The
cas et al., 2002). We expect the relationship between open- second aspect of conscientiousness is inhibitive and is related
ness to experience and self-direction to be particularly strong, to the motivation for impulse control (Costa & McCrae,
as both relate to curiosity and creativity. Individuals who 1988), expressed in conformity values (Olver & Mooradian,
score highly on openness to experience are also likely to 2003; Roccas et al., 2002). Hence, conscientiousness should
value universalism, as universalism values include tolerance be related to achievement (+) and conformity (+) values.
and openness to ideas and behaviors that are different from
what one is accustomed to (Olver & Mooradian, 2003; Roc- Emotional stability. People who score highly on this trait tend
cas et al., 2002; see also Schwartz, 1992). To summarize, we to be less prone to negative affect (Costa & McCrae, 1988;
expect openness to experience to exhibit relationships with John & Srivastava, 1999). They are not easily distressed and
the following values: stimulation (+), self-direction (+), uni- have healthy coping strategies (Gunthert, Lawrence, &
versalism (+), conformity (−), tradition (−), and security (−). Armeli, 1999). As this trait is primarily affective, and as val-
ues do not tend to have direct relations to well-being or dis-
Agreeableness. Agreeable individuals are oriented toward tress (Roccas et al., 2002; Sagiv, Roccas, & Hazan, 2004),
helping others and cooperating with them (e.g., Graziano & this trait is likely to be unrelated to values.
Eisenberg, 1997; John & Srivastava, 1999). This orientation
is similar to the motivation underlying benevolence values,
which aims to enhance the well-being of people in one’s
Sinusoid Patterns of Correlations
immediate social environment (family, friends, etc.). Coop- According to values theory (Schwartz, 1992, 1996), if theory
eration with others requires some willingness to adapt to predicts that a certain variable (such as religiosity) is associ-
group norms as well. Conformity values express the motiva- ated with a certain value (such as tradition), this variable
tion to fulfill the expectations of others in one’s social groups. should also exhibit positive relations with compatible types
Similarly, tradition values express the motivation to maintain of values (those that are adjacent to it on the circle; in this
the customs, traditions, and hierarchy of one’s social groups. example, conformity, security, and benevolence) and nega-
Therefore, both conformity and tradition are likely to be pos- tive relations with conflicting types of values (those that are
itively related to the cooperative aspect of agreeableness opposite to it on the circle; in this example, hedonism and
(Luk & Bond, 1993; Olver & Mooradian, 2003; Roccas et stimulation). Because the value circle is based on a motiva-
al., 2002). In contrast, power values express the motivation tional continuum, related variables should have a systematic
for dominance and control, sometimes at the expense of oth- pattern of correlations with the entire value system. The vari-
ers. They are therefore incongruent with agreeableness able of interest should be most positively related to the value
(Olver & Mooradian, 2003; Roccas et al., 2002). We thus that is most clearly positively linked with it, and the correla-
expect agreeableness to exhibit relationships with the fol- tions should become less and less positive as one moves
lowing values: benevolence (+), conformity (+), tradition around the circle and away from that value, eventually mov-
(+), and power (−). ing to negative relationships. The negative relationships
should reach their maximum with the value on the circle that
Extraversion. Extraverts need stimulation. They are highly directly opposes the value with the strongest positive corre-
energetic, ambitious, assertive, and reward-seeking (e.g., lation. If one graphs the correlations, with the values pro-
Costa & McCrae, 1992; John & Srivastava, 1999). This ten- vided on the graph from left to right in order as one moves
dency to seek rewards and to be ambitious is highly compat- clockwise around the circle, the subsequent line should form
ible with achievement values (Luk & Bond, 1993; Roccas et a sinusoid curve (a sine wave, with one major peak and one
al., 2002). In addition, being energetic and having a high major valley).
need for stimulation are highly compatible with stimulation We expect openness to experience, agreeableness, and
values (Bilsky & Schwartz, 1994; Luk & Bond, 1993; Roc- extraversion to display a sinusoidal pattern of correlations
cas et al., 2002). We therefore expect extraversion to be with the full set of values. Specifically, the highest positive
related to achievement (+) and stimulation (+) values. correlation should be with the corresponding value accord-
ing to the research reviewed above, and the correlations with
Conscientiousness. This trait dimension describes socially other values should decrease monotonically going around
prescribed impulse control that facilitates task and the circle of values, creating a sinusoid shape. We do not

expect this pattern with conscientiousness, because it should 1997; Schwartz, 2011b). Finding that these relationships are
be most positively related to two non-adjacent values universal would support the view that the links between traits
(achievement and conformity). and values are based on processes that are largely unaffected
by culture. This moderator analysis is therefore particularly
interesting, as its results may inform our fundamental under-
Possible Moderators
standing of the relationships between traits and values.
What might affect the strength of relations between traits and
values? We consider five moderators, starting with modera-
Values Instruments
tors based on theoretical considerations followed by modera-
tors based on methodological issues. Traits and values have been measured by different instru-
ments, which makes results difficult to generalize because
differences across studies may be the result of using different
Culture scales that assess traits or values in a somewhat different
Culture is often defined as a shared system of meaning (e.g., fashion. However, the variety of scales used, particularly for
Smith & Bond, 1998), and cultures differ in the meanings traits, meant that few studies used the same measures. As a
they attribute to events. Different cultures may result in dif- result, we could not conduct fully hierarchical moderator
ferent trait–value relations. Of the possible cultural dimen- analyses considering both the traits measure and the values
sions that one could consider, individualism versus measure. For values, however, most studies used one of two
collectivism and tightness versus looseness seem to have the instruments: the Schwartz Value Survey (SVS; Schwartz,
potential to moderate the relationships between traits and 1992) or the Portrait Values Questionnaire (PVQ; Schwartz
values. et al., 2001). We therefore conducted a basic moderator anal-
Individualism versus collectivism is the most studied cul- ysis using the values instrument as a moderator.
tural dimension (Taras, Kirkman, & Steel, 2010). According The PVQ was developed from the SVS with the intention
to Hofstede (1980), individualistic cultures emphasize indi- of creating an instrument that was less abstract and less cog-
viduality—the uniqueness of the individual and his or her nitively complex. Rather than rating the importance of each
right to pursue personal goals. In contrast, collectivistic cul- value, respondents read descriptions of individuals (i.e., por-
tures emphasize the importance of one’s group and, as a traits) in terms of values and rate the extent to which the
result, the obligations to one’s group. Cultures also differ in described person is similar to them. An important difference
strength—in the pervasiveness of social norms and in the tol- between these instruments for the purposes of the current
erance to deviant behavior from those norms (Pelto, 1968). investigation involves how directly they measure values. The
Tight societies have a culture with very strong norms and SVS measures values directly, because participants rate
severe sanctions for the violation of those norms, whereas abstract goals in terms of how important they are as a guiding
loose cultures have more ambiguous norms and are more principle in their lives. Unlike the SVS, in the PVQ, partici-
permissive of possible deviance (Gelfand, Nishii, & Raver, pants read a description of a person and rate how similar that
2006; Gelfand et al., 2011). Cultures that are tighter or more person is to them. The description of the person includes two
collectivistic are thus likely to encourage more normative sentences. Often, one sentence describes the person in terms
value endorsement, whereby individuals within the culture of a goal that is important to him or her. This is a direct mea-
would be more likely to subscribe to the dominant values of sure of a value. For example, the first part of an item that
the culture, rather than those consistent with the individual’s measures security values is “It is important to her to live in
traits. In contrast, members of looser cultures or more indi- secure surroundings.” Yet the other sentence sometimes
vidualistic cultures might be more likely to endorse values involves trait-like elements. For example, the second sen-
that are consistent with their individual personality traits, tence of the item above is “She avoids anything that might
leading to higher correlations between values and traits in endanger her safety.” As the PVQ has trait-like elements, it is
more individualistic or looser cultures (for a similar argu- likely to result in inflated trait–value correlations compared
ment regarding the links between values and behavior, see with the SVS.
Roccas & Sagiv, 2010).
Alternatively, the relationships between traits and values
Personality Instruments
may be universal as they may stem from the same psycho-
logical processes across cultures. Indeed, although ample Although 11 personality instruments were utilized in the
research demonstrates that the means of measures of values studies that we examine, two were used frequently enough to
and traits vary cross-culturally (see, for example, Allik, include them in a moderator analysis: the Neuroticism,
2012, regarding traits; Schwartz, 2011b, regarding values), Extraversion, and Openness to Experience (NEO; Costa &
the links between the two systems may be universal, just as McCrae, 1992) and the Big Five Inventory (BFI; John,
the intercorrelations among traits and those among values Donahue, & Kentle, 1991). Arguably the best-validated of
are largely universal (see, for example, McCrae & Costa, personality inventories, the NEO is commercially available

with differing versions based on age and gender, and it is Other Moderators
widely used in a variety of settings (John & Srivastava,
1999). McCrae and Costa developed the NEO initially from Many studies are conducted with university students, who
the work of previous personality researchers, most notably are a more homogeneous group than the general population
Cattell. They later added Agreeableness and and can therefore generate somewhat different results in
Conscientiousness, creating the NEO-PI-R. This inventory studies compared with the general population (Peterson,
provides 240 items measuring six facets each for each of the 2001). We therefore examined study population as a poten-
five factors. Items are provided in sentence form, with partial moderator. Because the studies included varied greatly in
ticipants rating their level of agreement (or disagreement). sample size, we also tested whether sample size moderates
McCrae and Costa subsequently developed a short version trait–value correlations. Finally, as is often done in meta-
(60 items; the NEO-FFI) that only assesses the five factor- analyses, we tested whether publication status moderates
level constructs. Although it is not identical, the NEO-PI-R trait–value relations to account for a possible publication
and the NEO-FFI are substantially correlated (John & bias.
Srivastava, 1999), hence we group them together in our mod-
erator analysis. Meta-Analytic Method
Researchers developed the BFI (John et al., 1991) in an
effort to achieve some convergence among differing views of Literature Search
the content of the factors of the Big Five, which were con- To locate articles for inclusion, we conducted searches using
ceptualized somewhat differently by different researchers. major electronic databases such as PsycINFO. We used the
Ten judges with psychology backgrounds reviewed the lit- keywords “personality,” “traits,” and “values” to search for
erature on all the existing versions of the Big Five, and cat- related articles. We additionally scanned the reference sec-
egorized 300 items from the Adjective Check List (ACL) tions from the articles produced by the initial search to see
into five categories. The researchers then retained the items whether additional studies could be located in this fashion.
for which there was a high level of agreement and from there, For studies located that did not include correlations tables,
culled the list down to 44 representative items. Because we contacted authors to request their raw data. Finally, we
adjectives can sometimes have more than one meaning, they contacted researchers in this domain to request unpublished
created short phrases for items (John & Srivastava, 1999). studies and posted requests for data on related list serves.
The BFI and the NEO are highly correlated but not exactly Studies were collected up until April 2013; the search yielded
the same; corrected correlations range from .83 (Extraversion) 88 possible studies.
to .97 (Agreeableness; John & Srivastava, 1999). Because
the NEO and the BFI are highly correlated and because both Inclusion criteria. We restricted our meta-analysis to the dom-
rely on longer descriptions (sentences or phrases), we expect inant taxonomies for categorizing traits and values reviewed
them to yield fairly similar results. above—the FFM and the Schwartz values theory (see Tables
1 and 2), and to those studies that examined traits and values
at the individual (rather than group) level. One study (Wijnen
Statistical Adjustments for Values Scale Use
et al., 2007) divided openness to experience into two facets,
Values research requires a somewhat different approach to labeled “self-rated intelligence” and “creativity.” Correla-
data analysis compared with traits. People make decisions tions were averaged across these facets to create a factor-
about how to behave not based on the absolute importance of level correlation for the meta-analysis. One study (von
a value but rather on its importance relative to other values. In Collani & Grumm, 2009) grouped values into four broader
other words, we cannot predict behavior based solely on how categories rather than using the 10 value types; it was elimi-
high a person’s score is on benevolence values; we need to nated from our study (results of analyses for the four higher
know how high it is relative to other values that the person also order value dimensions are available in supplementary files
endorses. Moreover, individuals differ in their use of the scale at http://psr.sagepub.com/supplemental).
such that some people tend to attribute high importance to val- Some studies were excluded because they did not provide
ues across items, and some low. Schwartz (1992) therefore the necessary quantitative data for a meta-analysis.
recommended controlling for mean importance of values. Due Specifically, 8 studies did not include a correlation matrix
to these theoretical reasons, we anticipate that the expected (and we were unable to obtain a correlation matrix from the
trait–value relations will be more accurate when scale-use ten- author[s]); two studies included only significant correlations;
dency of values is controlled. Whereas the majority of studies and 1 study used canonical correlations. These were
controlled for scale use in the correlation matrix, some reported excluded. This yielded a total of 60 studies (listed in Table
only the zero-order correlations (in some cases, these studies 3), including 29 from published articles, 1 book chapter, 25
subsequently controlled for scale use in their regression equa- unpublished data sets, 2 data sets from conference presenta-
tions or path estimates, which would typically be viewed as tions, and 3 samples from dissertations. Several studies had
more critical to hypothesis testing). very large sample sizes; 10 studies had samples of more than

Table 3. Studies Included in the Meta-Analysis; Possible Moderators.

