CULTURAL AND EVOLUTIONARY COMPONENTS

OF MARITAL SATISFACTION
A Multidimensional Assessment of Measurement Invariance

TODD LUCAS
Wayne State University
MICHELE R. PARKHILL
University of Washington
CRAIG A. WENDORF
University of Wisconsin–Stevens Point
E. OLCAY I⋅ MAMOĞLU
Middle East Technical University, Ankara, Turkey
CAROL C. WEISFELD
University of Detroit Mercy
GLENN E. WEISFELD
Wayne State University
JILIANG SHEN
Beijing Normal University

Couples assess their satisfaction with one another according to numerous culturally determined criteria.
However, evolutionary perspectives on marriage emphasize that husbands and wives are also concerned with
their adaptive fitness, and this suggests that some aspects of marital satisfaction may be cross-culturally homoge-
nous. We examined whether marital satisfaction reflects both ‘culturally unique’ and ‘adaptively universal’ con-
cerns of husbands and wives. Approximately 2000 couples from Britain, Turkey, China and the United States
completed a multidimensional measure of marital satisfaction that we assessed for measurement invariance.
Measures of romantic love and spousal support functioned similarly for couples within all four cultures, indi-
cating the possibility of a ubiquitous pair-bonding component of marital satisfaction. However, invariant mea-
surement structure was less robust across these samples, suggesting a culturally derived component of marital
satisfaction. In general, results suggest that invariance analyses may be used to elucidate cultural and evolution-
ary perspectives on marriage.

Keywords: marriage; marital satisfactionl; invariance; love; Turkey; China; United Kingdom; United States

Interpersonal relationships are heavily guided by norms, customs, and expectations that
are derived from culture (for reviews, see Berscheid, 1995; Fiske, Kitayama, Markus, &
Nisbett, 1998). In particular, satisfaction with one’s spouse may largely depend on the
degree to which a marriage fulfills culturally determined expectations and obligations. On

AUTHORS’ NOTE: Portions of this article were prepared while the first author was a postdoctoral researcher at the Center for
Behavioral and Decision Sciences in Medicine (Ann Arbor Veterans Affairs Hospital and University of Michigan). Portions of
this research were presented at the 2003 annual meeting of the Human Behavior and Evolution Society, in Lincoln, Nebraska.
We are grateful to Robin J. H. Russell and Pamela A. Wells for their contributions to this research. In addition, we acknowledge
the helpful comments of Cindy Gallois and two anonymous reviewers on prior versions of this article. Correspondence concern-
ing this article may be addressed to Todd Lucas, Division of Occupational and Environmental Health, Department of Family
Medicine and Public Health Sciences, Wayne State University, 3800 Woodward Avenue, Suite 808; Detroit, MI 48201. E-mail:
tlucas@med.wayne.edu.

JOURNAL OF CROSS-CULTURAL PSYCHOLOGY, Vol. 39 No. 1, January 2008 109-123

DOI: 10.1177/0022022107311969
© 2008 Sage Publications
109
110 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

the other hand, evolutionary research and theory has emphasized that besides satisfying
culture-specific functions, marriage must also enhance the basic adaptive objectives of
both husbands and wives. Specifically, marriage should facilitate not only procreation but
also appropriate caregiving behaviors for offspring, and marital satisfaction may be impor-
tant to the extent that it enables this.
The juxtaposition of cultural and evolutionary perspectives poses an interesting question for
marital satisfaction research—namely, in what ways can satisfying marriages be both culturally
unique and universally adaptive? In the present study, we examined the measurement invari-
ance of two indices of marital satisfaction (romantic love and communication/support) in cou-
ples from the United States, Britain, Turkey, and China. To the extent that evolutionary and
cultural perspectives respectively emphasize some alternative themes of marriage, our purpose
was to determine whether two specific criteria that couples might use to define a satisfying mar-
riage should be considered inherently universal or culturally unique.

