Content uploaded by Shalom H Schwartz
Author content
All content in this area was uploaded by Shalom H Schwartz on Aug 24, 2016
Content may be subject to copyright.
1
The Refined Theory of Basic Values
Shalom H. Schwartz
The Hebrew University of Jerusalem and National University Higher School of Economics,
Russian Federation
To appear in S. Roccas & L. Sagiv (Eds.), Values and behavior: Taking a cross-cultural
perspective, Springer.
This chapter was prepared within the framework of the Basic Research Program at the
National Research University Higher School of Economics (HSE) and supported within the
framework of a subsidy granted to the HSE by the Government of the Russian Federation for
the implementation of the Global Competitiveness Program
2
The theory of basic individual values (Schwartz, 1992) divided the value space into
ten distinct values. This was an arbitrary scientific convenience. Nonetheless, for over 20
years, researchers have used this division and the four higher order values meant to
summarize the ten values virtually without challenge. This was certainly not what I expected.
Instead, I anticipated that an improved theory would divide the space into discrete values
“with greater universal heuristic and predictive power” (Schwartz, 1992: 46). This chapter
presents such a theory, the refined theory of basic values first introduced in 2012 (Schwartz,
et al. 2012). I call it ‘refined’ rather than ‘revised’ because it does not contradict the original
theory. Rather, it identifies values with greater “heuristic and predictive power” in the same
value space.
The core of the value theory is the claim that all values are ordered relative to one
another on a circular motivational continuum. Terms that represents values (value items) are
located on this circular continuum according to the compatibility or conflict between the
motivations they express. In most if not all languages, there are thousands of value terms that
express nuances of motivation. The value theory pictures these value terms (=items) as filling
a two- or higher-dimensional space with no gaps. Multi-dimensional scaling (MDS) can
represent all the value items as points in a two-dimensional space. In this space, the more
highly correlated items are with one another and the more similar their correlations with other
items, the closer they are. The less positive the correlations among items, the more distant
they are in the space. Thus, the distance between any two items indicates how similar or
different they are from one another. MDS imposes no constraints on the observed relations
between items such as cluster or factor analyses do.
In order to work effectively with values as scientific constructs, we must divide this
crowded space of items into discrete values. The problem is how to partition the space and
how many values to distinguish. Because the values and the value items form a continuum
3
according to theory, any decisions about where to place boundaries are arbitrary. Moreover,
the boundaries of the resulting discrete values are necessarily fuzzy; value items on one side
of a boundary differ little from those on the other side. The original and refined value theories
apply to the same value space. They differ in how they partition the space and in the number
of values they distinguish.
The refined theory distinguishes 19 values. Given that parsimony should be preferred,
why go from 10 to 19? One can split almost any construct into finer subparts which yield
more precise understanding. The finer the distinctions between values, the clearer it is that the
values do form a continuum. But how far should we go? Two main criteria guided me in
developing both the original value theory and the refined theory. First, there must be evidence
that individuals across cultures actually discriminated these values. Second, there must be
evidence that each value has unique associations with other variables—background, attitudes,
and behavior—so there is some benefit in distinguishing them.
Developing the Refined Theory
In presenting the original value theory, I defined basic values as trans-situational
goals, varying in importance, which serve as guiding principles in the life of a person or
group. This definition also applies to the refined values. To qualify as a candidate to be a
basic refined value, I expected a value to exhibit the same key characteristics as the original
values. They should be grounded in one or more of three universal requirements of human
existence with which people must cope: needs of individuals as biological organisms,
requisites of coordinated social interaction, and survival and welfare needs of groups.
Moreover, they should fulfill the dynamic functions that Schwartz (1992, 2006) attributed to
basic values. They should (1) focus on attaining personal or social outcomes, (2) express
openness to change or conservation of the status quo or (3) serve self-interests or
4
transcendence of self-interests in the service of others, and (4) promote growth and self-
expansion or protect against anxiety and threat to self.
The multifaceted definitions of some of the original values provided one clue to
potential refined values. For example, Schwartz (1992) noted that the universalism value
included facets related to nature/beauty, to tolerance, and to concern for others’ well-being
and that the security value included facets related to personal security and to security in the
wider society. A second clue was spaces in the MDS projections of relations among items
intended to measure the original ten values. Such spaces could appear within a region
intended to measure an original value, for example, between the items measuring the wealth
facet of power and the dominance items. Spaces could also appear between adjacent basic
values on the motivational continuum. This often occurred between the power and security
regions.
On these bases, my colleagues and I (Schwartz, et al., 2012) generated distinct
conceptual definitions for 19 values. We checked our theorizing about the potential, narrowly
defined values against two types of available data. We first scrutinized MDS projections of
the correlations among the value items in each of 344 samples from 83 countries. The
samples had completed the Schwartz Value Survey (SVS: Schwartz, 1992, 2006) or the
Portrait Values Questionnaire (PVQ: Schwartz, 2006; Schwartz, et al., 2001). We sought
visual evidence for the potential subtypes we theorized. That is, we examined whether the 19
values emerged in distinct sub-regions in the MDS projections. Second, we examined results
of confirmatory factor analysis (CFA) studies of values. One available study analyzed SVS
data (Schwartz & Boehnke, 2004) and three analyzed PVQ data (Beierlein, Davidov,
Schmidt, Schwartz, & Rammstedt, 2012; Cieciuch & Schwartz, 2012; Saris, Knoppen &
Schwartz, 2013).
The 19 Values in the Refined Theory
5
Table 1 lists the 19 values that we expected people in each country to discriminate,
each defined in terms of its motivational goal. Below, I briefly present the conceptual grounds
for distinguishing each of these values and some of the empirical evidence that led us to
propose it. Schwartz et al. (2012) provide more complete details.
(1) Self-Direction Thought
(2) Self-Direction Action
The conceptual definition of the original self-direction value implies two potential
subtypes, autonomy of thought and of action. Both the SVS and the PVQ measured both
components. Items relevant to autonomy of thought referred to creativity, curiosity, and
interest. Items relevant to autonomy of action referred to choosing own goals, independence,
and self-reliance. The MDS projections in many samples revealed separable sub-regions for
thought and action items in the self-direction region. Autonomy of thought items were nearer
to universalism and autonomy of action items nearer to stimulation. The CFA analyses also
suggested that distinguishing thought and action subtypes was possible.