Sample Method of
Study N Personality measure Values measure Country population analysisa Publication status
1. Arthaud-Day, Rode, and 582 IPIP SVS United States Students Partialled Published article
Turnley (2012)
2. Bardi (2005) 103 BFI SVS United Kingdom Students Partialled Unpublished
3. Bardi (2008) 677 TIPI SVS United Kingdom Students Partialled Unpublished
4. Bardi, Buchanan, Goodwin, 136 BFI SVS United Kingdom Police trainees Partialled Published article
Slabu, and Robinson (2014)
5. Bardi, Bull, and Brown (2008) 65 BFI SVS United Kingdom Students Partialled Unpublished
6. Bardi and Guerra (2011) 163 BFI SVS World Students Partialled Published article
7. Bardi and John (2006) 110 BFI SVS United Kingdom Students Partialled Unpublished
8. Bardi, Lee, Hofmann-Towfigh, 128 BFI SVS United Kingdom Students Partialled Published article
and Soutar (2009) (a)
9. Bardi et al. (2009) (b) 196 BFI SVS United Kingdom Students Partialled Published article
10. Bardi, Levontin, and John 586 BFI, NEO, and SVS and PVQ United States Students Partialled Unpublished
(2011) Saucier’s mini-
markers
11. Bardi, Loader, Keen, and 232 BFI SVS World Students Partialled Unpublished
Martin (2004)
12 Barrick, Giluk, Shaffer, and 126 PCI PCVS United States Students Correlations Unpublished
Stewart (2006)
13. Barrick et al. (2005) 166 PCI SVS United States Students Correlations Unpublished
14. Blickle, Schlege, Fassbender, 226 NEO SVS Germany 76 white-collar Correlations Published article
and Klein (2006) criminals; 150
managers
15. Burns and Postlethwaite 159 PCI SVS United States Students Correlations Unpublished
(2007)
16. Caprara, Schwartz, Capanna, 4,349 BFQ PVQ Italy General Unknown Published article
Vecchione, and Barbaranelli population
(2006)
17. Caprara and Vecchione (2006) 944 BFQ PVQ Italy General Correlations Published article
population
18. Caprara, Vecchione, and 576 BFQ PVQ Italy Students Unknown Published article
Schwartz (2009)
19. Cohrs, Kielmann, Maes, and 512 NEO SVS World General Partialled Published book
Moschner (2005) population chapter
20. Collins and Blum (2011) 199 IPIP SVS United States Students Centered Unpublished
21. Dirilen-Gumus (2010) (a) 386 BFI PVQ United States Students Unknown Published article
22. Dirilen-Gumus (2010) (b) 382 BFI PVQ Turkey Students Unknown Published article
23. Dirilen-Gumus, Cross, and 278 BFI PVQ United States General Correlations Published article
Donmez (2012) population and
Students
24. Dollinger, Leong, and Ulicni 275 NEO RVS United States Students Standardized Published article
(1996) (a)
25. Dollinger et al. (1996) (b) 198 NEO RVS United States Students Standardized Published article
26. Goldberg (2008) 698 IPIP SVS United States General Partialled Unpublished
population
27. Haslam, Whelan, and Bastian 180 IPIP SVS Australia Students Partialled Published article
(2009)
28. Knafo (2007) 278 Saucier’s mini- SVS Israel Students Correlations Unpublished
markers
29. Kusdil (2000) (a) 147 NEO SVS United Kingdom General Partialled Dissertation
population
30. Kusdil (2000) (b) 329 NEO SVS Turkey General Partialled Dissertation
population
31. Lonnqvist and Versakalo 498 NEO SVS Finland Military (Reserve Correlations Unpublished
(2005) Officer School)
32. Lonnqvist and Walkowitz 120 Short Five PVQ Germany Students Correlations Unpublished
(2010)
33. Luengo Kanacri, Rosa, and Di 563 BFQ PVQ Italy General Correlations Published article
Giunta (2012) population and
Students
(continued)

Table 3. (continued)
Sample Method of
Study N Personality measure Values measure Country population analysisa Publication status
34. Luk and Bond (1993) 114 NEO SVS China Students Partialled Published article
35. MyPersonality—100 items 2,986 NEO SVS World Internet Partialled Unpublished
(Kosinski & Stillwell, 2011)
36. MyPersonality—20 items 1,487 NEO SVS World Internet Partialled Unpublished
(Kosinski & Stillwell, 2011)
37. MyType (Wilson, Gosling, & 15805 BFI PVQ World Internet Correlations Unpublished
Graham, 2012)
38. Olver and Mooradian (2003) 255 NEO and Saucier’s SVS United States Students Correlations Published article
mini-markers
39. Parks (2007) 367 IPIP RSVS United States Students Correlations Dissertation
40. Parks (2008) 74 IPIP RSVS United States Employees Correlations Unpublished
41. Parks-Leduc, Pattie, Pargas, 420 IPIP RSVS United States Students Correlations Unpublished
and Eliason (2014)
42. Poling, Woehr, Gorman, and 266 Saucier’s Unipolar PVQ United States Students Correlations Conference
Arciniega (2006) Markers paper
43. Roccas, Sagiv, and Porat 217 Saucier’s mini- PVQ Israel Students Partialled Unpublished
(2007) markers
44. Roccas, Sagiv, Schwartz, and 246 NEO SVS Israel Students Partialled Published article
Knafo (2002)
45. Saroglou and Munoz-Garcia 256 NEO SVS Spain Students Partialled Published article
(2008)
46. Stankov (2007) 1,255 IPIP SVS United States Students Unknown Published
article
47. Steca, Monzani, and Greco 4,285 BFQ PVQ Italy General Unknown Unpublished
(2011) population
48. Sverdlik and Sagiv (2007) 272 Saucier’s mini- SVS Israel Students Partialled Unpublished
markers
49. Trapnell (2007) 249 BFI PVQ Canada Students Partialled Unpublished
50. Uziel, Sagiv, and Roccas (2007) 170 Saucier’s mini- SVS Israel Students Unknown Unpublished
markers
51. Vecchione, Alessandri, 1,675 Unique to study PVQ Italy General Correlations Published article
Barbaranelli, and Caprara population
(2011)
52. Vecchione, Caprara, Schoen, 981 BFQ PVQ Italy General Partialled Published article
Castro, and Schwartz (2012) population
(a)
53. Vecchione et al. (2012) (b) 352 BFQ PVQ Spain General Partialled Published article
population
54. Vecchione et al. (2012) (c) 190 NEO PVQ Germany General Partialled Published article
population
55. Vecchione and Mebane (2007) 1,089 BFQ PVQ Italy General Correlations Published article
population
56. Wijnen, Vermier, and Van 311 Mervielde’s Big Five SVS Belgium Students Correlations Published article
Kenhove (2007) Scale
57. Wolfradt and Dalbert (2003) 212 NEO SVS Austria 104 students; Correlations Published article
107 general
population
58. Xu (2005) 126 IPIP SVS United States Employees Correlations Conference
and China paper
59. Yik and Tang (1996) 216 Unique to study SVS Hong Kong Students Correlations Published article
60. YourMorals.org (Graham et 7,543 BFI SVS World Internet Correlations Unpublished
al., 2011)
Note. BFI = Big Five Inventory; BFQ = Big Five Questionnaire; IPIP = International Personality Item Pool; NEO = Neuroticism, Extraversion, and Openness to Experience; PCI =
Personal Characteristics Inventory; PCVS = Pairwise Comparison Values Survey; PVQ = Portrait Values Questionnaire; RSVS = Revised Schwartz Value Survey; SVS = Schwartz
Value Survey; TIPI = Ten Item Personality Inventory.
a. Method of analysis refers to how the researcher(s) presented the correlations. While some researchers just provide the correlations without statistical adjustment, many
values researchers control for scale use by partialling out the mean value score or via some other method.

1,000 participants. The largest 4 data sets were data gathered Table 4. Average Reliabilities (Coefficient α).
from web sites and social networks—MyType (N = 15,805;
Construct Reliability
Wilson, Gosling, & Graham, 2012), YourMorals.org (N =
7,543; Graham et al., 2011), and 2 from MyPersonality.org Openness to experience .76
(N = 2,986 and 1,487; Kosinski, Stillwell, & Graepel, 2013). Agreeableness .73
MyType and MyPersonality.org can be accessed from Extraversion .78
Facebook and include a variety of surveys designed to tell Conscientiousness .80
people more about themselves, including a Myers–Briggs Emotional stability .81
test, a Big Five personality traits test, the SVS, and others. Power .72
MyType surveys have been taken by more than 17,000 users. Achievement .76
YourMorals.org is a site developed by social psychologists to Hedonism .72
enhance the study of various topics related to moral, social, Stimulation .73
Self-direction .67
and political psychology; it also includes a variety of sur-
Universalism .78
veys, including a Big Five measure and the SVS (Graham,
Benevolence .72
Haidt, & Nosek, 2009; Graham et al., 2011).
Conformity .68
Tradition .63
Meta-analytic procedure. Meta-analytic procedures were Security .65
based on Hunter and Schmidt (2004). We corrected correla-
tions for unreliability and sampling error (SPSS syntax
adopted from Field & Gillett, 2010). Two studies collected
data on traits or values using more than one scale. This the moderators, but we are cautious in our conclusions. Our
yielded two correlations that were not independent (because cultural moderators were modeled as continuous variables;
they came from the same respondents). For one of these stud- all other moderators were categorical.
ies (Bardi & John, 2006), composite correlations were calcu-
lated before inclusion in the meta-analysis. Composites were Culture. The primary studies included in the meta-
calculated based on formulas provided by Hunter and analysis were conducted in 13 countries in North America,
Schmidt (2004). Because composite calculations require that Europe, and Asia. This provided cross-cultural variabil-
intercorrelations between all scales be provided, we were ity that enabled us to consider whether cultural differences
unable to perform the same procedure for the other study could create a moderator effect, that is, whether trait–value
(Olver & Mooradian, 2003). For this study, therefore, the relationships might vary by culture. Several studies were
average correlations were calculated and these were included conducted across multiple countries and cultures; these were
instead. removed from this set of analyses.
Most studies reported coefficient alpha reliabilities, but To conduct the individualism/collectivism moderator
a few did not. When possible, correlations were corrected analysis, studies were assigned a number representing their
individually. When reliability estimates were not provided, level of individualism versus collectivism, based on data
artifact distribution was used to provide a mean reliability; from Hofstede (1980; for previous use of this method, see
this was then used to correct for unreliability (see Table 4 Bardi & Guerra, 2011; Suh, Diener, Oishi, & Triandis, 1998).
for average reliabilities). None of the studies reported reli- Based on his empirical data, Hofstede assigned culture scores
ability estimates accounting for transient error, and as a ranging from 1 to 100, with 1 representing very high collec-
result, they are likely to overestimate the true reliability of tivism and 100 indicating very high individualism. For
the scales. No corrections were made for range restriction, example, the United States is very individualistic and has a
as data were not available to calculate range restriction on score of 91. Spain has a score of 51, about the middle of the
values. Past research on traits has found little evidence of scale. Hong Kong, with a much more collectivist culture, has
range restriction, however, and the same is likely to be true a score of 25. Our studies were skewed toward the high end
of values. of the scale; the majority of the studies were conducted in
countries with moderate to high individualism scores.
Moderator analyses. Hunter and Schmidt (2004) recom- Because this moderator was modeled as a continuous vari-
mended hierarchical moderator analyses when a sufficient able, weighted least squares (WLS) regression was used for
number of studies exists. If moderators are related and the the moderator analysis rather than sub-grouping, as recom-
analyses are not hierarchical, failing to conduct hierarchical mended by Steel and Kammeyer-Mueller (2002). With WLS
moderator analyses can lead to false conclusions (i.e., mod- regression, the correlation is considered the dependent vari-
eration could be attributed to the wrong construct). However, able, and the moderator the independent variable, in a regres-
the number of studies was insufficient to perform hierarchi- sion analysis (with each study weighted by the inverse of the
cal moderator analyses considering all moderators. As a sample error variance). Significant betas indicate that mod-
result, we conducted separate moderator analyses for each of eration has occurred.