MARITAL SATISFACTION AND CULTURE

Marriages are generally more successful when spouses establish a sense of satisfaction with
one another. However, researchers tend to diverge in their view of how satisfaction with one’s
spouse is attained. Some of the most important and interesting distinctions are posited by cul-
tural and evolutionary perspectives on marriage. From a cultural vantage point, marital satis-
faction may be enhanced to the extent that a marriage fulfills culturally determined expectations
and obligations of husbands and wives. In particular, the criteria for a satisfying marriage may
be highly varied and may depend on a unique set of culturally enforced norms, values, and
obligations. For example, a traditional Chinese marriage may be satisfying to the extent that it
fulfills familial duties that include the production of a male heir for the continuance of a fam-
ily line, the acquisition of a daughter-in-law who will provide support for the husband’s parents,
and the begetting of sons who will provide for the security of the couple in their old age (Wang,
1994). In addition, traditional Chinese marriages often represent the formation of an alliance of
two extended families, whose interests supplant those of the to-be-married couple (Dion &
Dion, 1993). On the other hand, Western cultures generally view marriage as serving fewer
instrumental and more personal functions and are thus thought to be satisfying to the extent that
they fulfill happiness or hedonistic goals of husbands and wives (e.g., Lalonde, Hynie, Pannu,
& Tatla, 2004). In addition, although social obligations are a defining feature of marriage in
many Eastern cultures, such influences may be viewed as obstacles to personal happiness in
Western cultures (Fiske et al., 1998).

WITHIN-CULTURE HETEROGENEITY AND MARITAL SATISFACTION

Cross-national differences are one defining feature of the cultural perspective. However, an
equally important facet concerns the influence of economic and sociopolitical contexts within
a culture (e.g., Zavalloni, 1975). In China, for example, the New Marriage Law of 1950 for-
mally abolished the traditional arranged marriage system and put in its place a system of free
choice and equal rights for both husbands and wives. China has also been informally influenced
by Western thought that has further altered attitudes about marriage (Higgins, Zheng,
Liu, & Hui Sun, 2002). However, a majority of the Chinese population live in rural areas,
where formal and informal influences are not as powerful. Accordingly, covertly evolved
forms of traditional marriages can still occur in China (Wang, 1994). Thus, there is also
 Lucas et al. / MARITAL SATISFACTION INVARIANCE 111

within-culture heterogeneity of Chinese marriage in that it comprises both modern and

more traditional arrangements. Similar heterogeneity has also been observed in Turkey
(I⋅ mamoğlu & Yasak, 1997) and undoubtedly exists in many other nations. Generally, then,
sociopolitical contexts may create varied criteria for marital satisfaction not only between
cultures but also within them.
In addition to the heterogeneity created by sociopolitical variation, differences between
husbands and wives from the same culture may also contribute to variability in marital sat-
isfaction criteria. In this vein, researchers have hypothesized that men and women hold dif-
ferent instrumental and interpersonal views not only of marriage (e.g., Bernard, 1972) but
also of romantic relationships (e.g., Sprecher & Toro-Morn, 2002). One often overlooked
possibility is that spousal discrepancies in marital satisfaction may also depend on culture.
For example, although British husbands and wives may think similarly about romantic
love but differently about spousal support, Chinese couples may agree on aspects of
spousal support but differ in their views of romantic love.

MARITAL SATISFACTION AND EVOLUTION

Bolstered by recognition of both between- and within-culture differences, a general

consensus is that cultural perspectives on marriage often emphasize heterogeneous quali-
ties of marital satisfaction criteria. A somewhat contrasting viewpoint is offered by an evo-
lutionary perspective on marriage, in which emphasis is often given to more universal
criteria. Although evolutionary mechanisms may produce cultural heterogeneity, evolu-
tionary research has tended to articulate a more ubiquitous and primal view not only on
mating and marital preferences (e.g., Buss et al., 1990) but also on marital satisfaction
(Shackelford & Buss, 1997). In this vein, evolutionary research has revealed that many
patterns in mate selection and marital stability are cross-culturally stable.
From an evolutionary perspective, mating behavior—including mate selection, marital
satisfaction, and divorce—is favored to the extent that it promotes individual reproductive
success. Both sexes tend to seek mature, opposite-sex members of their species who are not
close kin, which makes adaptive sense. However, men and women also exhibit somewhat dif-
ferent ideal criteria in mate selection. For example, cross-cultural surveys suggest that men
generally prefer to marry younger, more attractive women who are high in reproductive
potential, whereas women prefer to marry older and wealthier men, who generally excel in
providing material resources and social status for the family (Buss, 1989). Other universal
features of mating behavior can also be explained in adaptive terms. For example, marital sta-
bility around the world is associated with fecundity: the more children, the less chance of
divorce (Goode, 1993). In addition, divorce in almost every culture is predicted by infertility
and infidelity, both of which may hinder individual fitness (Betzig, 1989).
Marriage itself may also be interpreted in comparative terms as pair bonding—an
evolved characteristic of species with highly dependent young that is thought to facilitate
parental caregiving (Ember & Ember, 1979; Kleiman, 1977). Humans qualify as belong-
ing to those species with highly dependent young. Therefore, from an evolutionary stand-
point, it is not surprising that pair bonding, or marriage, is found in virtually every human
culture and is the generally preferred arrangement for raising children. Evidence abounds
that children generally develop best if raised by both biological parents (e.g., Amato, 1993;
Daly & Wilson, 1988), and marital satisfaction and stability are generally lower with
stepchildren than with biological children (Rogers & White, 1998; Teachman & Heckert,
112 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

