Shalom H.

Schwartz
The PVQ40 is available at no cost for academic research purposes.
For commercial use, a license is required.
Contact [email protected] to obtain a license.

Scoring and Analysis Instructions

June 2021
For a presentation of the theory underlying the PVQ40 and validating data, see Schwartz
(1992, 2004a, 2004b and 2016) in the references below. For instructions for different types of
statistical analysis, see below ‘correcting for scale use bias’. In languages that distinguish
gender grammatically, the separate Male and Female versions should be used.

Scoring Key for PVQ 40 (aka PVQIV) Value Scale

The score for each value is the mean of the raw ratings given to the items listed below for that
value.

Values Items in Index

Conformity 7,16,28,36

Tradition 9,20,25,38

Benevolence 12,18,27,33

Universalism 3,8,19,23,29,40

Self-Direction 1,11,22,34

Stimulation 6,15,30

Hedonism 10,26,37

Achievement 4,13,24,32

Power 2,17,39

Security 5,14,21,31,35

Scoring Key for four Higher Order Values in the PVQ40 Value Scale

Self-Transcendence Combine means for universalism and benevolence

Self-Enhancement Combine means for achievement and power

Openness to change Combine means for self-direction, stimulation and hedonism

Conservation Combine means for security, conformity and tradition

Correcting for scale use biases

Shalom H. Schwartz
For many types of analyses, it is highly desirable to make a correction for individual
differences in use of the response scale.1 Below are instructions indicating whether a
correction is appropriate for each type of analysis and how to make the correction. Failure to
make the scale use correction when appropriate may lead to mistaken conclusions!

Individuals and cultural groups differ in their use of the response scale.2 Scale use differences
often distort findings and lead to incorrect conclusions.3 To correct for scale use:

(A) Compute scores for the 10 values by taking the means of the items that index it
(above). If you wish to check internal reliabilities, do so for these value scores before the
next steps.
(B) Compute each individual’s mean score across all 40 value items. This is the
individual’s Mean RATing of all values. Call this MRAT.4
(C) Subtract MRAT from each of the 10 value scores. This centers the scores of each of
the individual’s 10 values (computed in A) around that individual’s Mean (MRAT).

1. For correlation analyses: Use the centered value scores (C).

2. For group mean comparisons, analysis of variance or of covariance (t- tests, ANOVA,
MANOVA, ANCOVA, MANCOVA): Use the centered value scores as dependent
variables.

3. For regression:

a. If the value is your dependent variable, use the centered value score.

b. If the values are predictor variables:

1. Enter uncentered values as predictors in the regression.
a’ If all 10 values are included, the single regression coefficients for
the values are not clearly interpretable because the values are
completely interdependent. This is so even if the
multicollinearity statistics do not look problematic.
b’ If only some of the values are relevant to the topic, include only
those a priori, to reduce interdependence and obtain more
interpretable regression coefficients.
2. If you are interested only in the total variance accounted for by
values and not in the regression coefficients, you may include all 10
uncentered values as predictors. The R2 is meaningful but, because
the 10 values are exactly linearly dependent, the coefficients for each
value are not precisely interpretable.
3. If you use only one value as a predictor, use the centered value
because this is equivalent to correlation.

c. In publications, it is advisable to provide a table with the correlations between

the centered values and the dependent variables in addition to the regression
results. These correlations will aid in understanding results and reduce
confusion due either to multicolinearity or to intercorrelations among the
values.

4. For multidimensional scaling (MDS), both centered and uncentered item responses
work equally well.
Shalom H. Schwartz
5. For SEM and for canonical, discriminant, or confirmatory factor analyses:
Use raw (uncentered) value scores for the items or for the 19 value means.5

6. Exploratory factor analysis is not suitable for discovering the theorized set of relations
among values because they form a quasi-circumplex, which EFA does not reveal. Factors
obtained in an EFA with rotation will only partly overlap with the 10 values, will exploit
chance associations, and will combine them to form larger factors. The first unrotated
factor represents the way respondents use the response scale. It may represent an
acquiescence bias, social desirability, the overall importance of values to the person, or
some combination of these and other influences. It does not represent specific value
preferences. A crude representation of the circular structure of values can be obtained
using EFA by plotting the value items in a two-dimensional space according to their
loadings on factors 2 and 3 of the unrotated solution.