To examine the tightness/looseness of culture as a mod- method, and 22 clearly reported zero-order correlations (we
erator, studies were assigned a tightness score based on data were unable to code some studies due to missing information
from Gelfand and colleagues (2011), with higher scores rep- in the articles; these were eliminated from this analysis). Not
resenting a tighter culture. For example, Turkey had a rela- all studies controlled for participants’ scale-use tendency in
tively high score of 9.2, representing strong cultural norms. the same way: The majority reported correlations after par-
Israel, in contrast, scored a 3.1, suggesting more acceptance tialling out the personal mean value score, but 1 study sub-
of behavior that is inconsistent with cultural norms. Included tracted the mean value score from each value domain score
studies were fairly normally distributed in terms of this cul- (Collins & Blum, 2011), and 2 studies subtracted the mean
tural moderator; most were in the middle, whereas a few value score from each value domain score and then divided
studies were at either end of the range. As in the previous that difference by the standard deviation of the scores (Doll-
analysis, WLS regression was used to conduct the moderator inger, Leong, & Ulicni, 1996). Given that they were intended
analysis. to fulfill the same purpose and probably would have resulted
in similar correlations, we treated these methods as suf-
Values instruments. The SVS has a list of 56 (Schwartz, ficiently similar to group them together for the moderator
1992) or 57 (Schwartz, Sagiv, & Boehnke, 2000) value analyses.
items (e.g., social power, daring) that participants rate as a
guiding principle in their own life on a 9-point scale, from
−1 (opposed to my principles) to 0 (not important) to 7 (of Meta-Analytic Results
supreme importance). The scale is asymmetric to capture
discriminations between values, as all values are desirable
Individual Correlation Estimates
in society (Schwartz & Bardi, 2001). Rather than rating the Tables 5 through 9 provide the results of the main meta-anal-
importance of each value, respondents to the PVQ (Schwartz ysis (one table for each trait). As expected, traits and values
et al., 2001) read two-sentence descriptions of individu- are related in consistent ways. Openness to experience and
als (i.e., portraits) in terms of values and rate the extent to agreeableness, in particular, exhibit several strong relation-
which the described person is similar to them. There are ships with values. The meta-analysis reveals that openness to
two versions of the questionnaire—one for males and one experience is strongly correlated with self-direction (ρ =
for females—to allow for gender-specific pronouns. Sample .52); has moderate positive relationships with stimulation
items include “He likes to be in charge and tell others what to (ρ = .36) and universalism values (ρ = .33); and has a moder-
do. He wants people to do what he says” (power); “She looks ate negative relationship with tradition, conformity, and
for adventures and likes to take risks. She wants to have an security values (ρ = −.31, −.27, and −.24, respectively).
exciting life” (stimulation). The PVQ includes 40 descriptive Agreeableness relates most strongly with benevolence val-
“portraits” (value items); respondents indicate how much the ues (ρ = .61). It also has moderate relationships with power
described person is similar to them on a 6-point scale ranging (ρ = −.42), universalism (ρ = .39), conformity (ρ = .26), and
from 1 (very much like me) to 6 (not like me at all). A total tradition values (ρ = .22).
of 30 studies used the SVS, and 18 used the PVQ (1 of those Extraversion and conscientiousness exhibit fewer signifi-
used the shortened, 20-item version of the PVQ, but was still cant relationships (and no strong relationships) with the
included). Schwartz value domains. Extraversion demonstrates moder-
ate relationships with stimulation, power, achievement, and
Personality instruments. As stated previously, 11 studies hedonism values (ρ = .36, .31, .31, and .20, respectively).
using the NEO (Costa & McCrae, 1992) and 14 studies using Conscientiousness is moderately related to security (ρ = .37),
the BFI (John et al., 1991) were included in this moderator conformity (ρ = .27), and achievement values (ρ = .17).
analysis. The NEO includes either 240 or 60 items depending Finally, emotional stability did not demonstrate any signifi-
on whether or not it is examining facets. A sample item is “I cant relationships with the values domains. In total, 18 of the
am a productive person who always gets the job done” (John 50 correlations yielded generalizable relationships (those for
& Srivastava, 1999); participants rate their level of agree- which the 80% credibility interval around ρ did not include
ment/disagreement with the statement. The BFI includes 44 zero; see Table 10 for a summary of these relationships).
short statements such as “Is original, comes up with new Eleven of the 50 relationships have 90% confidence intervals
ideas,” for which participants utilize a 5-point scale to rate (CI; around mean r) that do not include 0 (see Tables 5-9 for
their level of agreement/disagreement in terms of how well mean r and 90% CIs).
the statement describes them. The strength of the relationships between traits and values
was generally consistent with our theoretical expectations, as
Statistical adjustments to control for values scale use. To the more cognitively based traits tended to have more and
control for scale-use tendency, researchers often partial out stronger relationships with values, and the primarily affec-
the overall mean score of values. Twenty-eight of the stud- tive trait (emotional stability) was unrelated to values. The
ies included in the meta-analysis clearly used a partialling specific relationships were also largely consistent with

Table 5. Meta-Analytic Results for Openness to Experience and Values.
Ten-factor value domains k N Mean r ρ SDρ CVLL CVUL CILL CIUL % Var
Power 52 54,274 −.04 −.06 .11 −.20 .08 −.27 .15 13
Achievement 54 54,747 .08 .11 .12 −.04 .26 −.12 .34 11
Hedonism 53 54,165 .07 .09 .12 −.06 .24 −.14 .32 11
Stimulation 51 53,692 .27 .36 .13 .20 .52 .11 .61 8
Self-Direction 55 54,959 .37 .52 .11 .38 .66 .31 .74 11
Universalism 53 54,165 .25 .33 .10 .20 .46 .12 .53 12
Benevolence 54 54,747 .10 .13 .11 −.01 .27 −.08 .34 13
Conformity 55 54,959 −.20 −.27 .16 −.48 −.07 −.59 .04 6
Tradition 51 53,692 −.21 −.31 .15 −.50 −.12 −.60 −.02 8
Security 54 54,377 −.17 −.24 .17 −.45 −.02 −.56 .09 6
Note. Bolded mean r values have a 95% confidence interval that does not include 0. Bolded ρ values have an 80% credibility interval that does not include
0. k = number of studies; N = total number of individuals across all studies; Mean r = the average of the uncorrected correlations; ρ = the estimated true
score correlation; SDr = standard deviation of mean r; SDρ = the standard deviation of the corrected correlations; CVLL and CVUL = lower and upper
bounds, respectively, of the 80% credibility interval; 90% CV = the 90% Credibility Value—this is the upper limit of the 80% credibility interval; CILL and
CIUL = lower and upper bounds, respectively, of the 95% confidence interval around the corrected mean correlation; % Var = the percentage of variance
that was accounted for by statistical artifacts (sampling error and unreliability of measures).
Table 6. Meta-Analytic Results for Agreeableness and Values.
Power 54 54,599 −.31 −.42 .21 −.70 −.15 −.84 .00 3
Achievement 55 54,946 −.18 −.24 .24 −.54 .07 −.71 .23 3
Hedonism 53 54,165 −.08 −.11 .11 −.24 .03 −.32 .10 13
Stimulation 51 53,692 −.04 −.05 .11 −.19 .09 −.26 .17 13
Self-Direction 55 54,959 −.04 −.07 .19 −.31 .18 −.45 .32 5
Universalism 54 54,364 .29 .39 .12 .23 .54 .15 .62 9
Benevolence 56 55,072 .45 .61 .17 .39 .82 .28 .94 4
Conformity 55 54,959 .18 .26 .11 .12 .39 .05 .47 14
Tradition 51 53,692 .15 .22 .12 .08 .37 .00 .45 13
Security 54 54,377 .00 .00 .17 −.22 .22 −.34 .34 6
expectations: The 14 predicted relationships all generalized, credibility intervals. This indicates that moderators are likely
and 4 relationships generalized that we did not hypothesize present in the data, suggesting the need for a more fine-tuned
(agreeableness with universalism; extraversion with power analysis of the data.
and hedonism; and conscientiousness with security). If we
define “strong” relationships as those with a rho greater than
Sinusoidal Pattern of Correlations
.50 (see Cohen, 1988), then 2 of the 50 relationships are
strong: agreeableness with benevolence (ρ = .61), and open- Schwartz (1992, 1996) suggested that correlations between
ness to experience with self-direction (ρ = .52). With only 2 values and any other variable should be represented graphi-
strong relationships out of 50 (after correcting for statistical cally with a sinusoidal curve. In this type of analysis, values
artifacts), traits and values are clearly distinct constructs. If are listed on the horizontal axis in order (i.e., moving around
we rely on the more conservative mean r, then none of the the circle) and the correlations with the other variable of
relationships would be classified as strong (the strongest interest are then plotted on the vertical axis. Figures 2 to 6
would be agreeableness with benevolence; mean r = .45). In provide plots of the meta-analytic estimates (rho) of the rela-
nearly every case, the percentage of variance accounted for tionships between traits and values. With the exception of
by statistical artifacts was small, resulting in generally wide emotional stability, which does not exhibit any significant

Table 7. Meta-Analytic Results for Extraversion and Values.
Power 54 54,599 .23 .31 .17 .09 .52 −.02 .63 5
Achievement 55 54,946 .23 .31 .16 .10 .52 −.01 .63 6
Hedonism 53 54,165 .16 .20 .09 .09 .32 .02 .39 15
Stimulation 51 53,692 .28 .36 .07 .27 .45 .22 .50 22
Self-Direction 55 54,959 .12 .17 .17 −.05 .38 −.16 .49 6
Universalism 54 54,364 −.05 −.05 .15 −.25 .14 −.35 .24 7
Benevolence 56 55,072 −.04 −.05 .25 −.37 .26 −.54 .43 3
Conformity 55 54,959 −.13 −.17 .20 −.42 .09 −.56 .23 4
Tradition 51 53,692 −.18 −.25 .20 −.51 .01 −.64 .15 4
Security 54 54,377 −.04 −.05 .17 −.27 .17 −.39 .29 6
Table 8. Meta-Analytic Results for Conscientiousness and Values.
Power 54 54,599 .04 .05 .09 −.07 .17 −.13 .23 17
Achievement 55 54,946 .12 .17 .11 .02 .31 −.05 .39 12
Hedonism 54 54,391 −.15 −.19 .15 −.38 .00 −.49 .11 6
Stimulation 51 53,692 −.12 −.16 .18 −.38 .07 −.50 .19 5
Self-direction 55 54,959 .01 .01 .24 −.29 .31 −.45 .47 3
Universalism 54 54,364 −.01 −.02 .18 −.26 .22 −.38 .35 4
Benevolence 56 55,072 .05 .07 .16 −.14 .28 −.25 .39 6
Conformity 55 54,959 .20 .27 .11 .13 .41 .05 .49 12
Tradition 51 53,692 .07 .10 .12 −.05 .25 −.13 .33 13
Security 54 54,377 .27 .37 .19 .14 .61 .01 .74 5
Note. Bolded mean r values have a 95% confidence interval that does not include 0. Bolded ρvalues have an 80% credibility interval that does not include
relationships with values, the pattern shows the expected regarding moderator analyses are available from the first
sinusoid patterns as predicted by the values theory. Especially author.
for openness to experience and agreeableness, one clear peak
and one clear valley is observed such that the relationships Culture. The WLS regression results for individualism/col-
increase and decrease monotonously as one moves around lectivism as a moderator are summarized in Table 11. We
the circle. Conscientiousness, as anticipated, has two peaks. proposed that more individualistic cultures would allow indi-
Even the weaker meta-analytic effects provide meaningful viduals greater flexibility in choosing values, such that they
information, as they follow the sinusoidal pattern predicted would be more likely to choose values consistent with their
by the structure of a circle, so even effects that do not gener- traits. Thus, we expected that trait–value relationships would
alize are interesting within the larger picture of how values be stronger in more individualistic cultures. However, only 4
relate to other variables. of the 50 regression equations were significantly moderated
by individualism/collectivism. The relationships for which
the moderator effect was significant were agreeableness with
Moderator Analyses hedonism and emotional stability with achievement, stimula-
Given the volume of data involved, we present our findings tion, and self-direction. None of these were hypothesized to
in an abbreviated format in Tables 11 to 15. Additional data be related, and none were strong relationships in the main

Table 9. Meta-Analytic Results for Emotional Stability and Values.
Power 52 54,274 .02 .03 .08 −.07 .13 −.13 .18 21
Achievement 54 54,747 −.01 −.01 .10 −.14 .12 −.21 .19 14
Hedonism 53 54,165 .01 .01 .05 −.05 .08 −.09 .12 35
Stimulation 51 53,692 .01 .02 .13 −.14 .18 −.23 .27 9
Self-direction 55 54,959 −.01 −.01 .09 −.13 .11 −.19 .18 18
Universalism 53 54,165 −.03 −.03 .07 −.12 .05 −.16 .10 25
Benevolence 54 54,747 −.01 −.01 .10 −.14 .11 −.20 .18 14
Conformity 55 54,959 −.04 −.05 .05 −.12 .01 −.15 .04 43
Tradition 51 53,692 −.02 −.03 .04 −.09 .02 −.12 .05 53
Security 54 54,377 −.02 −.03 .08 −.13 .07 −.18 .12 24
Table 10. Summary of the Main Meta-Analysis Results (ρ).
Openness to Experience Agreeableness Extraversion Conscientiousness Emotional stability

Power −.06 −.42 .31 .05 .03
Achievement .11 −.24 .31 .17 −.01
Hedonism .09 −.11 .20 −.19 .01
Stimulation .36 −.05 .36 −.16 .02
Self-direction .52 −.07 .17 .01 −.01
Universalism .33 .39 −.05 −.02 −.03
Benevolence .13 .61 −.05 .07 −.01
Conformity −.27 .26 −.17 .27 −.05
Tradition −.31 .22 −.25 .10 −.03
Security −.24 .00 −.05 .37 −.03
Note. Generalizable results (in bold) refer to results for which the 80% credibility interval does not include 0.
meta-analysis (the strongest was agreeableness with hedo- main meta-analysis (agreeableness with universalism; ρ =
nism, ρ = −.11). Graphing the line resulting from each of the .39). For this relationship, the slope was positive and the line
regression results shows that in each case, the slope was did not cross zero, suggesting that as the culture gets tighter,
positive, and that the line crosses the x-axis (such that the the relationship between the trait and the value gets stronger
relationship goes from negative to positive as the culture (counter to our expectations). Of the remaining six signifi-
goes from more collectivistic to more individualistic). The cant effects, all crossed zero. Three were negative, suggest-
results do not support our premise that the relationship ing that as the culture became tighter, the relationships
between personality and values is stronger in more individu- between traits and values went from positive to negative. The
alistic cultures. other three were positive, indicating that as the culture
We also expected that looser cultures would allow for became tighter, the relationships went from negative to posi-
stronger relationships between traits and values. For tight- tive. Thus, the results of the tightness/looseness moderator
ness/looseness of culture, the results (Table 12) indicate that do not suggest a consistent effect of culture on these
this aspect of culture significantly moderated 7 of the 50 relationships.
relationships: openness to experience with hedonism; agree- In sum, none of the hypothesized relationships (from the
ableness with universalism and security; extraversion with main analyses) showed evidence of moderation in our cul-
tradition; conscientiousness with power and universalism; tural moderator analyses. Those relationships that showed
and emotional stability with power. Of those 7 significant evidence of cultural moderation did not provide results that
effects, none were hypothesized relationships, although one were either systematic or supportive of our hypothesis. The
relationship was moderate in strength and generalized in the number of studies (which ranged from 41-47) was quite

Figure 2. Plot of relationships between openness to experience and values.