(3) Stimulation
The conceptual definition of stimulation mentions excitement, novelty, and challenge,
three somewhat different concepts. The SVS and PVQ included one item relevant to each.
However, all the items formed a single narrow spatial region in the MDS analyses and a
single distinct factor in the CFA analyses. This suggests that the concepts are conceptually
close and empirically inseparable. We therefore concluded that stimulation is sufficiently
narrowly defined.
(4) Hedonism
The conceptual definition of the original hedonism value refers to sensual pleasure.
All the analyses suggested that hedonism constitutes a single construct.
(5) Achievement
6
The conceptual definition of the original achievement value refers to “personal
success through demonstrating competence according to social standards.” This was intended to
specify what the achievement literature calls performance motivation—pursuit of normative
competence rather than mastery of knowledge and skills (Elliot & McGregor, 2001). Many
SVS and PVQ items were ambiguous regarding whether they expressed the goal of normative
success (demonstrating success) or of mastery (being competent) or both. Hence, the MDS
and CFA analyses could not provide clear evidence of whether these two types of
achievement were discriminable. To avoid this ambiguity and define achievement values as
originally intended, we narrowed the original definition. We defined achievement as
expressing only the underlying motivation that others judge one successful. We dropped the
concept of competence because self-direction thought and self-direction action values capture
the goals of mastering knowledge and skills, respectively.
(6) Power Dominance
(7) Power Resources
The conceptual definition of the original power value suggests three potential
subtypes, dominance over people, control through material resources, and maintaining
prestige or face. All three share the goal of promoting own interests by controlling what
happens and thereby minimizing or avoiding anxiety-arousing threats. Items relevant to
dominance referred to social power and being in charge. Items relevant to resources referred
to wealth and being rich. Items relevant to face, present only in the SVS, referred to
preserving one’s public image and being respected. The MDS projections and CFA analysis
of SVS items in many samples revealed separable sub-regions for all three subtypes. The
dominance and resources subtypes were close to one another and near to achievement values.
The face subtype was equidistant between the power and security regions. With only two
7
dominance items and one resource item in the PVQ, analyses could only provide hints that
these two subtypes were distinguishable.
(8) Face
We proposed face as a separate refined value for two reasons. First, the motivation to
maintain one’s public image has a strong defensive connotation absent in the more pro-active
power values. Second, in many MDS projections of PVQ data, there was an empty space
between the power and security regions; two face items usually occupied this space in the
projections of SVS data. This suggested a distinct type of motivation that falls between power
and security. We did not view face as a subtype of security because its self-interested and
manipulative orientation toward others differs from security.
(9) Security-Personal
(10) Security-Societal
The conceptual definition of the original security value implies two subtypes,
personal security and societal security. Both the SVS and the PVQ contained items that
measure both components. Items relevant to personal security referred to avoiding danger and
feeling cared about. Items relevant to societal security referred to social order and
government stability. The MDS projections in many samples revealed separable sub-regions
for the personal and societal items in the security region. Personal security items were nearer
to power and societal security items nearer to conformity and tradition. The CFA analyses
also supported the personal vs. societal distinction. In both the SVS and PVQ, there was one
item referring to health. It did not align consistently with either subtype. Staying healthy is an
important aspect of personal security. To emphasize the security connotation of the health
concept, we operationalized it as avoiding becoming sick.
(11) Conformity-Rules
(12) Conformity-Interpersonal
8
The conceptual definition of the original conformity value includes two components,
complying with social norms and avoiding upsetting others. However, the distinction between
the components is not clear because not upsetting others is itself a social norm. Consequently,
several SVS and PVQ items were ambiguous. We proposed a sharper definition of the first
component, limiting it to conforming to rules and formal obligations. We labeled the second
component interpersonal conformity. In the MDS analyses, the few clearly differentiated items
(e.g., ‘follow rules’ and ‘politeness’) provided some support for distinguishing the subtypes.
The rules items were closer to tradition and security and the interpersonal items were closer
to self-transcendence values. Schwartz (1992, 2006) described the motivation underlying
conformity values as protecting the self and avoiding negative reactions from others. The
location of the interpersonal conformity items near self-transcendence suggests that concern
for the other can also motivate interpersonal conformity.
(13) Tradition
The original definition of tradition specifies a conceptually narrow value, maintaining
cultural and religious traditions. Two SVS and two PVQ items clearly expressed this value
(e.g., respect tradition/preserve customs and religious belief/do what religion requires).
However, two other SVS and PVQ tradition items (e.g., humble/modest and satisfied with
what one has/don’t ask for more) pointed to a conceptually distinct value not mentioned in
the definition of tradition. Both the MDS and CFA analyses supported separation of two
different sets of items. Not surprisingly, given this split, tradition had the lowest internal
reliability among the ten values across samples (Schwartz, 2005a,b). For the proposed
tradition value, we retained the original definition of tradition, but proposed a new humility
value to capture the connotations of the alternative items.
(14) Humility
9
We defined the motivational goal of humility as recognizing one’s insignificance in
the larger scheme of things. Valuing humility entails attributing important to accepting what
one has without expecting more, to being self-effacing rather than boastful. Additional SVS
items that expressed humility were self-effacing and accepting my portion/submitting to life’s
circumstances. Additional PVQ items were humble/don’t draw attention to self.
(15) Benevolence-Caring
(16) Benevolence-Dependability
The original conceptual definition of benevolence referred only to caring for the
welfare of in-group members. However, in the MDS analyses of the five SVS benevolence
items, two (responsible/dependable and loyal/faithful to friends) separated consistently from
the items that clearly operationalized caring (e.g., helpful/working for others welfare). This
alerted us to another possible subtype of benevolence, which we tentatively labelled
dependability. The PVQ items did not permit discriminating a dependability subtype because
all referred to caring. To sharpen the distinction between benevolence-caring and
benevolence-dependability, we defined the latter as “being a reliable and trustworthy member
of the in-group.”
(17) Universalism-Tolerance
(18) Universalism-Concern
(19) Universalism-Nature
The conceptual definition of the original universalism value was “understanding,
appreciation, tolerance, and protection for the welfare of all people and for nature.” Schwartz
(1992) noted that this implied three subtypes. The MDS analyses of both the SVS and the PVQ
revealed distinct regions for all three universalism subtypes, tolerance (e.g., broadminded,
tolerant), societal concern (e.g., equality, social justice), and protecting nature (e.g., protect
the environment). The CFA study supported separating nature and concern components and
10
did not assess tolerance. We proposed three universalism values with distinct motivational
goals (see Table 1). All three entail concern for the welfare of those beyond the in-group.