1985). Thus, evolutionists suggest that marriage is more than simply a cultural institution;
it has an evolved basis. Because of its ubiquity, it may have generally been the most effec-
tive arrangement in our species’ past for raising offspring and passing on the parents’
genes. Of course, this is not to say that heterosexual marriage is the only arrangement for
successfully raising children in contemporary societies.
Evidence for an adaptively evolved propensity to marry may also include the universal
emotional tendency to fall in love (e.g., Buss, 1988; Jankowiak & Fischer, 1992). This
emotion usually includes sexual attraction as well as amorousness (Money & Ehrhardt,
1972). Although some have suggested that the importance of romantic love in marriage is
culturally varied (for review, see Landis & O’Shea, 2000), evidence for an evolved basis
to this social emotion includes its hormonal supports, especially the effect of oxytocin on
social bonding in humans and other mammals (Insel, 1997). Recently, a hormonal link
between pair bonding and paternal care has also been reported: Husbands of expectant
women undergo a rise in prolactin, which increases paternal inclinations after the birth
(Story, Walsh, Quinton, & Wynne-Edwards, 2000). In sum, although cultural perspectives
have emphasized unique criteria for marital satisfaction, evolutionary research has sug-
gested that marital arrangements serve adaptive fitness interests and that these should
relate to some aspects of marital satisfaction in a cross-culturally transcendent manner.

THE PRESENT RESEARCH

Although they differ in their emphasis on some important tenets of marital satisfaction,
cultural and evolutionary outlooks are not mutually exclusive of one another. Undoubtedly,
evolutionary mechanisms are likely to produce some unique cultural transmissions, and
many cultural norms and expectations serve some adaptive function. Furthermore, it would
be dubious to suggest that cultural and evolutionary perspectives respectively emphasize
only heterogeneous and homogenous qualities of marital satisfaction or that they are oth-
erwise incompatible with one another. Nevertheless, to the extent that cultural and evolu-
tionary perspectives often lead to theory and research on unique versus universal
characteristics of marriage, there are at least some alternative points of emphasis. The jux-
taposition of culture and evolution thus frames an important question for marriage
research—namely, in what ways are marital satisfaction criteria universally similar and
culturally unique? In addition, do similarities and differences in the composition of mari-
tal satisfaction across cultures map onto themes that seem to be emphasized by evolution-
ary and cultural viewpoints?
One way of formally exploring evolutionary and cultural perspectives on marital satis-
faction is to undertake a cross-cultural examination of the measurement invariance of mar-
ital satisfaction indices. Invariance testing is a form of covariance structure analysis
designed to assess whether measures, such as those composing marital satisfaction, are
defined similarly across different groups (e.g., Byrne & Campbell, 1999; Cheung &
Rensvold, 1999; Vandenberg & Lance, 2000). Invariance testing is increasingly commonly
used in test construction and scale development endeavors to assess the measurement
equivalency of newly established instruments (e.g., Lucas, Alexander, Firestone, &
LeBreton, 2007; Lucas, Michalopoulou, Falzerano, Menon, & Cunningham, in press).
However, invariance testing has also gained prominence in cross-cultural research by pro-
viding a method for analyzing similarities and differences in the composition of measures
that are compared cross-culturally. Thus, one ready benefit of invariance testing is that it
may be used to assess a desirable statistical property of measures prior to their use in
 Lucas et al. / MARITAL SATISFACTION INVARIANCE 113