Note: Pow = Power, Ach = Achievement, Hedn = Hedonism, Stim = Stimulation, Sdir = Self-direction, Univ = Universalism, Benv = Benevolence, Conf =
Conformity, Trad = Tradition, Scty = Security.
Figure 3. Plot of relationships between agreeableness and values.

modest to detect moderation using regression. With more Values instruments. Results of the analyses using the value
studies, and studies from a wider range of countries, there instrument (SVS or PVQ) are provided in Table 13. Credi-
may be greater potential for finding meaningful moderation bility intervals were still generally wide, and the percent
effects by culture. Nevertheless, the current evidence does variance accounted for relatively small, after taking into
not support the idea that culture meaningfully moderates the account the values scale that was used. We conducted a
relationships between traits and values. series of z tests to determine whether the 95% CI overlapped

Figure 4. Plot of relationships between extraversion and values.

Figure 5. Plot of relationships between conscientiousness and values.

when comparing the two types of studies; results indicate moderator. As additional evidence, more of the relationships
that 29 of the 50 relationships are significantly different generalize when we separate the studies based on values
between studies using the SVS versus studies using the measure used (i.e., the credibility intervals are more narrow
PVQ. Thus, the values measure appears to act as a and less likely to include 0; so there is more similarity within

Figure 6. Plot of relationships between emotional stability and values.

Table 11. Personality and Values; Individualism/Collectivism of Culture as a Moderator.
Openness to experience Agreeableness Extraversion Conscientiousness Emotional stability
k β R2 k β R2 k β R2 k β R2 k β R2
Power 44 −.01 .00 45 .14 .02 45 −.09 .01 45 −.07 .00 44 .20 .04
Achievement 46 .16 .03 47 .25 .06† 47 −.03 .00 47 .16 .02 46 .38 .14*
Hedonism 45 −.07 .01 45 .30 .09* 45 −.04 .00 46 .14 .02 45 .21 .04
Stimulation 43 −.13 .02 43 −.07 .01 43 −.03 .00 45 .06 .00 43 .34 .12*
Self-direction 47 .14 .02 47 −.14 .02 47 −.05 .00 47 −.06 .00 47 .31 .10*
Universalism 45 .08 .01 46 −.04 .00 46 .08 .01 46 −.08 .01 45 −.03 .00
Benevolence 46 −.01 .00 47 −.02 .00 47 .20 .04 47 .11 .01 46 −.16 .02
Conformity 47 .00 .00 47 −.05 .00 47 .16 .03 47 .00 .00 47 −.09 .01
Tradition 43 .08 .01 43 .19 .03 43 .10 .01 43 .14 .02 43 −.20 .04
Security 46 .05 .00 46 .17 .03 46 .16 .03 46 .02 .00 46 −.01 .00
Note. k = the number of studies. β = the standardized beta weight of the independent variable (individualism/collectivism).
†
Significant at p < .10.
*
each of the two groups of studies than there is across all example, the value item “wisdom” from the SVS (part of
studies). While 18 relationships generalized in the overall universalism) is not covered in the PVQ. This is probably
analysis, 24 relationships generalized when using just the the value item that relates most strongly to valuing intellect,
SVS, and 24 relationships generalize when using just the and therefore, its absence from the PVQ may have weak-
PVQ. This pattern suggests that the PVQ and the SVS are ened the relationships of the PVQ’s universalism with the
not measuring entirely the same content, and researchers trait openness; indeed, the PVQ’s universalism was more
should bear this in mind in future research. Although the weakly linked with trait openness compared with the SVS’s
intention in developing the PVQ was to develop a measure universalism. In general, however, as we expected, the PVQ
with identical content compared with the SVS, slight differ- had stronger relationships with traits than the SVS. Specifi-
ences in the coverage of content may be present. For cally, of the 15 hypothesized relationships, 10 were stronger

Table 12. Personality and Values; Tightness/Looseness of Culture as a Moderator.

2 2 2 2
k β R k β R k β R k β R k β R2
Power 42 .14 .02 43 −.27 .07† 43 .06 .00 43 −.32 .10* 42 .39 .15*
Achievement 44 −.01 .00 45 −.27 .07† 45 .22 .05 45 −.20 .04 44 −.01 .00
Hedonism 43 .33 .11* 43 −.29 .08† 43 .10 .01 44 −.24 .06 43 .11 .01
Stimulation 41 .30 .09† 41 −.05 .00 41 .19 .04 41 .17 .03 41 .08 .01
Self-direction 45 .03 .00 45 .25 .06 45 .27 .07† 45 .22 .05 45 −.16 .03
Universalism 43 −.01 .00 44 .40 .16* 44 −.11 .01 44 .44 .19* 43 −.06 .00
Benevolence 44 .26 .07† 45 .21 .05 45 −.08 .01 45 .02 .00 44 −.17 .03
Conformity 45 −.01 .00 45 −.16 .03 45 −.19 .03 45 −.14 .02 45 −.08 .01
Tradition 41 −.19 .04 41 −.31 .09† 41 −.33 .11* 41 −.14 .02 41 −.21 .04
Security 44 −.11 .01 44 −.36 .13* 44 −.17 .03 44 −.12 .02 44 .19 .04
Note. k = the number of studies. β = the standardized beta weight of the independent variable (tightness/looseness of culture).
†
*
Table 13. Meta-Analytic Rho (and SD-rho) for Personality and Values; Values Measure as Moderator.
SVS PVQ SVS PVQ SVS PVQ SVS PVQ SVS PVQ
ρ SDρ ρ SDρ z ρ SDρ ρ SDρ z ρ SDρ ρ SDρ z ρ SDρ ρ SDρ z ρ SDρ ρ SDρ z
Power −.11 .13 −.03 .06 2.82 −.26 .15 −.53 .17 5.56 .14 .10 .43 .07 11.8 .06 .09 .05 .09 .37 .05 .06 .01 .08 1.82
Achievement .13 .15 .11 .06 .63 −.07 .16 −.36 .21 5.02 .20 .12 .39 .15 4.55 .22 .09 .12 .11 3.24 −.05 .04 .02 .12 2.39
Hedonism .08 .05 .09 .14 .29 −.10 .13 −.12 .08 .64 .14 .06 .25 .08 4.97 −.14 .10 −.22 .17 1.81 .03 .05 .01 .06 1.17
Stimulation .33 .10 .38 .13 1.38 −.03 .12 −.06 .10 .91 .30 .06 .40 .05 6.06 −.08 .05 −.21 .21 2.58 −.10 .05 .10 .10 7.86
Self-direction .51 .12 .54 .09 .98 −.02 .19 −.10 .19 1.40 .12 .14 .19 .18 1.41 .04 .14 −.01 .29 .68 −.08 .05 .05 .07 6.87
Universalism .37 .09 .30 .10 2.39 .29 .11 .45 .07 6.03 .06 .12 −.13 .12 5.24 .01 .14 −.03 .21 .71 .01 .04 −.05 .07 3.30
Benevolence .10 .11 .16 .10 1.91 .46 .16 .71 .06 7.70 .12 .12 −.20 .24 5.28 .14 .09 .01 .18 2.86 −.03 .05 0 .12 1.01
Conformity −.20 .14 −.32 .15 2.73 .28 .09 .26 .10 .69 .01 .10 −.29 .16 7.14 .30 .09 .25 .12 1.52 −.05 .03 −.06 .05 .77
Tradition −.22 .11 −.38 .13 4.30 .27 .11 .20 .11 2.09 −.07 .10 −.40 .12 9.65 .12 .10 .08 .13 1.11 −.03 .04 −.04 .05 .71
Security −.17 .17 −.28 .15 2.30 .14 .14 −.09 .12 5.94 .08 .13 −.13 .15 4.88 .28 .09 .44 .20 3.19 .01 .04 −.06 .07 3.86
Avg % Variance 18% 9% 14% 8% 20% 8% 22% 5% 55% 17%
Note. Generalizable results (bolded ρ) refer to results for which the 80% credibility interval does not include 0. The number of studies for the SVS ranged from 27 to 30
(N ranged from 19,276 to 20,183); for the PVQ there were 18 studies (N = 32,707). The z test is used to test for whether the 95% confidence intervals overlap (Hunter &
Schmidt, 1990); z > 1.96 indicates that the scores are significantly different at p ≤ .05; z > 2.56 indicates that the scores are significantly different at p ≤ .01. SVS = Schwartz
Value Survey (Schwartz, 1992); PVQ = Portrait Values Questionnaire (Schwartz et al., 2001). SDρ = the standard deviation of the corrected correlations; Avg % Variance = the
average percent variance accounted for.
with the PVQ compared with 5 that were stronger with the BFI, although this was not true for the NEO (although this
SVS, probably because the PVQ includes some trait-like could be a result of combining two different versions of the
elements in its items. NEO). Relationship with values also tended to be stronger
for the BFI as compared with the NEO, although we would
Personality instruments. Results of this moderator analysis are not have predicted this a priori, and it is not clear why this
presented in Table 14. As with the values scales, the modera- should be the case. Given that the NEO and the BFI were
tor analysis with personality scales yielded generally wide developed using different methods, perhaps it is not surpris-
credibility intervals and results that did not account for a ing that the results would differ.
large percentage of variance. We again conducted a series of
z tests to determine whether the 95% CIs overlapped; results Statistical adjustments to control for values scale use. Abbrevi-
indicate that 30 of the 50 relationships are significantly dif- ated results of this moderator analysis are provided in Table
ferent. Contrary to our expectations, therefore, the personal- 15. The z test showed that 37 of the 50 relationships were
ity measure significantly moderated the observed significantly different when using different methods, so sta-
relationships between personality and values. In addition, tistical adjustments act as a moderator. As further evidence,
more of the relationships generalized when using just the 28 of the 50 relationships generalized when limiting our