The 19 values partition the same circular motivational continuum as the original ten
values (see Figure 1). Cieciuch et al. (2014) demonstrated this with a third order CFA of
values data from nine countries, gathered with the first version of the instrument developed to
measure the 19 refined values. Each of the 14 refined values derived from subtypes of the
original values loaded on a latent factor that represented the broad original value from which
it was split. Moreover, these ten latent values, together with the two new values introduced in
the refined theory (face and humility), loaded on latent factors representing the four higher
order values, self-transcendence, self-enhancement, conservation, and openness to change.
Reflecting their location between higher order values, face loaded on both conservation and
self-enhancement and humility loaded on both conservation and self-transcendence.1
Measuring the 19 Refined Values: The PVQ-RR
Researchers have noted various problems of measurement with instruments that
operationalized the original ten values, specifically, multicollinearity between adjacent
values, low internal reliabilities of some indexes, and cross-loadings of items on multiple
factors (e.g., Davidov, et al., 2008; Saris, Knoppen & Schwartz, 2013). The attempt to cover
multiple, diverse, substantive components in the conceptual definition of each basic value
with only a few items (Schwartz, 1992, 2005a, b) largely caused these problems. The revised
Portrait Values Questionnaire (PVQ-RR) to measure the 19 values reduces them. By defining
the broad values more narrowly, the refined theory made it possible to generate more sets of
items to measure each value that are more homogeneous.
The PVQ-RR includes three items per value, the minimum necessary for CFA
analyses. Because each value presumably consists of a single, narrowly defined conceptual
1 Figure 2 in Cieciuch et al. (2014) shows conformity with an unexpected secondary loading on self-
transcendence in addition to its primary loading on conservation. Conformity-interpersonal accounts for this. If
it is allowed to have a secondary loading on self-transcendence, conformity loads only on conservation.
11
component, the same number of items can be used to measure each value. Each item portrays
a different person in terms of his or her goals or aspirations that point implicitly to a value.
Respondents compare each person to themselves and indicate how similar the person is to
them. All items begin with the phrase, “It is important….” This is intended to insure that
people recognize what follows as a value aspired to, not as a person’s behavioral trait. Items
are gender-matched and consist of only one sentence, thereby avoiding double-barreled items.
Several characteristics of the PVQ method merit mention. First, it assumes that people
have latent, basic values that are expressed in and can be inferred from their responses, not
that they have articulated values. Second, it asks respondents to engage in a common
everyday activity, social comparison. This is easier than reporting one’s abstract values,
because people rarely think about such values. Third, respondents compare others, whose
value-relevant characteristics are described, to themselves, not themselves to the other. This
directs attention to the specific, value-relevant characteristics of the other, so the similarity
judgment is likely to focus on them. Comparing self to other might focus attention on salient,
but value-irrelevant aspects of one’s self-concept (Schwartz et al., 2001). Finally, it uses an
asymmetric response scale. Because values are generally socially desirable, people tend to
consider others who attribute importance to values as similar to self. Providing more
similarity than dissimilarity options captures this psychological asymmetry and permits finer
discrimination where needed on the scale (Schwartz, 2005b).
Respondents’ own values are inferred from the values of the people they describe as
similar to themselves. For example, a respondent who indicates that a person described by “It
is important to her to enjoy life’s pleasures” is similar to herself presumably attributes
importance to hedonism values. Additional examples are, “It … never to be humiliated”
(face) and “It … to be humble” (humility). Respondents indicate how similar each person is
to themselves on a 6pt scale: 1 – not like me at all, 2 – not like me, 3 – a little like me, 4 –
12
moderately like me, 5 – like me, 6 – very much like me. The full scale, which usually takes 4-
8 minutes to complete, is available from the author with coding and analysis instructions.
The desirability of values makes self-reports vulnerable to social desirability and
acquiescence biases; people tend to use the more favorable (higher) end of the response
scales. This is a problem when comparing values of different individuals or groups or
correlating values with other variables. To cope with the biases in these types of analyses,
Schwartz (1992, 2006a) recommended centering respondents’ ratings of each item on their
own mean rating of all the items by subtracting the mean from the score for each item.
Alternatively, the mean of all items can be used as a covariate. Verkasalo, Tuomivaara,
& Lindeman (1996) suggested yet another method with similar effects.
Distinctiveness of the 19 Refined Values
Two types of analyses provide evidence of whether people actually distinguish the 19
values and whether the measurement instruments adequately capture the distinctions people
make, CFA and MDS. Schwartz et al. (2012) report analyses with the PVQ5X, the first,
experimental scale designed to measure the 19 values. They gathered data from 15 samples
from 10 countries: Finland, Germany, Israel, Italy, New Zealand, Poland, Portugal,
Switzerland, Turkey, and the United States. Separate CFAs on the pooled within-sample
covariance matrixes, allowing no cross-loadings or covariances between uniquenesses,
yielded an acceptable fit, after dropping nine items. MDS analyses within each of the 15
samples revealed that at least 17 of the 19 values formed a distinct region in 80% of samples.
In almost all the remaining cases, the two intermixed values were values theorized to be
adjacent in the circle. Every value was discriminated in at least 12/15 samples and most were
discriminated in all samples (see Schwartz, et al., 2012, Table 7).
Schwartz and Butenko (2014) replaced the problematic items in the PVQ5X. They
then assessed the distinctiveness of the 19 values in a sample of 266 Russian students. An
13
MDS plot of these data revealed that every one of the 19 values formed a distinct region.
They also performed CFA analyses, allowing no cross-loadings or covariances between
uniquenesses. However, they examined each of the four higher order sets of values separately
(cf. Cieciuch & Schwartz, 2012; Saris, Knoppen & Schwartz, 2013). This eliminates
irrelevant sources of misspecification due to negative cross-loadings between values on
opposite sides of the motivational circle. Such loadings do not reflect on the distinctiveness
of adjacent values (Davidov, Schmidt, & Schwartz, 2008). After dropping five items, all four
models exhibited adequate fit statistics and supported the distinctiveness of the 19 values.
Schwartz et al. (submitted) assessed the distinctiveness of the values in samples from
Italy, Poland, Russia, and the USA with CFA. They used the same instrument and method of
analysis as Schwartz and Butenko (2014) and dropped an average of four items. The CFA in
each country supported distinct latent factors for the 19 values.