Footnotes

1. For a review of some conceptual and statistical issues in correcting for scale use with
basic values, including a simulation of the effects of the simple method of centering
(ipsatizing), see Rudnev (2021). Other methods that correct for scale use may also be
used (e.g., Van Rosemalen et al., 2010).

2. For a discussion of the general issue, see Saris (1988). Rudnev (2021) summarizes
studies that examine meanings of such scale use as an individual difference variable.
Smith (2004) discusses correlates of scale use differences at the level of cultures.

3. Two critical assumptions underlie these corrections.

(1) The motivational circle captured by the set of 10 individual level values is
reasonably comprehensive of the full motivational space of values recognized
across individuals and cultural groups. Empirical evidence supports this
assumption (Schwartz, 1992, 2004; Schwartz & Cieciuch, 2021).
(2) Studies of value priorities are concerned with the importance of particular
values as part of the value system of a person or group. This is because the
way values affect cognition, emotion, and behavior is through a trade-off or
balancing among multiple values that are simultaneously relevant to a
decision or action (cf. Schwartz et al., 2016). The relevant values often have
opposing implications for the decision or action. The absolute importance of
a single value across individuals or across groups ignores the fact that values
function as a system (Schwartz, 1996). The scale use correction converts
absolute value scores into scores that indicate the relative importance of each
value in the value system, i.e., the individual’s value priorities.

4. We center within person rather than standardizing (i.e., we do not divide by

individuals’ standard deviation across the 40 items). This is because individual
differences in variances of value ratings are usually meaningful. Even if, on average,
individuals attribute the same mean importance to the set of values, some individuals
discriminate more sharply among their values and others discriminate less sharply.
Standardizing that makes everyone’s variance the same (i.e., 1) would eliminate these
real differences in the extent to which individuals discriminate among their values

5. Centering creates a small degree of linear dependence among the items. This may be
problematic in these analyses. The scale use problem is avoided or eliminated by
Shalom H. Schwartz
other aspects of these analyses without centering. See Closs (1996) and Cornwell &
Dunlop (1994).

References [starred are available as electronic files]

Closs, S. J. (1996). On the factoring and interpretation of ipsative data. Journal of

Occupational and Organizational Psychology, 69, 41-47.

Cornwell, J. M., & Dunlop, W. P. (1994). On the questionable soundness of factoring ipsative
data: A response to Saville and Willson (1991). Journal of Occupational and Organizational
Psychology, 67, 89-100.

Van Rosmalen J, Van Herk H, Groenen PJF. (2010) Identifying response styles: A latent-
class bilinear multinomial logit model. Journal of Marketing Research, 47, 157–172.

Rudnev M. (2021) Caveats of non-ipsatization of basic values A review of issues and a

simulation study. Journal of Research in Personality, 93, 104-118.

Saris, W. E. (Ed) (1988). Variation in response functions, A source of measurement error in

attitude research. Amsterdam: Sociometric Research Foundation.

*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. (1996). Value priorities and behavior: Applying a theory of integrated value
systems. In C. Seligman, J.M. Olson, & M.P. Zanna (Eds.), The Psychology of Values: The
Ontario Symposium, Vol. 8 (pp.1-24). Hillsdale, NJ: Erlbaum.

*Schwartz, S. H. (2004a). Basic human values: Their content and structure across countries.
In A. Tamayo & J Porto (Eds.), Valores e trabalho [Values and work]. Brasilia: Editora
Universidade de Brasilia.

*Schwartz, S. H. (2004b). Robustness and fruitfulness of a theory of universals in individual

human values. In A. Tamayo & J. Porto (Eds.), Valores e trabalho [Values and work].
Brasilia: Editora Universidade de Brasilia.

*Schwartz, S. H. (2016). Basic individual values: Sources and consequences. In D. Sander

and T. Brosch (Eds.), Handbook of value (pp.63-84). Oxford: UK, Oxford University Press.

Smith, P. B. (2004). Acquiescent response bias as an aspect of cultural comminications style.

Journal of Cross-Cultural Psychology, 35, 50-61.