Table 14. Meta-Analytic Rho (and SD-rho) for Personality and Values; Personality Measure as Moderator.
BFI NEO BFI NEO BFI NEO BFI NEO BFI NEO
Power −.08 .05 −.14 .20 .83 −.50 .23 −.46 .05 .63 .33 .13 .09 .02 6.77 .09 .04 .04 .11 1.24 −.02 .06 .02 .02 2.28
Achievement .10 .08 −.05 .14 3.05 −.38 .25 −.14 .15 2.93 .27 .05 .10 .08 5.94 .18 .08 .16 .09 .56 −.09 .03 −.06 .02 2.94
Hedonism .01 .07 .04 .07 1.04 −.14 .05 −.19 .10 1.46 .19 .04 .10 .08 3.28 −.29 .09 −.18 .06 3.65 −.02 .04 .05 .03 4.90
Stimulation .31 .08 .22 .09 2.35 −.08 .07 −.15 .08 2.06 .36 .04 .27 .04 5.08 −.25 .16 −.15 .03 2.27 .07 .15 −.11 .02 4.42
Self-direction .52 .08 .39 .11 3.29 −.17 .13 −.13 .07 .98 .11 .08 −.04 .09 4.34 −.11 .19 −.08 .06 .56 .02 .11 −.03 .05 1.51
Universalism .33 .05 .39 .08 2.10 .39 .04 .21 .10 5.39 −.05 .15 −.07 .04 .48 −.11 .15 −.12 .11 .19 .01 .00 .03 .04 1.58
Benevolence .13 .03 .05 .11 2.24 .68 .10 .32 .14 6.96 −.15 .30 .01 .08 1.90 −.02 .15 .08 .07 2.18 .07 .07 −.04 .06 4.13
Conformity −.29 .10 −.32 .11 .70 .31 .04 .22 .12 2.39 −.21 .20 −.09 .06 2.13 .30 .04 .25 .05 2.71 −.04 .02 −.04 .00 .00
Tradition −.29 .08 −.28 .08 .28 .30 .05 .21 .11 2.19 −.28 .21 −.12 .10 2.41 .14 .05 .04 .07 3.56 −.02 .02 −.02 .00 .00
Security −.25 .10 −.33 .06 2.48 −.01 .18 −.03 .00 .42 −.06 .18 −.02 .06 .78 .50 .14 .17 .09 7.14 −.08 .04 .03 .03 7.86
Avg % Variance 20% 23% 15% 30% 13% 41% 15% 34% 46% 77%
Note. Generalizable results (bolded ρ) refer to results for which the 80% credibility interval does not include 0. Fourteen studies used the BFI (N =
25,776); for the NEO, the number of studies ranged from 8 to 11 (N ranged from 6,289 to 6,974). The z test is used to test for whether the 95%
confidence intervals overlap (Hunter & Schmidt, 1990); z > 1.96 indicates that the scores are significantly different at p ≤ .05; z > 2.56 indicates that the
scores are significantly different at p ≤ .01. BFI = Big Five Inventory (John, Donahue, & Kentle, 1991); NEO = Neuroticism, Extraversion, and Openness to
Experience, NEO-PI-R or NEO-FFI (Costa & McCrae, 1992). SDρ = the standard deviation of the corrected correlations; Avg % Variance = the average
percent variance accounted for.
Table 15. Meta-Analytic Rho (and SD-rho) for Personality and Values; Method as Moderator.
Openness to Experience Agreeableness Extraversion Conscientiousness Emotional stability
Partialled Corr.s Partialled Corr.s Partialled Corr.s Partialled Corr.s Partialled Corr.s
Power −.10 .12 −.05 .08 1.68 −.57 .18 −.25 .13 7.13 .30 .19 .26 .09 0.95 .07 .09 .03 .09 1.53 −.03 .08 .07 .03 5.78
Achievement .03 .11 .17 .07 5.36 −.37 .24 −.12 .14 4.57 .21 .13 .34 .10 3.96 .14 .07 .19 .16 1.34 −.07 .08 .01 .07 3.69
Hedonism −.01 .07 .14 .04 9.22 −.16 .08 −.07 .12 2.95 .17 .07 .20 .09 1.25 −.27 .11 −.15 .15 3.11 0 .04 .02 .05 1.49
Stimulation .28 .08 .39 .09 4.31 −.12 .08 .01 .06 6.21 .34 .06 .35 .05 0.61 −.27 .13 −.09 .14 4.45 .08 .14 −.08 .06 5.09
Self-direction .46 .10 .55 .09 3.31 −.21 .13 .05 .08 8.59 .05 .12 .25 .10 6.36 −.16 .14 .14 .15 7.17 .05 .08 −.08 .07 6.06
Universalism .33 .07 .33 .15 0 .35 .12 .37 .10 0.63 −.12 .08 .07 .14 5.55 −.16 .12 .11 .10 8.50 0 .04 −.03 .05 2.23
Benevolence .10 .09 .14 .11 1.35 .59 .21 .59 .09 0 −.21 .21 .16 .15 7.26 −.04 .13 .20 .08 8.02 .05 .08 −.07 .06 5.94
Conformity −.33 .11 −.18 .16 3.74 .27 .09 .27 .12 0 −.26 .14 .01 .18 5.76 .27 .07 .33 .09 2.56 −.04 .05 −.07 .05 2.09
Tradition −.32 .09 −.24 .14 2.24 .25 .09 .25 .13 0 −.32 .18 −.10 .16 4.34 .11 .08 .16 .08 2.09 −.02 .03 −.06 .05 3.20
Security −.30 .11 −.16 .15 3.36 −.09 .10 .15 .13 7.07 −.14 .11 .09 .17 5.45 .45 .20 .33 .14 2.43 −.06 .09 0 .06 2.75
Avg % Variance 17% 20% 12% 17% 12% 15% 14% 16% 30% 44%
Note. Generalizable results (bolded ρ) refer to results for which the 80% credibility interval does not include 0. Partialled refers to studies that partialled out the mean value
score; corr.s = studies that just provided the straight correlation. SD ρ = the standard deviation of the corrected correlations; Avg % Variance = the average percent variance
accounted for. There were 24 to 28 studies that clearly partialled out the mean value score (N ranged from 26,761 to 28,015) and 21 to 22 studies that clearly reported a
straight correlation (N ranged from 15,698 to 15,910). The z test is used to test for whether the 95% confidence intervals overlap (Hunter & Schmidt, 1990); z > 1.96 indicates
that the scores are significantly different at p ≤ .05; z > 2.56 indicates that the scores are significantly different at p ≤ .01.
analyses to studies using the partialling method, and 22 rela- Other moderators. We also examined study population, sam-
tionships generalized with studies using zero-order correla- ple size, and publication status as potential moderators.
tions. In comparing the methods, some relationships were However, the three moderators were confounded with one
stronger when partialling, some were stronger when using another, as several of the larger studies were based on unpub-
zero-order correlations, and some were about the same across lished data from Internet studies involving the general popu-
the two methods. However, “stronger” does not necessarily lation. None of these moderator analyses yielded clear
mean better or more accurate. As the partialling method has results; in all three sets of analyses, some differences emerged
strong theoretical support, partialling out the mean value between the subgroups, but across the 50 correlations, no
score should provide a clearer picture of the relationships subgroup was clearly superior to another. Given that the
because correlations that are controlled for scale use are results are not particularly meaningful, in the interest of
more accurate. space, we have not included details from these analyses

(additional details regarding these moderator analyses are Agreeableness. Agreeableness is also strongly related to val-
available on request from the first author). ues. Individuals who score high on agreeableness tend to
value being prosocial, particularly toward people in their
close environment (benevolence values) but also toward
Discussion people in society in general (universalism values).They also
The current article presents the first meta-analysis of the tend to value restraining their impulses to fit in (conformity
relationships between personality traits and personal values, and tradition), possibly to facilitate getting along with others.
focusing on the most studied models of traits and values— In contrast, individuals who score low on agreeableness tend
FFM and the Schwartz (1992) value theory. The results show to value having resources and being dominant (power val-
meaningful relationships, most of which were predicted a ues). Similar to openness, agreeableness had a perfect sinu-
priori and generally follow the sinusoidal pattern predicted soidal pattern of correlations with values.
by the values circle. As we expected, the strength of these
relationships was a product of two factors: the nature of the Extraversion. As we expected, extraversion correlated less
traits (more cognitively based traits have stronger relation- strongly with values than did openness and agreeableness,
ships with values) and content overlap between the traits and but these correlations are nevertheless theoretically mean-
values. The pattern of results also suggests that although val- ingful and generally follow a sinusoidal pattern. Individuals
ues and traits are related, the two constructs are distinct. who score high on extraversion tend to value excitement and
Moderator analyses suggest that the choice of instrument variety (stimulation values), as well as value enhancing their
(for both traits and values) sometimes affects the results. own interests through dominance, success, and having fun
Hypothesized links tend to be stronger when using the PVQ (power, achievement, and hedonism values).
to measure values as compared with the SVS, and when
using the BFI to measure personality as compared with the Conscientiousness. Conscientiousness is the only trait that
NEO. Furthermore, relationships are more consistent with was not expected to have a sinusoidal pattern of correlations
theory when response tendencies for values are statistically with values; instead, it was expected to have two peaks of
controlled. Additional moderator analyses do not support the correlations with values that are not adjacent to one another
idea that culture (individualism vs. collectivism and tight- in the value circle (i.e., conformity and achievement). Its cor-
ness vs. looseness cultural dimensions) affects these relation- relations with values generally followed the expected pat-
ships. We next elaborate on the specific meta-analytic results tern, but it was most strongly associated with security values.
and discuss their theoretical and practical implications. We Hence, this meta-analysis established that conscientious
then discuss some limitations and point to important future people tend to value order, adherence to rules, and the avoid-
directions for research. ance of risks. They also tend, to a lesser degree, to value fit-
ting in(conformity) and having socially recognized
accomplishments (achievement). In general, however, con-
Trait–Value Associations scientiousness is less strongly associated with values com-
In discussing the findings of associations between traits and pared with openness and agreeableness.
values, we refer both to the strengths of the links found and
to the patterns of associations. With regard to the latter, as Emotional stability. As expected, emotional stability did not
values are structured in a circle, relations with other vari- correlate with values. This is consistent with our premise that
ables should generally follow a sinusoidal pattern of gradual an affectively based trait should not be strongly related to
change in correlations as one moves around the circle. values, which are a cognitively based construct. Our meta-
analysis also confirms the idea expressed in a review on val-
Openness to experience. We expected this trait to have the ues and well-being (Sagiv et al., 2004) that values are not
strongest correlations with values based on the nature of directly related to well-being.
openness as the most cognitively based trait and on the over-
lap in content between openness and values. And indeed, of
Moderators
the five traits, openness to experience had some of the stron-
gest correlations with values, showing the clearest pattern of Culture. Our analyses were the first to examine cross-cultural
correlations and with an almost perfect sinusoidal pattern of differences in the relationships between traits and values.
correlations. Individuals scoring high on openness to experi- The moderator analyses showed little evidence for a cultural
ence tend to value novelty (self-direction and stimulation effect on the strength of trait–value relationships, although a
values) and particularly novel ideas (self-direction) and bro- larger sample size is needed before drawing firm conclu-
admindedness (universalism values). In contrast, individuals sions. Still, the occasional and unsystematic cultural effects
who score low on openness to experience tend to value main- found may suggest that although people in various cultures
taining the world as it is and the safety it provides (tradition, vary on their levels of traits (e.g., Costa, Terracciano, &
conformity, and security values). McCrae, 2001) and values (e.g., Schwartz, 2011b),

the relationships between traits and values remain largely Theoretical Implications
consistent across cultures, at least in the current range of cul-
tures (see John, Naumann, & Soto, 2008, for traits; Schwartz, The research literatures on personality traits and personal
2011b, for values) and for the cultural variables that we con- values share a common heritage: research on both sets of
sidered. Additional dimensions of culture do exist and could constructs originated with a reliance on the lexical hypoth-
also be considered, although we tested those that we believed esis for identifying relevant content. The lexical hypothe-
were most likely to moderate these relationships. As more sis proposes that meaningful differences in the
samples accumulate, the statistical power for answering this characteristics of individuals are encoded in language,
important question will increase. such that a review of the dictionary for terms describing
individuals will yield a comprehensive list of important
Survey instrument. Using the PVQ to measure values often characteristics (Goldberg, 1993; McCrae & Costa, 1997;
results in stronger hypothesized trait–value correlations. The Schwartz, 1994). When Allport (1937) undertook his study
PVQ scale may inflate correlations between the two con- of personality traits based on the lexical hypothesis, he
structs because its items often include trait-like components. stated that personality traits should be non-evaluative and
Should researchers continue to use it? If an important part of was explicit in his efforts to remove evaluative terms that
a study is to measure values alone without any traces of related to an individual’s “character,” or values. Until quite
traits, strivings, or preferences (all exist in parts of some recently, that separation of values and traits has led the two
items), and if distinguishing between values and these other sets of constructs to be studied mostly independently of the
characteristics of the person is important, then researchers other.
might want to use the SVS rather than the PVQ. In all other Traits and values are independently examined as impor-
instances, the PVQ should be adequate for use, and it has tant predictors of a multitude of outcome variables in numer-
some clear advantages over the SVS (detailed in Schwartz, ous contexts in various areas such as educational psychology
2005). A newly constructed measure of values (Schwartz et (e.g., Knafo & Schwartz, 2004; Poropat, 2009), organiza-
al., 2012) builds on the PVQ and fewer of its items have trait- tional psychology (e.g., Berson, Oreg, & Dvir, 2007; Lim &
like components compared with the PVQ, but it is still not Ployhart, 2004), health psychology (e.g., Bergin, 1991;
completely trait-free. Researchers may also wish to bear in Terracciano & Costa, 2004), political psychology (e.g.,
mind that the content of the two instruments may sometimes Saucier, 2000; Schwartz, Caprara, & Vecchione, 2010), envi-
be slightly different. ronmental psychology (e.g., Grunert & Juhl, 1995;
Similarly, the choice of personality instrument is likely to Ramanaiah, Clump, & Sharpe, 2000), sports psychology
have an impact on the strength of the observed relationships (e.g., Courneya & Hellsten, 1998; M. Lee, Whitehead,
between traits and values. As with the values measure, Ntoumanis, & Hatzigeorgiadis, 2008), occupational psychol-
researchers should choose personality measures based on a ogy (e.g., Gottfredson, Jones, & Holland, 1993; Sagiv, 2002),
particular study’s purpose and design. We also encourage social psychology (e.g., Cohrs, Moschner, Maes, &
researchers to evaluate personality instruments at the item Kielmann, 2005; Sibley & Duckitt, 2008), and others. Only
level to determine whether the scale is accurately measuring recently have researchers started to explore the combined
traits or whether it also includes some value-laden items. We effects of traits and values on various outcomes (see, for
did not expect, a priori, for the BFI to exhibit stronger rela- example, Parks & Guay, 2012; Roccas et al., 2002). The
tionships with values as compared with the NEO—We can- present article offers the first meta-analysis of the relation-
not therefore say at this point which personality instrument ships between personality traits and personal values, thereby
yields more accurate results, just that the relationships clarifying these relationships. The results demonstrate that
differ. traits and values are related in predictable ways based both
on the extent to which the trait is cognitively based and on
Statistical adjustments to control for values scale use. Control- the extent to which the contents of traits and values are con-
ling for how respondents use the rating scales in statistical ceptually similar.
analyses of correlations should result in more accurate cor- As noted previously, researchers tend to adhere to one of
relations between traits and values. Such treatment of the three basic views of traits and values: (a) They are different
data is in line with the understanding that values exist in a ways of measuring the same thing; (b) they are unique and
system of values; therefore, the important element in linking separate constructs at the same level of abstraction; and (c)
to another variable is not so much the absolute importance they are both part of a hierarchy of personality, but they exist
given to the value but rather its importance compared with at different levels in that hierarchy. This meta-analysis
all other values—its prioritization over other values in one’s clearly demonstrates that the first viewpoint is inaccurate. If
value system. This consideration leads to the important rec- traits and values were different ways of measuring the same
ommendation for researchers to control for scale-use ten- thing, then the correlations between them should have been
dency in values when correlating values with other much stronger, demonstrating convergent validity. Yet only
variables. 18 of the 50 relationships generalized, and of those, only 2