No CFAs have been done yet on data gathered with the PVQ-RR. However, we expect
CFAs to support the distinctions among the 19 values. This is because the MDS findings and
the reliability statistics of the values for 31 samples (12 adult, 14 student, 3 mixed student and
adult, and 2 adolescent samples) (M alpha = .71, sd = .09) are as good as or better than with
the previous versions. Samples are from 18 countries: Basque region, , Costa Rica, England,
France, French Canada, Greece, Iceland, Macedonia, Poland, Serbia, South Africa, Spain,
Switzerland, Russia, Turkey, Ukraine, USA, and Vietnam. Eleven values were distinguishable
in the MDS in every sample, seven in at least 27/31 samples, and security-personal was
distinguishable in 25 samples. The values that intermixed were two self-direction (once), two
security (three times), two benevolence (twice), and conformity-interpersonal with humility
(three times). These findings strongly suggest that respondents across diverse countries
discriminate the 19 values, measured with the PVQ-RR.
Circular Order of the 19 Values
14
The most common order of the 19 values in the refined theory corresponds to Figure
1. However, in the MDS of the typical sample one of the 19 values is slightly misplaced.
Across the 31 PVQ-RR samples, only two values moved outside the region of their higher
order value. The subtypes of the original ten values were all adjacent to one another. As
compared with the order in the original theory, the order of the benevolence and universalism
items reversed. Often, however, the benevolence subtypes emerged toward the center of the
circle, with the universalism subtypes peripheral to them. Schwartz, et al. (2012) speculate
about possible explanations for this change, but clarifying why and under what the conditions
this occurs requires further research. Schwartz, et al. (2012) also discuss the rationales for the
order of the subtypes within each broad value. I note only a few of the rationales that reflect
the impact of the dynamic underlying functions on the order (see Figure 1).
Hedonism falls between openness to change and self-enhancement. It is closer to
openness to change in most samples (Schwartz, 2006), however, because it promotes self-
expansion rather than protecting against anxiety. Achievement falls between the self-
protection and growth sectors because it has two goals, success according to social standards
in order to gain admiration and power (self-protection) and public confirmation of one’s
capabilities (growth). Face falls between power and security because it shares a concern with
power for maintaining control and with personal security for avoiding personal threat.
Humility falls between self-transcendence and conservation because its renunciation
of self-interest may reflect either self-expansive concern for others or self-restrictive
compliance with social expectations. Conformity-interpersonal emerges next to humility and
sometimes intermixes with it or reverses positions. Such conformity may motivate
individuals to consider the desires of others, not only to avoid their potential negative
reactions. In the original theory and MDS analyses, tradition values were peripheral to
conformity values, presumably reflecting their greater abstractness and stronger opposition to
15
openness (Schwartz, 1992, 2006). The more narrowly defined tradition value falls between
conformity-rules and societal security and is toward the periphery in about half the samples.
Relations of Values to Demographic Variables and Attitudes
Across societies, people discriminate the 19 more narrowly defined values. But do
researchers gain heuristic and explanatory power by distinguishing among these values? I
next provide some examples of insights that can be gained. I discuss relations of values with
gender, age, religiosity and a variety of attitudes. To obtain the most reliable possible
findings, I combined the samples that measured the refined values with the PVQ-RR (30
samples from 18 countries) and with the earlier PVQ5X (15 samples from 10 countries) and
PVQ-R (five samples from five countries) questionnaires. These 50 samples came from a
total of 25 different countries from six continents. I mention only statistically significant
differences (p < .05, 2 tailed).
Gender. First, consider relations of several of the refined values to gender. Schwartz
and Rubel (2005) reported that women consistently rated universalism values more important
than men did. This was also the case for the more narrowly defined facets of universalism-
concern and universalism-tolerance across the 50 samples. There was, however, no gender
difference at all for the universalism-nature facet. Women attributed more importance than
men only to the two facets of universalism values that involve relations to people; they
showed no greater concern for nature. Although women rated both universalism-concern and
universalism-tolerance values more important than men did, the association with gender was
stronger for universalism-concern than for universalism-tolerance in 72% of samples.
Concern for the welfare of others outside one’s in-group is more in accord with the
communal, feminine caring role than the more intellectual tolerance of them is.
Schwartz & Rubel (2005) also reported an inconsistent tendency for women to rate
security values more important that men did. For the narrowly defined security facets, women
16
attributed more importance to personal security values than men did in 88% of samples. This
fits the evolutionary and role-based reasons given to explain why women value security more
(Schwartz & Rubel, 2005). Women have a greater need to protect themselves and their
infants during the period of early child rearing. Moreover, their smaller size, lower status, and
greater dependence on others’ support, make them more vulnerable than men. These reasons
apply only to personal and not to societal security, however. Regarding societal security
values, men attributed more importance to this aspect of security than women did in 66% of
samples. This, too, fits both evolutionary and role-based analyses. In evolutionary history,
men were the ones who were typically tasked with hunting in peer groups for the community
and men dealt more with protecting the tribe in contacts with other tribes in the region. Even
today, in most countries men are more involved in the affairs of the wider society than
women are.
Schwartz & Rubel (2005) further reported that benevolence values were the values
women most consistently rated more important than men did. Across the 50 samples. women
also rated both facets of benevolence values more important than men did. However,
benevolence-caring correlated more positively with gender than benevolence-dependability
did in most samples. This accords with the idea that caring for close others is a more
communal, feminine role whereas being the one on whom others depend is a more agentic,
masculine role. We measured mastery and performance goal orientations (Elliot & McGregor,
2001) in 15 samples (Schwartz et al., 2012). In 14 of these samples, benevolence-caring
related more negatively than benevolence-dependability did to performance-avoidance, that
is, to performing worse than others, as a goal at work or school. This competitive goal is
incompatible with benevolence in general, but less so with the more self-assertive facet of
benevolence-dependability.
17
The newly defined value of face falls between the values power-resources and
personal security on the circle. Across the 50 samples, face was more important to females
than to males whereas power-resources was more important to males than to females.
Women’s greater concern with their public image and men’s greater concern with controlling
resources accord with evolutionary arguments about advantageous mating strategies (Buss,
1993/2004; Wright, 1994). The various gender differences for the refined values, while
consistent across samples, were generally small, as in Schwartz and Rubel (2005).
Age. Relations of values to age reveal interesting differences between the two facets
of conformity values. Conformity-rules values show a linear increase in importance with age
across the 17,714 respondents in the 50 samples. This fits the pattern of increased
embeddedness in and commitment to the institutions of society over the life course
(Hutchison, 2011) and the decreasing respect for government authority among younger
cohorts in many countries (Norris, 1999). Conformity-interpersonal values are uncorrelated
with age, however. Instead, their importance in different age groups reveals a U-shaped
association. There is a linear increase in their importance after age 40, peaking above age 60,
as one might expect for conformity values. Conformity-interpersonal values are also high
among teenagers. This may reflect the substantial concern with evaluation by peers common
during this developmental period (Brown, Clasen, & Eicher, 1986). However, the importance
of conformity-interpersonal values is lowest between ages 21 to 40. Perhaps this is because
people are more likely to be members of multiple groups that pose conflicting interpersonal
expectations during this period (e.g. work-family conflict).2
Religiosity. Relations of values to religious commitment confirm the importance of
separating the components of the original tradition value into more narrowly defined tradition
values and humility values. The respective correlations of religious commitment with
humility and with the narrower tradition value were, respectively, .041/.421 across samples
2 I thank Sonia Roccas for this suggestion.
18
from six historically Roman Catholic countries, .125/.495 across three Eastern Orthodox
countries, .037/.438 across four Protestant countries, .126/.559 in one Jewish country, and
-.037/.414 in one Muslim country. The finding that humility values are largely unrelated to
religiosity is striking. While maintaining traditional practices and ways of thinking is strongly
linked with religiosity, valuing being humble is just as important to secular as to religious
people.
Attitudes. Relations of values to a variety of attitudes also confirm the importance of
using the narrowly defined values to distinguish facets of the original broad values. Analyses
are based on attitude data available only for the 15 samples from 10 countries that responded
to the PVQ5X. Universalism-concern correlated more positively than universalism-tolerance
with the attitude that the government should reduce income differentials between rich and
poor in 14 samples. On the other hand, willingness to accept immigrants, an expression of
accepting diversity, related more positively to universalism-tolerance than to universalism-
concern in 13 samples.
Attitudes also support the distinction between self-direction-thought and self-direction
action. Self-direction thought correlated more positively with following politics closely and
wanting to learn as much as possible at work or school in 14 samples. Self-direction-action
correlated more positively in these samples with agreeing that individual freedom is an
inviolable right that must be maintained at all costs. These differences reflect the contrast
between a focus on intellectual vs. physical autonomy.
Several attitudes clarify the distinction between power-resources values and power-
dominance values. Not surprisingly, in every sample, power-resources correlated more highly
than power-dominance with agreeing that “it is hard to get ahead without a lot of money” and
that “a high income is most important in choosing a job.” More interestingly, in 14 samples,
valuing power-dominance correlated more highly than power-resources with agreeing that it
19
is “alright to impose our views on those who disagree with us” whereas power-resources
correlated more highly than power-dominance with agreeing that “the police should be given
more power so they can protect us.” And in all samples, power-dominance correlated more
with following politics closely than power-resources did. Indeed, across samples, the
correlation with following politics was positive for power-dominance and negative for power-
resources. This pattern of associations indicates that power-resources values concern being
financially well off personally and having one’s wealth protected, regardless of how power is
exercised in the society. In contrast, power-dominance values focus on exerting one’s power
over others and on the way power is distributed and used in the society.
Relations of Values to Behavior
Mechanisms Linking Values to Behavior
In most instances, attitudes mediate relations between values and behavior.
Consequently, correlations between values and specific behaviors are rarely very strong.
Later chapters in this book elaborate ways in which values are linked with behaviors. Here, I
briefly describe linking mechanisms that have guided my thinking (Schwartz 2009, 2016). In
order for a value to exert influence on an attitude or behavior, it must first be activated.
Accessibility increases the probability that a value will be activated, and more important
values are more accessible (Bardi, 2000). So important values are activated more often and
exert influence. Activation experiments are particularly important because they show that
values can cause behavior (cf. Sagiv, Sverdlik & Schwarz, 2011).
Second, a value must be experienced as relevant to the attitude or behavior in
question. Because basic values are abstract, their relevance is often not obvious. People may
see justice, for example, as relevant to discrimination based on race but not on gender. A
behavior is relevant to a value if performing that behavior has consequences for promoting or
inhibiting attainment of the goals of that value. Value activation often occur outside of
20
conscious awareness. People are more likely to become aware of drawing on their values to
assess potential alternatives if the alternatives entail high costs or have conflicting
implications for important values. Abstract values relate more consistently to specific
behaviors that have become instantiations of a value through experience or learning (Maio
2010). People then know specific ways to express the value in real-life situations so it may
even have an unmediated impact on behavior.
Third, values determine the valence people assign to the perceived consequences of
available actions. Actions are more attractive to the extent that they promote or protect valued
goals. High-priority values are central to the self-concept. Sensing an opportunity to attain
them sets off an automatic, positive, affective response to actions that will serve them.
Sensing a threat to value attainment sets off a negative affective response. Some behaviors
are ‘value-expressive’, that is, they are primarily compatible with one value and incompatible
with the opposing values in the circle (Bardi and Schwartz 2003). For example, manipulative
behaviors are compatible with power values and incompatible with benevolence values.
Values predict ‘value-expressive behaviors’ more strongly than behaviors that are compatible
with mutually conflicting values, called ‘value-ambivalent’ (Lönnqvist et al. 2013). For
example, for a person who ascribes low importance to stimulation but high importance to
conformity values, agreeing to go skydiving when pressured by friends is a ‘value
ambivalent’ behavior.
Fourth, values influence behavior by affecting attention, perception, and interpretation
of situations. People attend more to the aspects of situations that threaten important values or
offer opportunities to attain them. People may perceive a job offer as an opportunity for
greater self-direction or as a threat to security, depending on their value priorities. They may
interpret such an offer as a way to improve their family’s well-being or as a way to gain
personal power. Each perception or interpretation promotes a different line of action.
21
Finally, important values influence behavior because they promote planning
(Gollwitzer 1996). People are more likely to formulate plans for how to go about attaining the
goals of more important values. By promoting planning, value importance increases value-
consistent behavior. It focuses people on the pros of desired actions rather than the cons. It
increases people’s belief in their ability to reach valued goals and their persistence in the face
of obstacles and distractions.