were strong relationships. The pattern of results therefore Limitations

clearly demonstrates that although values are meaningfully
linked to traits, the two constructs are distinct. It is important to outline the unavoidable limitations of this
A review of the traits–values literature provides additional meta-analysis, which stemmed largely from lack of suitable
evidence in support of this premise; many studies have studies. First, the reliability corrections, as stated previously,
examined the differential impact of traits and values on vari- were based only on coefficient alpha, which does not take
ous outcomes and constructs, providing evidence of diver- into account transient error. As a result, they are likely to be
gent validity. For example, Roccas and colleagues (2002) overestimates of the true reliability. Second, information was
found that values and traits have different patterns of correla- not available to determine whether range restriction occurred
tions with religiosity (associated primarily with values but in the data, although we do not think that range restriction is
not traits) and with positive affect (associated primarily with a large issue in the studies examined here. Third, many of
traits but not values). Other researchers have replicated these these studies were conducted in different languages, using
findings with subjective well-being and life satisfaction translated scales. Given the number of studies available and
(associated primarily with traits; Haslam, Whelan, & Bastian, the fact that not all studies indicated the language, we made
2009; Saiz, Alvaro, & Martinez, 2011) and with religiosity no attempt to control for differences caused by the language
and spirituality (associated primarily with values; Saiz et al., of the instrument. Fourth, not enough studies were available
2011; Saroglou & Munoz-Garcia, 2008). Parks and Guay in each category to enable conducting hierarchical moderator
(2012) found that traits and values have differential relation- analyses, making it difficult to pinpoint which construct is
ships with the motivational processes of goal content and causing the moderation, because moderators may be corre-
striving. Finally, several studies have shown that values are lated (as noted above). Fifth, focusing exclusively on studies
better predictors of voting preference and choice than are that used the FFM for traits and the Schwartz (1992) value
traits (Caprara, Schwartz, Capanna, Vecchione, & circle for personal values led us to eliminate some studies
Barbaranelli, 2006; Caprara, Vecchione, & Schwartz, 2009; that might be worth including when a sufficient number of
Dirilen-Gumus, Cross, & Donmez, 2012). Clearly, these are such studies is available. Sixth, we examined only factor-
not just different ways of measuring the same thing, and level scores for traits, as most studies did not include facet-
including both traits and values in the same study has the level scores. Finally, as noted, the moderator analyses
potential to improve predictions of a wide array of regarding culture are likely to provide clearer results with
outcomes. more studies, especially studies of people from collectivist
As the different views on the nature of the relationships cultures. When more studies accumulate, a meta-analysis
between traits and values can cause confusion in the litera- can be conducted that overcomes these limitations.
ture, researchers who study traits and values should state
their underlying assumptions and provide a consistent theo-
Future Directions
retical conceptualization connecting their work with other
trait–value studies that conceptualize the link in a similar This meta-analysis points to the importance of three broad
way. As an example, the HEXACO personality inventory future directions that would enhance the integrative under-
(K. Lee & Ashton, 2004) is a six-factor model that includes standing of the person: (a) incorporating more elements of
Honesty–Humility (not included in our study because it is the person in one study to enhance the understanding of the
not a Big Five inventory). The Honesty–Humility factor structure of enduring psychological characteristics of the
seems to largely tap values; the developers of the scale state person, (b) examining causal directions of different charac-
that the common adjectives used to define the factor are teristics of the person to enhance the understanding of life-
“Sincere, honest, faithful/loyal, modest/unassuming, fair- span development of the person, and (c) integrating research
minded” (Ashton & Lee, 2007, p. 154). These descriptors from other domains, such as neuroscience, to further inform
overlap considerably with values items; “honest” and and develop our understanding of the constructs of traits and
“loyal” are both benevolence items on the SVS, whereas the values and how they are related.
tradition scale includes the item “humble” (Schwartz, 1992,
p. 7), which seems quite similar to modest and unassuming. Incorporating more elements of the person in the same
Fair-minded would seem to fit with universalism values, study. This meta-analysis established links between two
which are concerned with equality and fairness in society enduring characteristics of the person: traits and values.
(Schwartz, 1992). If one is of the opinion that traits and val- However, more elements should be included in the same
ues are both aspects of personality, then the HEXACO scale study, and the simultaneous relationships among multiple
may be a legitimate method for measuring personality. elements should be established empirically. In addition to
However, if a researcher views personality as an aggregate traits and values, these could include goals, needs, beliefs,
of traits (and not values), then this scale may not be appro- attitudes, and temperaments, among others. Some work has
priate for measuring personality. We therefore encourage been conducted in this direction, mainly examining two
researchers to make explicit their assumptions and types of constructs at a time (e.g., Calogero, Bardi, & Sutton,
definitions. 2009; Goodwin, Polek, & Bardi, 2012; Roberts & Robins,

2000), but much more work is needed as well as the inclu- preliminary exploration of the links between them. This
sion of more constructs. meta-analysis establishes, for the first time, the relationships
between personality traits (of the most widely used trait
Longitudinal directions of relations between enduring characteris- model—the FFM) and personal values (of the most widely
tics of the person. Although this meta-analysis provides an used value model—the Schwartz value theory). It also estab-
understanding of how traits and values relate to one another, it lishes for the first time a conceptual underpinning for under-
represents a “snapshot” of those relationships at one point in standing these relationships, as more cognitively based traits
time, which may not reflect how the two influence one another demonstrate stronger relationships with values, and more
across time. Theorists have suggested mechanisms for recip- emotionally based traits exhibit weaker (or no) relationships
rocal effects of traits and values (Caprara et al., 2006; Roccas with values. We now know that openness to experience and
et al., 2002), and several researchers have begun to theorize agreeableness are the most strongly and coherently related to
and examine the possible interrelations between the two (e.g., values; extraversion and conscientiousness also have some
Bilsky & Schwartz, 1994; Olver & Mooradian, 2003; Roccas meaningful relations to values; and emotional stability is
et al., 2002; Schermer et al., 2011). Traits are often considered generally unrelated to values. We also have established that
more innate; hence, one could argue that if one of these were traits and values are distinct constructs and that their rela-
to influence the other, it would be traits. Alternatively, one tionships show little variation cross-culturally. Our modera-
could argue that values influence traits—for example, when tor analyses also resulted in a recommendation for researchers
someone becomes a parent, the value domains of benevolence to control for scale-use tendency in values, and to consider
and security could become more important to them. This carefully which survey instrument to use as some measures
should change their behavior, and that behavioral change (i.e., the PVQ and the BFI) tend to yield stronger trait–value
might, over time, lead to modest changes in traits. We do not, relationships than do others. The findings from this meta-
at this point, fully understand the reciprocal nature of these analysis will enable researchers to use traits and values more
relationships over time, yet this is an important step for gain- effectively in their studies. Moreover, this new knowledge
ing an integrative view of the person and understanding how has prepared the ground for developing models that advance
this integration comes about. We therefore encourage research- our integrative understanding of the person.
ers to study this question in longitudinal research.
Acknowledgment
Integrating other research domains. The question of how traits We wish to thank Frank Schmidt and In-Sue Oh for their feedback
and values are related may be more easily answered if we on various meta-analytic issues we faced in developing this article.
search broadly in the research literature for other ways to Also, we are extremely grateful to the many researchers who shared
understand these constructs. Both traits and values are psy- their unpublished data sets with us for this study.
chological constructs—they represent complex processes
taking place within the brain. A better understanding of those Declaration of Conflicting Interests
neural processes may inform our understanding of how traits The author(s) declared no potential conflicts of interest with respect
and values relate to one another, how much they influence to the research, authorship, and/or publication of this article.
one another, and how much they are influenced by genetics
(nature) versus environment (nurture). For example, the Funding
frontal lobe develops later in life than the mid-brain, not The author(s) received no financial support for the research, author-
reaching full maturity until early adulthood (Sowell, Thomp- ship, and/or publication of this article.
son, Holmes, Jernigan, & Toga, 1999). This fact suggests
that psychological processes involving the frontal lobe might
Supplemental Material
be more influenced by external influences into early adult-
hood, whereas those that are processed primarily in the mid- The online supplemental material is available at http://pspr.
brain may be less prone to such influence. If this is true, then sagepub.com/supplemental.
stating that traits represent nature and values represent nur-
References
ture is overly simplistic, because traits themselves appear to
vary in the extent to which they involve the frontal lobe ver- *References marked with asterisk are included in the meta-analysis.
sus the mid-brain. Understanding at a more fundamental Allik, J. (2005). Personality dimensions across cultures. Journal of
level how traits and values are related to one another should Personality Disorders, 19, 212-232.
Allik, J. (2012). National differences in personality. Personality
help move the field forward in developing a more integrated
and Individual Differences, 53, 114-117.
conceptualization of characteristics of the individual. Allport, G. W. (1937). Personality: A psychological interpretation.
New York, NY: Henry Holt.
Conclusion *Arthaud-Day, M. L., Rode, J. C., & Turnley, W. H. (2012). Direct
and contextual effects of individual values on organizational
The past decade has seen initial attempts to clarify the simi- citizenship behavior in teams. Journal of Applied Psychology,
larities and differences between traits and values and a 97, 792-807.