Associations of Refined Values with Behavior
A major reason for studying values is the assumption that they can explain, influence,
and predict behavior (Roccas & Sagiv, 2010; Rokeach, 1973). Do each of the 19 values in the
refined theory relate in a distinctive manner to behaviors? Three studies have addressed this
question (Schwartz & Butenko, 2014; Schwartz et al., submitted; Lebedeva, Schwartz &
Plucker, submitted). The following presentation draws on analyses of value-behavior
relations in four countries, Italy, Poland, Russia, and the USA.
Most previous value-behavior studies focused on the values that propel behaviors. In
doing so, they ignored a key postulate of the value theory inherent in the notion of the
motivational circle: Any behavior is a product of tradeoffs between the values that promote
and oppose it. If a value on one side of the circle promotes a particular behavior, values on
the opposing side of the circle should simultaneously inhibit that behavior. I focus mainly on
the Schwartz et al. (submitted) study for two reasons. First, it examined the tradeoffs between
opposing values in predicting each behavior. Second, it used both self- and other-reports of
the frequency with which people performed each behavior, combining the two rating sources
to measure behavior more reliably.
The researchers generated a set of everyday behaviors they expected to be ‘value-
expressive’ of each of the 19 values. Each set was expected to be propelled primarily by one
value and to be inhibited by values on the opposing side of the motivational circle. Table 2
22
presents results of the multiple regressions of corresponding behaviors on each value and on
the opposing higher order value. These are results of hierarchical linear modeling analyses
(HLM6.1) across the samples from the four countries. The Table also shows the average
correlation across samples between the behaviors and the values expected to propel them and
the higher order values expected to inhibit them. Values were indexed with the centered factor
scores derived from a multi-group confirmatory factor analysis that established the invariance
of value measurement across countries. The Table provides one example of a behavior for
each value. The sets of behaviors included between three and five behaviors.
Table 2 about here
Table 2 reveals that each of the 19 refined values correlated positively and
significantly across the four samples with the behavior for which it was expected to be the
primary motivator. In the regression analyses, which also included the opposing higher order
value, each value also predicted the behavior it was expected to propel significantly across
samples. In the separate country samples, these correlations and regression coefficients were
significant and positive in almost every case. Prediction was most problematic for security-
personal values. They predicted behavior significantly in only two of the four countries,
although the coefficients were positive in the other countries too.
Table 2 also shows that each opposing higher order value correlated negatively and
significantly, across samples, with the behavior it was expected to inhibit. Higher order
values were used to assess the value tradeoffs because multiple values oppose each value
motivationally. In the regression analyses, only nine of the opposing values added significant
predictive power to the values that propel the behaviors. Self-transcendence values inhibited
self-enhancing behaviors (achievement excepted) and conservation values inhibited openness
to change behaviors. Stated more broadly, concern for the impacts of one’s behavior on others
(social-focused values) inhibited engaging in behaviors aimed at promoting or protecting
23
one’s own benefits and status and engaging in behaviors that might change the status quo. In
contrast, examining the distinctive contribution of self-enhancement values revealed that they
had little or no inhibiting effect on self-transcending behaviors (universalism-tolerance
excepted) and openness to change values had little or no distinctive inhibiting effect on
conservation behaviors. Thus, person-focused values did not inhibit behaviors aimed at
benefitting others or maintaining a safe, predictable environment.
The tradeoffs between the values that propel and those that oppose each behavior
explained an average of 16.0% of the variance in the behaviors. Values explained more than
25% of the variance in hedonism, stimulation, and power-resources behaviors, but less than
10% of the variance in personal- and societal-security behaviors. Perhaps, the relatively weak
impact of values on security behaviors indicates that inborn traits or temperaments or
situational factors play a larger role in driving these behaviors than in driving other
behaviors.
The above findings speak to the predictive validity of the 19 refined values. They do
not demonstrate the discriminant validity of the values, however. That requires establishing
that each value relates more strongly to the behaviors it presumably propels than any other
value does. For example, universalism-concern values should correlate more strongly than
universalism-tolerance values do with behaviors like collecting clothing for the needy, even
though these two facets of universalism are highly intercorrelated. Comparisons of the value-
behavior correlations in the matrix of the 19 values by 19 behaviors averaged across samples
supported the discriminant validity of the values. Each value correlated more highly with the
behavior it presumably propels than any other value did in 98% of the 342 (19 x 18)
comparisons. In 97% of comparisons, the correlation difference was significant.
The value with the weakest discriminant validity was achievement. Power-dominance,
power-resources, and face values correlated nonsignificantly more highly with such behaviors
24
as studying hard and trying to impress one’s boss than achievement values did. This suggests
that all the aspects of self-enhancement motivated these behaviors. Schwartz, et al.
(submitted) bring evidence that the deviant findings for achievement values may reflect
inclusion of two concepts in this value, success in attaining goals and recognition. Additional
research should determine whether it is better to split the achievement value.
For the study just described, the authors began with the 19 refined values. They then
chose behaviors that they presumed would be propelled primarily by one value and inhibited
primarily by one opposing higher order value. This permitted systematic and comprehensive
evaluations of the predictive and discriminant validity of all 19 refined values. Lebedeva,
Schwartz, and Plucker (submitted) began with the behaviors that interested them, overall
creative behavior and creativity in six specific domains. They used the PVQ-RR to identify
the values that underlie creative behavior. They generated hypotheses by considering the
values that creativity in each domain was likely to express or threaten. For example, they
expected stimulation values, which motivate pursuit of novel and exciting experience, to
promote creativity and tradition values, which motivate preserving accepted ways of
thinking, to inhibit creativity. This analysis led them to combine some facets into broad
values (e.g., combining self-direction thought and self-direction action into a broad self-
direction value) for the hypotheses but to use refined values for others (e.g., universalism-
nature [linked to aesthetics] and security-personal).
Lebedeva et al. (submitted) tested their hypotheses in representative samples of adults
from two cultural regions of Russia (N = 2046). A regression of overall creativity on the
values and demographics confirmed the hypotheses. Self-direction, universalism-nature, and
stimulation values promoted overall creativity. In contrast, humility, face, security-personal,
tradition, conformity, and power values inhibited overall creativity.3 The study also examined
3 Lebedeva et al. (submitted) combined humility and tradition values. I separated them and reran the
analyses.