Ashton, M. C., & Lee, K. (2007). Empirical, theoretical, and practi- *Caprara, G. V., Schwartz, S., Capanna, C., Vecchione, M., &
cal advantages of the HEXACO model of personality structure. Barbaranelli, C. (2006). Personality and politics: Values,
Personality and Social Psychology Review, 11, 150-166. traits, and political choice. Political Psychology, 27, 1-28.
*Bardi, A. (2005). [Values, traits, and ambiguity tolerance]. (Correlation matrix obtained on request from first author)
Unpublished raw data. *Caprara, G. V., & Vecchione, M. (2006). Le ragioni degli elet-
*Bardi, A. (2008). [Values and traits of first year psychology and tori: tratti, valori e stabilita del voto [The rationale of voters:
business students in autumn 2006]. Unpublished raw data. traits, values, and stability in voting]. Giornale Italiano do
*Bardi, A., Buchanan, K. E., Goodwin, R., Slabu, L., & Robinson, Psicologia, 3, 501-522.
M. P. (2014). Value stability and change during self-chosen life *Caprara, G. V., Vecchione, M., & Schwartz, S. H. (2009).
transitions: Self-selection vs. socialization effects. Journal of Mediational role of values in linking personality traits to politi-
Personality and Social Psychology, 106, 131-147. doi:10.1037/ cal orientation. Asian Journal of Social Psychology, 12, 82-94.
a0034818 Cervone, D. (2005). Personality architecture: Within-person structures
*Bardi, A., Bull, S., & Brown, H. L. (2008). [Values, goals, and and processes. Annual Review of Psychology, 56, 423-452.
traits of undergraduate students in spring 2008]. Unpublished Cloninger, C. R. (1994). Temperament and personality. Current
raw data. Opinion in Neurobiology, 4, 266-273.
*Bardi, A., & Guerra, V. M. (2011). Cultural values predict cop- Cohen, J. (1988). Statistical power analysis for the behavioral sci-
ing using culture as an individual difference variable in multi- ences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.
cultural samples. Journal of Cross-Cultural Psychology, 42, *Cohrs, J. C., Kielmann, S., Maes, J., & Moschner, B.
908-927. (2005). Zur Reliabilität und Validität der Erfassung von
*Bardi, A., & John, O. P. (2006). [Value meanings]. Unpublished Persönlichkeitsmerkmalen im Internet: Ergebnisse einer
raw data. Längsschnittstudie [Reliability and validity of personality
*Bardi, A., Lee, J. A., Hofmann-Towfigh, N., & Soutar, G. (2009). assessment over the Internet: Results of a longitudinal study].
The structure of intraindividual value change. Journal of In K.-H. Renner, A. Schütz, & F. Machilek (Eds.), Internet
Personality and Social Psychology, 97, 913-929. und Persönlichkeit [Internet and personality] (pp. 38-53).
*Bardi, A., Levontin, L., & John, O. P. (2011). [Traits and values]. Göttingen, Germany: Hogrefe.
Unpublished data. Cohrs, J. C., Moschner, B., Maes, J., & Kielmann, S. (2005). The
*Bardi, A., Loader, N. S., Keen, A., & Martin, Z. C. (2004). motivational bases of right-wing authoritarianism and social
[Values, traits, coping, and well-being of international stu- dominance orientation: Relations to values and attitudes in
dents]. Unpublished data. the aftermath of September 11, 2001. Personality and Social
Barenbaum, N. B., & Winter, D. G. (2008). History of modern Psychology Bulletin, 31, 1425-1434.
personality theory and research. In O. P. John, R. W. Robins, *Collins, S., & Blum, T. (2011). Mechanisms matter: An exami-
& L. A. Pervin (Eds.), Handbook of personality: Theory and nation of motivational linkages between individual differences
research (3rd ed., pp. 3-26). New York, NY: Guilford Press. and leader emergence. Unpublished manuscript.
*Barrick, M. R., Giluk, T., Shaffer, J., & Stewart, G. S. (2006). Costa, P. T., Jr., & McCrae, R. R. (1988). From catalog to classi-
[Personality and values data collected 2005-2006]. Unpublished fication: Murray’s needs and the five-factor model. Journal of
data. Personality and Social Psychology, 55, 258-265.
*Barrick, M. R., Stewart, G., Zimmerman, R., Parks, L., Darnold, Costa, P. T., Jr., & McCrae, R. R. (1992). NEO-PI-R: Professional
T., & Dustin, S. (2005). [Correlates of individual values, data manual. Odessa, FL: Psychological Assessment Resources.
collected 2003-2004]. Unpublished data. Costa, P. T., Jr., Terracciano, A., & McCrae, R. R. (2001). Gender
Bergin, A. E. (1991). Values and religious issues in psychotherapy differences in personality traits across cultures: Robust and sur-
and mental health. American Psychologist, 46, 394-403. prising findings. Journal of Personality and Social Psychology,
Berson, Y., Oreg, S., & Dvir, T. (2007). CEO values, organiza- 81, 322-331.
tional culture and firm outcomes. Journal of Organizational Courneya, K. S., & Hellsten, L. M. (1998). Personality correlates of
Behavior, 29, 615-633. exercise behavior, motives, barriers, and preferences: An appli-
Bilsky, W., & Schwartz, S. H. (1994). Values and personality. cation of the five-factor model. Personality and Individual
European Journal of Personality, 8, 163-181. Differences, 24, 625-633.
*Blickle, G., Schlege, A., Fassbender, P., & Klein, U. (2006). Some De Fruyt, F., Van De Wiele, L., & Van Heeringen, C. (2000).
personality correlates of business white-collar crime. Applied Cloninger’s psychobiological model of temperament and char-
Psychology: An International Review, 55, 220-233. acter and the five-factor model of personality. Personality and
*Burns, M., & Postlethwaite, B. (2007). [Personality and values Individual Differences, 29, 441-452.
data collected 2004-2005]. Unpublished data. *Dirilen-Gumus, Ö. (2010). The effect of religiosity on political
Buss, A. H. (1989). Personality as traits. American Psychologist, ideology via value types and personality traits: A comparison
44, 1378-1388. between Turkey and USA. Procedia - Social and Behavioral
Calogero, R. M., Bardi, A., & Sutton, R. (2009). A need basis for Sciences, 5, 12-17.
values: Associations between the need for cognitive closure *Dirilen-Gumus, Ö., Cross, S. E., & Donmez, A. (2012). Who
and value priorities. Personality and Individual Differences, voted for whom? Comparing supporters of Obama and McCain
46, 154-159. on value types and personality traits. Journal of Applied Social
Caprara, G. V., Alesandri, G., & Eisenberg, N. (2012). Prosociality: Psychology, 42, 2879-2900.
The contribution of traits, values, and self-efficacy beliefs. *Dollinger, S. J., Leong, F. T. L., & Ulicni, S. K. (1996). On traits
Journal of Personality and Social Psychology, 102, 1289- and values: With special reference to openness to experience.
1303. doi:10.1037/a0025626 Journal of Research in Personality, 30, 23-41.

Field, A. P., & Gillett, R. (2010). How to do a meta-analysis. British John, O. P., Naumann, L. P., & Soto, C. J. (2008). Paradigm shift to
Journal of Mathematical and Statistical Psychology, 63, 665-694. the integrative Big Five trait taxonomy: History, measurement,
Fiske, D. W. (1994). Two cheers for the Big Five! Psychological and conceptual issues. In O. P. John, R. W. Robins, & L. A.
Inquiry, 5, 123-124. Pervin (Eds.), Handbook of personality: Theory and research
Gelfand, M. J., Nishii, L. H., & Raver, J. L. (2006). On the nature (3rd ed., pp. 114-158). New York, NY: Guilford Press.
and importance of cultural tightness–looseness. Journal of John, O. P., & Srivastava, S. (1999). The Big Five trait taxonomy:
Applied Psychology, 91, 1225-1244. History, measurement, and theoretical perspectives. In L. A.
Gelfand, M. J., Raver, J. L., Nishii, L., Leslie, L. M., Lun, J., Lim, Pervin & O. P. John (Eds.), Handbook of personality: Theory
B. C., . . . Yamaguchi, S. (2011). Differences between tight and and research (2nd ed., pp. 102-138). New York, NY: Guilford
loose cultures: A 33-nation study. Science, 332, 100-104. Press.
Goldberg, L. R. (1993). The structure of phenotypic personality Kandler, C. (2012). Nature and nurture in personality development:
traits. American Psychologist, 48, 26-34. The case of neuroticism and extraversion. Current Directions
*Goldberg, L. R. (2008). The Eugene-Springfield Community in Psychological Science, 21, 290-296.
Sample: Information available from the research participants *Knafo, A. (2007). [Traits and Values]. Unpublished data.
(ORI Technical Report, Vol. 48, No. 1). Oregon Research Knafo, A., & Schwartz, S. H. (2004). Identity formation and par-
Institute. (Data obtained by request from author) ent-child value congruence in adolescence. British Journal of
Goodwin, R., Polek, E., & Bardi, A. (2012). The temporal reciproc- Developmental Psychology, 22, 439-458.
ity of values and beliefs: A longitudinal study within a major Knafo, A., & Spinath, F. M. (2011). Genetic and environmental
life transition. European Journal of Personality, 26, 360-370. influences on girls’ and boys’ gender-typed and gender-neutral
doi:10.1002/per.844 values. Developmental Psychology, 47, 726-731. doi:10.1037/
Gottfredson, G. D., Jones, E. M., & Holland, J. L. (1993). Personality a0021910
and vocational interests: The relation of Holland’s six interest *Kosinski, M., Stillwell, D. J., & Graepel, T. (2013). Private traits
dimensions to five robust dimensions of personality. Journal of and attributes are predictable from digital records of human
Counselling Psychology, 40, 518-524. behavior. Proceedings of the National Academy of Sciences.
Graham, J., Haidt, J., & Nosek, B. A. (2009). Liberals and conser- *Kusdil, M. E. (2000). Value socialisation in cultural context: A
vatives rely on different sets of moral foundations. Journal of study with British and Turkish families (Unpublished doctoral
Personality and Social Psychology, 96, 1029-1046. dissertation). The University of Sussex, Brighton, UK.
*Graham, J., Nosek, B. A., Haidt, J., Iyer, R., Koleva, S., & Ditto, P. Lee, K., & Ashton, M. C. (2004). Psychometric properties of the
H. (2011). Mapping the moral domain. Journal of Personality HEXACO Personality Inventory. Multivariate Behavioral
and Social Psychology, 101, 366-385. Research, 38, 329-358.
Graziano, W. G., & Eisenberg, N. (1997). Agreeableness: A dimen- Lee, M., Whitehead, J., Ntoumanis, N., & Hatzigeorgiadis, A.
sion of personality. In R. Hogan, J. A. Johnson, & S. R. Briggs (2008). Relationships among values, achievement orientations,
(Eds.), Handbook of personality psychology (pp. 795-824). San and attitudes in youth sport. Journal of Sport and Exercise
Diego, CA: Academic Press. Psychology, 30, 588-610.
Grunert, S. C., & Juhl, H. J. (1995). Values, environmental attitudes, Lim, B., & Ployhart, R. E. (2004). Transformational leadership:
and buying organic foods. Journal of Economic Psychology, Relations to the Five-Factor Model and team performance in
16, 39-62. typical and maximum contexts. Journal of Applied Psychology,
Gunthert, K. C., Lawrence, L. H., & Armeli, S. (1999). The role of 89, 610-621.
neuroticism in daily stress and coping. Journal of Personality Locke, E. A. (1997). The motivation to work: What we know. In
and Social Psychology, 77, 1087-1100. M. L. Maehr & P. R. Pintrich (Eds.), Advances in motivation
*Haslam, N., Whelan, J., & Bastian, B. (2009). Big Five traits medi- and achievement (Vol. 10, pp. 375-412). Greenwich, CT: JAI
ate associations between values and subjective well-being. Press.
Personality and Individual Differences, 46, 40-42. *Lonnqvist, J.-E., & Versakalo, M. (2005). [Traits and Values].
Heine, S. J., & Buchtel, E. E. (2009). Personality: The universal and Unpublished data.
the culturally specific. Annual Review of Psychology, 60, 369-394. *Lonnqvist, J.-E., & Walkowitz, G. (2010). [Personality and values
Hitlin, S., & Piliavin, J. A. (2004). Values: Reviving a dormant con- of students at the University of Bonn]. Unpublished data.
cept. Annual Review of Sociology, 30, 359-393. *Luengo Kanacri, B. P., Rosa, V., & Di Giunta, L. (2012). The
Hofstede, G. (1980). Culture’s consequences: International differ- mediational role of values in linking personality traits to
ences in work-related values. Beverly Hills, CA: SAGE. civic engagement in Italian youth. Journal of Prevention &
Hofstee, W. K. B. (1994). Who should own the definition of per- Intervention in the Community, 40, 8-21.
sonality? European Journal of Personality, 8, 149-162. *Luk, C. L., & Bond, M. H. (1993). Personality variation and val-
Hogan, J., & Ones, D. S. (1997). Conscientiousness and integrity ues endorsement in Chinese university students. Personality
at work. In R. Hogan, J. A. Johnson, & S. R. Briggs (Eds.), and Individual Differences, 14, 429-437.
Handbook of personality psychology (pp. 849-872). San Diego, McAdams, D. P. (1995). What do we know when we know a per-
CA: Academic Press. son? Journal of Personality, 63, 365-396.
Hunter, J. E., & Schmidt, F. L. (2004). Methods of meta-analysis: McAdams, D. P. (1996). Alternative futures for the study of
Correcting error and bias in research findings. Thousand human individuality. Journal of Research in Personality,
Oaks, CA: SAGE. 30, 374-378.
John, O. P., Donahue, E. M., & Kentle, R. L. (1991). The Big McAdams, D. P., & Pals, J. L. (2006). A new Big Five: Fundamental
Five Inventory: Technical report. Berkeley: University of principles for an integrative science of personality. American
California, Berkeley. Psychologist, 61, 204-217.

McClelland, D. C. (1996). Does the field of personality have a *Roccas, S., Sagiv, L., & Porat, R. (2007). [Personality traits and
future? Journal of Research in Personality, 30, 429-434. personal values]. Unpublished raw data.
McCrae, R. R., & Costa, P. T., Jr. (1997). Personality trait structure *Roccas, S., Sagiv, L., Schwartz, S. H., & Knafo, A. (2002). The
as a human universal. American Psychologist, 52, 509-516. Big Five personality factors and personal values. Personality
McCrae, R. R., & Costa, P. T., Jr. (2003). Personality in adulthood: and Social Psychology Bulletin, 28, 789-801.
A Five-Factor Theory perspective. New York, NY: Guilford Rohan, M. J. (2000). A rose by any name? The values construct.
Press. Personality and Social Psychology Review, 4, 255-277.
McCrae, R. R., & Costa, P. T., Jr. (2008). The Five-Factor Theory Rokeach, M. (1973). The nature of human values. New York, NY:
of personality. In O. P. John, R. W. Robins, & L. A. Pervin The Free Press.
(Eds.), Handbook of personality: Theory and research (3rd ed., Sagiv, L. (2002). Vocational interests and basic values. Journal of
pp. 159-180). New York, NY: Guilford Press. Career Assessment, 10, 233-257.
McCrae, R. R., & John, O. P. (1992). An introduction to the five- Sagiv, L., Roccas, S., & Hazan, O. (2004). Value pathways to
factor model and its applications. Journal of Personality, 60, well-being: Healthy values, valued goal attainment, and envi-
175-215. ronmental congruency. In P. A. Lingly & S. Josheph (Eds.),
*Olver, J. M., & Mooradian, T. A. (2003). Personality traits and Positive psychology in practice (pp. 68-85). Hoboken, NJ:
personal values: A conceptual and empirical integration. John Wiley.
Personality and Individual Differences, 35, 109-125. Saiz, J., Alvaro, J. L., & Martinez, I. (2011). Relacion entre rasgos
*Parks, L. (2007). Personality and values as predictors of moti- de personalidad y valores personales en pacientes dependien-
vated behavior (Unpublished doctoral dissertation). University tes de la cocaine [Relationships between personality traits and
of Iowa, Iowa City. personal values in cocaine-dependent patients]. Adicciones, 23,
*Parks, L. (2008). [Personality and values of bank employees]. 125-132.
Unpublished data. *Saroglou, V., & Munoz-Garcia, A. (2008). Individual differences
Parks, L., & Guay, R. P. (2009). Personality, values, and motiva- in religion and spirituality: An issue of personality traits and/or
tion. Personality and Individual Differences, 47, 675-684. values. Journal for the Scientific Study of Religion, 47, 83-101.
Parks, L., & Guay, R. P. (2012). Can personal values predict perfor- Saucier, G. (2000). Isms and the structure of social attitudes.
mance? Evidence in an academic setting. Applied Psychology: Journal of Personality and Social Psychology, 78, 366-385.
An International Review, 61, 149-173. Schermer, J. A., Vernon, P. A., Maio, G. R., & Jang, K. L. (2011).
*Parks-Leduc, L., Pattie, M. W., Pargas, F., & Eliason, R. (2014). A behavior genetic study of the connection between social val-
Self-monitoring as an aggregate construct: Relationships with ues and personality. Twin Research and Human Genetics, 14,
personality and values. Personality and Individual Differences, 233-239.
58, 3-8. Schwartz, S. H. (1992). Universals in the content and structure of
Pelto, P. J. (1968). The differences between “tight” and “loose” values: Theoretical advances and empirical tests in 20 coun-
societies. Society, 5, 37-40. tries. In M. P. Zanna (Ed.), Advances in experimental social
Peterson, R. A. (2001). On the use of college students in social psychology (pp. 1-65). San Diego, CA: Academic Press.
science research: Insights from a second-order meta-analysis. Schwartz, S. H. (1994). Are there universal aspects in the structure
Journal of Consumer Research, 28, 450-461. and contents of human values? Journal of Social Issues, 50(4),
*Poling, T. L., Woehr, D. J., Gorman, C. A., & Arciniega, L. M. 19-45.
(2006, April). The impact of personality and value diversity Schwartz, S. H. (1996). Value priorities and behavior: Applying
on team performance. Paper presented at the Annual Meeting a theory of integrated value systems. In C. Seligman, J. M.
for the Society for Industrial and Organizational Psychology, Olson, & M. P. Zanna (Eds.), The Ontario Symposium: Vol. 8.
Dallas, TX. (Individual-level data obtained by request from The psychology of values (pp. 1-24). Hillsdale, NJ: Lawrence
first author) Erlbaum.
Poropat, A. E. (2009). A meta-analysis of the five-factor model of Schwartz, S. H. (2005). Robustness and fruitfulness of a theory of
personality and academic performance. Psychological Bulletin, universals in individual human values. In A. Tamayo & J. B.
135, 322-338. Porto (Eds.), Valores e comportamento nas organizaç Atoes
Pytlik Zillig, L. M., Hemenover, S. H., & Dienstbier, R. A. (2002). [Values and behavior in organizations] (pp. 56-95). Petrópolis,
What do we assess when we assess a Big 5 trait? A content Brazil: Vozes.
analysis of the affective, behavioral, and cognitive processes Schwartz, S. H. (2011a). Studying values: Personal adventure,
represented in Big 5 personality inventories. Personality and future directions. Journal of Cross-Cultural Psychology, 42,
Social Psychology Bulletin, 28, 847-858. 307-319.
Ramanaiah, N. V., Clump, M., & Sharpe, J. P. (2000). Personality Schwartz, S. H. (2011b). Values: Individual and cultural. In F. J.
profiles of environmentally responsible groups. Psychological R. van de Vijver, A. Chasiotis, & S. M. Breugelmans (Eds.),
Reports, 87, 176-178. Fundamental questions in cross-cultural psychology (pp. 463-
Roberts, B. W., & Robins, R. W. (2000). Broad dispositions, broad 493). Cambridge, UK: Cambridge University Press.
aspirations: The intersection of personality traits and major life Schwartz, S. H., & Bardi, A. (2001). Value hierarchies across
goals. Personality and Social Psychology Bulletin, 26, 1284- cultures: Taking a similarities perspective. Journal of Cross-
1296. Cultural Psychology, 32, 268-290.
Roccas, S., & Sagiv, L. (2010). Personal values and behavior: Schwartz, S. H., & Bilsky, W. (1987). Toward a psychological
Taking the cultural context into account. Social & Personality structure of human values. Journal of Personality and Social
Psychology Compass, 4, 30-41. Psychology, 53, 550-562.

Schwartz, S. H., Caprara, G. V., & Vecchione, M. (2010). Basic Taras, V., Kirkman, B. L., & Steel, P. (2010). Examining the
personal values, core political values, and voting: A longitudi- impact of culture’s consequences: A three-decade, multilevel,
nal analysis. Political Psychology, 31, 421-452. meta-analytic review of Hofstede’s cultural value dimensions.
Schwartz, S. H., Cieciuch, J., Vecchione, M., Davidov, E., Fischer, Journal of Applied Psychology, 95, 405-439.
R., Beierlein, C., . . . Konty, M. (2012). Refining the theory Terracciano, A., & Costa, P. T., Jr. (2004). Smoking and the five-
of basic individual values. Journal of Personality and Social factor model of personality. Addiction, 99, 472-481.
Psychology, 103, 663-688. doi:10.1037/a0029393 *Trapnell, P. (2007). [Cultural Values of University of Winnepeg
Schwartz, S. H., Melech, G., Lehmann, A., Burgess, S., Harris, M., students]. Unpublished data.
& Owens, V. (2001). Extending the cross-cultural validity of the *Uziel, L., Sagiv, L., & Roccas, S. (2007). [Personality traits and
theory of basic human values with a different method of mea- personal values]. Unpublished data.
surement. Journal of Cross-Cultural Psychology, 32, 519-542. *Vecchione, M., Alessandri, G., Barbaranelli, C., & Caprara, G.
Schwartz, S. H., Sagiv, L., & Boehnke, K. (2000). Worries and val- (2011). Higher-order factors of the Big Five and basic val-
ues. Journal of Personality, 68, 309-346. ues: Empirical and theoretical relations. British Journal of
Sheldon, K. M. (2004). Optimal human being: An integrated multi- Psychology, 102, 478-498.
level perspective. Mahwah, NJ: Lawrence Erlbaum. *Vecchione, M., Caprara, G., Schoen, H., Castro, J. L. G., &
Sheldon, K. M., & Elliot, A. J. (1999). Goal striving, need satisfac- Schwartz, S. H. (2012). The role of personal values and basic
tion, and longitudinal well-being: The self-concordance model. traits in perceptions of the consequences of immigration: A
Journal of Personality and Social Psychology, 76, 482-497. three-nation study. British Journal of Psychology, 103, 359-
Shoda, Y., & Mischel, W. (2006). Applying meta-theory to achieve 377.
generalizability and precision in personality science. Applied *Vecchione, M., & Mebane, M. (2007). [Traits and Values].
Psychology: An International Review, 55, 439-452. Unpublished data.
Sibley, C. G., & Duckitt, J. (2008). Personality and prejudice: A von Collani, G., & Grumm, M. (2009). On the dimensional struc-
meta-analysis and theoretical review. Personality and Social ture of personality, ideological beliefs, social attitudes, and
Psychology Review, 12, 248-279. personal values. Journal of Individual Differences, 30, 107-
Smith, P. B., & Bond, M. H. (1998). Social psychology across cul- 119.
tures (2nd ed.). London, England: Prentice Hall. *Wijnen, K., Vermeir, I., & Van Kenhove, P. (2007). The rela-
Sowell, E. R., Thompson, P. M., Holmes, C. J., Jernigan, T. L., & tionships between traits, personal values, topic involvement,
Toga, A. W. (1999). In vivo evidence for post-adolescent brain and topic sensitivity in a mail survey context. Personality and
maturation in frontal and striatal regions. Nature Neuroscience, Individual Differences, 42, 61-73. (Personality and values cor-
2, 859-861. relations obtained by request from first author)
*Stankov, L. (2007). The structure among measures of personality, *Wilson, R. E., Gosling, S. D., & Graham, L. T. (2012).
social attitudes, values, and social norms. Journal of Individual A review of Facebook research in the social sciences.
Differences, 28, 240-251. (Correlations obtained on request Perspectives on Psychological Science, 7, 203-220.
from author) doi:10.1177/1745691612442904
*Steca, P., Monzani, D., & Greco, A. (2011). Valori, Motivi e Winter, D. G., John, O. P., Stewart, A. J., Klohnen, E. C., & Duncan,
Tratti di Personalità. Differenze tra Persone e Tipi di Persone L. E. (1998). Traits and motives: Toward an integration of two
[Values, motives, and traits. Individual differences and Types]. traditions in personality research. Psychological Review, 105,
In G. V. Caprara, E. Scabini, P. Steca, & S. H. Schwartz (Eds.), 230-251.
I valori nell’Italia contemporanea [Values in contemporary *Wolfradt, U., & Dalbert, C. (2003). Personality, values, and belief
Italy] (pp. 69-93). Milano, Italy: Franco Angeli. in a just world. Personality and Individual Differences, 35,
Steel, P. D., & Kammeyer-Mueller, J. D. (2002). Comparing meta- 1911-1918.
analytic moderator estimation techniques under realistic condi- *Xu, X. (2005, August). OCB through cultural lenses: Exploring
tions. Journal of Applied Psychology, 87, 96-111. the relations among personality, OCB, and cultural values.
Suh, E., Diener, E., Oishi, S., & Triandis, H. C. (1998). The shifting Paper presented at the Academy of Management Annual
basis of life satisfaction judgments across cultures: Emotions Conference, Honolulu, HI.
versus norms. Journal of Personality and Social Psychology, *Yik, M. S. M., & Tang, C. S. (1996). Linking personality
74, 482-493. and values: The importance of a culturally relevant per-
*Sverdlik, N., & Sagiv, L. (2007). [Traits and Values]. Unpublished sonality scale. Personality and Individual Differences, 21,
data. 767-774.
View publication stats

Pers Soc Psychol Rev 2015 Parks Leduc 3 29

Uploaded by

Copyright:

Available Formats

Pers Soc Psychol Rev 2015 Parks Leduc 3 29

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pers Soc Psychol Rev 2015 Parks Leduc 3 29

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Personality traits and personal values: A meta-analysis

Article in Personality and Social Psychology Review · April 2014

Laura Parks-Leduc Gilad Feldman

SEE PROFILE SEE PROFILE

Action-inaction biases View project

The user has requested enhancement of the downloaded file.

Personality and Social Psychology Review

Personality Traits and Personal Values:

Laura Parks-Leduc1, Gilad Feldman2, and Anat Bardi3

Downloaded from psr.sagepub.com by guest on December 31, 2014

Table 1. Five-Factor Model of Personality.

Construct Description: The extent to which individuals tend to be . . .

Table 2. Schwartz Value Taxonomy.

Construct Description/Items: Individuals who value this believe in the importance of . . .

importance, so that different people consider some values as

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

Table 3. Studies Included in the Meta-Analysis; Possible Moderators.

Downloaded from psr.sagepub.com by guest on December 31, 2014

Table 3. (continued)

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

Table 5. Meta-Analytic Results for Openness to Experience and Values.

Table 6. Meta-Analytic Results for Agreeableness and Values.

Downloaded from psr.sagepub.com by guest on December 31, 2014

Table 7. Meta-Analytic Results for Extraversion and Values.

Table 8. Meta-Analytic Results for Conscientiousness and Values.

Downloaded from psr.sagepub.com by guest on December 31, 2014

Table 9. Meta-Analytic Results for Emotional Stability and Values.

Table 10. Summary of the Main Meta-Analysis Results (ρ).

Openness to Experience Agreeableness Extraversion Conscientiousness Emotional stability

Downloaded from psr.sagepub.com by guest on December 31, 2014

Figure 2. Plot of relationships between openness to experience and values.

Figure 3. Plot of relationships between agreeableness and values.

Downloaded from psr.sagepub.com by guest on December 31, 2014

Figure 4. Plot of relationships between extraversion and values.

Figure 5. Plot of relationships between conscientiousness and values.

Downloaded from psr.sagepub.com by guest on December 31, 2014

Figure 6. Plot of relationships between emotional stability and values.

Table 11. Personality and Values; Individualism/Collectivism of Culture as a Moderator.

Openness to experience Agreeableness Extraversion Conscientiousness Emotional stability

Downloaded from psr.sagepub.com by guest on December 31, 2014

Table 12. Personality and Values; Tightness/Looseness of Culture as a Moderator.

Openness to experience Agreeableness Extraversion Conscientiousness Emotional stability

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

were strong relationships. The pattern of results therefore Limitations

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014

Downloaded from psr.sagepub.com by guest on December 31, 2014