25
predictors of creativity in the domains of crafts, plastic arts, music and dance performance,
theater, products for work, and machine graphics. In at least five of the six domains, self-
direction and universalism-nature values promoted creativity and humility and face values
inhibited creativity. Unexpectedly, benevolence values also promoted creativity in five
domains and overall. The authors suggested that benevolence does not motivate creativity
directly; it does so by motivating cooperation and contribution in creative group activity. As
expected, other values had no consistent associations with creativity.
Conclusion
The original theory of human values discriminated ten basic values, four higher order
values, and two polar value dimensions. As this volume attests, it has proven most fruitful.
The refined theory retains all the benefits of the original theory and adds to it the possibility
of representing the motivational continuum of human values even more finely. This chapter
has sought to demonstrate that making finer discriminations in the motivational circle is both
justified and desirable. It has shown that people in diverse cultures actually do discriminate
the 19 values in the refined theory. So using them in research does not mean imposing
scientific constructs on respondents that are not meaningful to them. Each of the 19 values
has unique associations with various background variables, attitudes, and behaviors. These
values do provide greater heuristic and predictive power, as anticipated in the seminal chapter
on the original theory (Schwartz, 1992).
Measuring the 19 values with the PVQ-RR (and perhaps other instruments yet to be
designed) yields reliable indexes. The PVQ-RR is easy for respondents and requires the same
or less time to complete than any reliable measure of the ten values. Of course, not every
topic we study will benefit from making all 19 distinctions, as the Lebedeva et al. (submitted)
study illustrates. It is therefore important to note that the 19 values can be combined to yield
indexes of the ten and four broader values. Moreover, these indexes are more reliable than
26
previously available indexes were. The added face and humility values may be kept separate
or included in the four broader values, face in self-enhancement or conservation, humility in
conservation or self-transcendence. The fact that these two values fall on boundaries and
include elements of adjacent higher order values reemphasizes the point that any partitioning
of the motivational continuum is arbitrary. Further research may yet suggest a more useful
partitioning.
In conclusion, this chapter is a call to researchers to adopt the refined theory of basic
values. The refined theory offers everything provided by the original theory and more.
27
References
Bardi, A. (2000). Relations of values to behavior in everyday situations. Unpublished
doctoral dissertation. The Hebrew University, Jerusalem.
Bardi, A., & Schwartz, S. H. (2003). Values and behavior: Strength and structure of relations.
Personality and Social Psychology Bulletin, 29, 1207-1220.
Beierlein, C., Davidov, E., Schmidt, P., Schwartz, S. H., & Rammstedt, B. (2012). Testing the
discriminant validity of Schwartz’ Portrait Value Questionnaire items – A replication
and extension of Knoppen and Saris (2009). Survey Research Methods, 6, 25-36.
Brown, B. B., Clasen, D. R., & Eicher, S. A. (1986). Perceptions of peer pressure, peer
conformity dispositions, and self-reported behavior among adolescents. Developmental
Psychology, 22, 521-530.
Buss, D. M. (1994/2003). The evolution of desire: Strategies of human mating. New York:
Basic Books
Cieciuch, J., Davidov, E., Vecchione, M., & Schwartz, S. H. (2014). A hierarchical structure
of basic human values in a third-order confirmatory factor analysis. Swiss Journal of
Psychology, 73, 177-182.
Cieciuch, J., & Schwartz, S. H. (2012). Assessing the number of distinct basic values and
their structure in Poland. Journal of Personality Assessment, 1-8.
Davidov, E., Schmidt, P., & Schwartz, S. H. (2008). Bringing values back in: The adequacy
of the European Social Survey to measure values in 20 countries. Public Opinion
Quarterly, 72, 420-445.
Elliot, A. J., & McGregor, H. A. (2001). A. 2 x 2 achievement goal framework. Journal of
Personality and Social Psychology, 80, 501–519.
Gollwitzer, P.M. (1996). The volitional benefits of planning. In: P.M. Gollwitzer and J.A.
Bargh (eds.), The psychology of action. New York: Guilford.
28
Hutchison, E. (2011). Dimensions of human behavior: The changing life course. London:
Sage.
Lebedeva, N., Schwartz, S. H., Plucker, J. A. (submitted). Domains of everyday creativity and
the personal values that underlie them.
Lönnqvist, J-E., Leikas, S., Paunonen, S., Nissinen, V., & Verkasalo, M. (2006). Conformism
moderates the relations between values, anticipated regret, and behavior. Personality
and Social Psychology Bulletin, 32, 1469-1481.
Norris, P. (ed.) (1999). Critical citizens: Global support for democratic government. Oxford:
Oxford University Press.
Roccas, S., & Sagiv, L. (2010). Personal values and behavior: Taking the cultural context into
account. Social and Personality Compass, 4, 31-41.
Rokeach, M. (1973). The nature of human values. New York: Free Press.
Sagiv, L, Sverdlik N, Schwarz N. (2011). To compete or to cooperate? Values' impact on
perception and action in social dilemma games. European Journal of Social
Psychology, 41, 64–77.
Saris, W. E., Knoppen, D., & Schwartz, S. H. (2013). Operationalizing the theory of human
values: Balancing homogeneity of reflective items and theoretical coverage. Survey
Research Methods, 7, 29-44.
Schwartz, S. H. (1992). Universals in the content and structure of values: Theory and
empirical tests in 20 countries. In M. Zanna (Ed.), Advances in experimental social
psychology (Vol. 25, pp. 1-65). New York: Academic Press.
Schwartz, S. H. (2005a). Basic human values: Their content and structure across countries. In
A. Tamayo & J. B. Porto (Eds.), Valores e comportamento nas organizações [Values
and behavior in organizations] pp. 21-55. Petrópolis, Brazil: Vozes.
29
Schwartz, S. H. (2005b). Robustness and fruitfulness of a theory of universals in individual
human values. In A. Tamayo & J. B. Porto (Eds.), op. cit. pp. 56-95.
Schwartz, S. H. (2006). Les valeurs de base de la personne: Théorie, mesures et applications
[Basic human values: Theory, measurement, and applications]. Revue Française de
Sociologie, 47, 249-288.
Schwartz, S. H., & Boehnke, K. (2004). Evaluating the structure of human values with
confirmatory factor analysis. Journal of Research in Personality, 38, 230-255.
Schwartz, S. H., & Butenko, T. (2014). Values and behavior: Validating the refined value
theory in Russia. European Journal of Social Psychology, 44, 719-813.
Schwartz, S.H., Cieciuch, J., Vecchione, M., Davidov, E., Fischer, R., Beierlein, C., Ramos,
A., Verkasalo, M., Lönnqvist, J.-E., Demirutku, K., Dirilen-Gumus, O., & Konty, M.
(2012). Refining the theory of basic individual values. Journal of Personality and
Social Psychology, 103, 663-688.
Schwartz, S.H., Cieciuch, J., Vecchione, M., Torres, C., Dirilem-Gumus, O., & Butenko, T.
(Submitted). Value tradeoffs and behavior in four countries: Validating 19 refined
values.
Schwartz, S. H., Melech, G., Lehmann, A., Burgess, S., Harris, M. & Owens, V. (2001).
Extending the cross–cultural validity of the theory of basic human values with a
different method of measurement. Journal of Cross–Cultural Psychology, 32, 519–542.
Schwartz, S. H., & Rubel, T. (2005). Sex differences in value priorities: Cross-cultural and
multi-method studies. Journal of Personality and Social Psychology, 89, 1010-1028.
Verkasalo, M., Tuomivaara, P., & Lindeman, M. (1996). 15-year-old pupils’ and their
teachers’ values, and their belief about the values of an ideal pupil. Educational
Psychology, 16, 35–47.
30
Wright, R. (1994). The moral animal: Evolutionary psychology and everyday life. London:
Vintage.
31
Table 1
The 19 values in the refined theory, each defined in terms of its motivational goal
Value Conceptual definitions in terms of motivational goals
Self-Direction—Thought Freedom to cultivate one’s own ideas and abilities
Self-Direction—Action Freedom to determine one’s own actions
Stimulation Excitement, novelty, and change
Hedonism Pleasure and sensuous gratification
Achievement Success according to social standards
Power—Dominance Power through exercising control over people
Power—Resources Power through control of material and social resources
Face Maintaining one’s public image and avoiding humiliation
Security—Personal Safety in one’s immediate environment
Security—Societal Safety and stability in the wider society
Tradition Maintaining and preserving cultural, family or religious
traditions
Conformity—Rules Compliance with rules, laws, and formal obligations
Conformity—Interpersonal Avoidance of upsetting or harming other people
Humility Recognizing one’s insignificance in the larger scheme of
things
Universalism—Nature Preservation of the natural environment
Universalism—Concern Commitment to equality, justice and protection for all
people
Universalism—Tolerance Acceptance and understanding of those who are different
from oneself
Benevolence—Caring Devotion to the welfare of in-group members
Benevolence—Dependability Being a reliable and trustworthy member of the in-group
From Schwartz et al. (2012).
32
Table 2. Correlations and regressions of behaviors on their corresponding value and opposing higher order value. Hierarchical linear model (HLM)
across four countries (N=2016)
Value-Behavior Domain
Opposing Higher Order Mean r b R2
Examples of Behavior ItemsA
How often did you/(she, he) engage in this behavior during the past year?
Self-Direction-Thought SDT
Conservation
.36**
-.33**
.25** .152
-.18**
Develop my/her own opinion on a topic in the news by studying what was written
on it.
Self-Direction-Action SDA
Conservation
.37**
-.43**
.16* .195
-.29**
Do something my/his way even if someone might disapprove.
Stimulation ST
Conservation
.48**
-.50**
.29** .288
-.34**
Look for exciting activities to break up my/her routine.
Hedonism HE
Conservation
.44**
-.46**
.28** .294
-.37**
Indulge myself/himself by buying things that I/he didn’t really need.
Achievement AC
Self-Transcendence
.36**
-.23**
.17** .130
-.04
Study or work late into the night before an exam or project due date even though
I/she was already well prepared.
Power-Dominance POD
Self-Transcendence
.43**
-.34**
.23** .213
-.34**
Manipulate others to get what I/he want/s.
Power-Resources POR
Self-Transcendence
.49**
-.38**
.34** .255
-.16**
Mention to other people how valuable some of my/her possessions are.
Face FAC .31** .15** .136 Feel anxious that someone might think I/he did something immoral.
33
Self-Transcendence -.32** -.22**
Security-Personal SEP
Openness to change
.21**
-.13**
.15** .043
-.02
Check the expiration date on products before buying or using them.
Security-Societal SES
Openness to change
.33*
-.18**
.28** .098
-.07*
Praise government actions against groups that endanger the public.
Tradition TR
Openness to change
.42**
-.27**
.23** .166
-.05
Celebrate national or ethnic group holidays.
Conformity-Rules COR
Openness to change
.26**
-.16**
.23** .115
-.02
Pay the full entry fee or fare, even when I/she could get away with not paying it.
Conformity-Interpersonal COI
Openness to change
.29**
-.13**
.28** .142
-.02
Keep my opinion to myself/himself rather than disagree openly with others.
Humility HU
ST, HE, AC, POD
.19**
-.11**
.18** .112
-.01
Refrain from mentioning something I/she did because it would be boasting.
Universalism-Nature UNN
Self-Enhancement
.40**
-.12**
.23** .147
-.02
Avoid buying items that might harm the environment.
Universalism-Concern UNC
Self-Enhancement
.25**
-.13**
.19** .113
-.03
Collect food, clothing, or other things for needy families.
Universalism-Tolerance UNT
Self-Enhancement
.28**
-.26**
.23** .155
-.13**
In a conversation, say we should be tolerant toward people with different lifestyles.
Benevolence-Caring BEC .35** .29** .144 Help friends or family with tasks like moving, getting to the airport or station, baby-
34
Self-Enhancement -.10** .03 sitting, etc.
Benevolence-Dependability
BED
Self-Enhancement
.34**
-.14**
.32** .137
-.03
Do what I/he committed myself to do for a family member.
** p < .01, * p < .05, 1-tailed
A Each behavior was measured on a 5pt scale of the frequency of performance relative to the number of times one had an opportunity to do so in the
past year. Scale points were 0—never, 1—rarely (about a quarter of the times), 2—sometimes (about half the times), 3—usually (more than half the
times), 4—always. ‘Never’ was used in less than 1% of responses. Self-reported and other-reported frequencies were combined. The behavior items
were gender matched.
35
Figure 1. Circular motivational continuum of 19 values in the refined value theory (from
Cieciuch, et al., 2014)