research (i.e., to ensure that cross-cultural comparisons are aptly based on unequivocally
functioning measures). However, a second benefit of invariance testing is that it may be
used on occasion to answer questions of theoretical substance that are of interest to
researchers, as when noninvariance itself is treated as a form of empirical evidence (e.g.,
Cheung & Rensvold, 1999). In this latter respect, invariance testing may afford opportuni-
ties to examine theoretical viewpoints, such as those posited by evolutionary and cultural
perspectives on marital satisfaction.
In the present research, we used invariance testing to examine the equivalency of marital sat-
isfaction. Specifically, we recruited couples from the United States, Britain, Turkey, and China
to assess the extent to which satisfying marriages were cross-culturally similar versus unique.
In accordance with multidimensional conceptualizations, we examined two possible domains
of marital satisfaction, including romantic love, and also communication and support. In addi-
tion, because we recruited intact couples, our data allowed us to examine not only between-
culture invariance (i.e., the extent to which satisfaction measures for husbands from all four of
our cultures were invariant) but also within-culture invariance (i.e., the extent to which mea-
sures for husbands and wives from the same culture were invariant with one another). Thus, our
study examined two independent facets of marital satisfaction invariance—between-culture
similarities of same-gendered spouses (i.e., cultural invariance) and within-culture similarities
of husbands and wives (i.e., spousal invariance).
In accordance with an evolutionary perspective on marriage, we interpreted culturally
invariant aspects of marital satisfaction as supportive of some adaptively evolved functions
of marriage (i.e., as facilitating procreation and pair bonding). Specifically, and in view of
literature that regards romantic love as an attachment process (Hazan & Shaver, 1987) and
affective bonds as universal (van Ijzendoorn & Sagi, 1999), we expected our measures of
spousal love and support to be cross-culturally invariant (Dion & Dion, 1993; Hazan &
Zeifman, 1994). In addition, because of their shared cultural heritage, we expected that
invariance would be stronger between husbands and wives from the same culture than for
same-sex spouses from different cultures.

METHOD

PARTICIPANTS

We recruited more than 2,000 married couples from the United States, Britain, China,
and Turkey (see Table 1). All couples were recruited from predominantly urban areas in
each of their respective countries. As a method of convenience sampling, this strategy was
appropriate to the extent that urban areas encompass a large proportion of married indi-
viduals living within each of our four selected nations. However, our samples were non-
random and therefore do not represent the entirety of marriage within each culture or the
entirety of culture within each nation. A total of 322 U.S. and 350 Turkish couples were
recruited using modified snowball sampling (Bailey, 1987). We recruited 1,031 British
couples through various techniques that included placing an advertisement in a women’s
magazine, hiring a market research company, and enlisting college students to carry out
convenience sampling. Finally, we recruited 232 Chinese couples by giving their children
a questionnaire at school to deliver home to their parents. In all four samples, husbands
were slightly older than their wives, and couples had been married for an average of at
least 11 years. However, there were some notable differences across our samples as well.
114 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

TABLE 1
Sample Demographics

United States Britain China Turkey

Sample Size (N = 322) (N = 1,031) (N = 232) (N = 350)

Husband age in years (SD) 42.35 (10.81)a 38.47 (12.44)b 39.84 (7.60)b 38.33 (10.09)b
Wife age in years (SD) 39.94 (10.31)a 36.14 (12.01)b 38.05 (6.97)c 34.19 (9.34)d
Length of marriage in years (SD) 15.32 (11.26)a 13.17 (10.43)b,d 13.94 (7.76)a,b,c 11.76 (9.46)d

NOTE: Sample size denotes number of couples. Values with noncommon superscripts indicate significant pair-
wise differences (Tukey Honestly Significant Difference, p < .05).

For example, the mean age of wives was different for all samples, with Turkish wives
being the youngest and American wives being the oldest.

MATERIALS

All participants completed the Marriage and Relationships Questionnaire (MARQ) devel-
oped by Russell and Wells (1986). The original version of the MARQ comprises 235 multi-
ple choice and true/false items and was developed using British couples in the late 1980s (see
Russell & Wells, 1993). The survey was originally designed to give a broad view of respon-
dents’ feelings both about themselves and about their relationship with their spouse. Parallel
versions of the MARQ exist for administration to both husbands and wives. The measure is
decidedly subjective in its attempt to capture husband and wife perceptions of themselves
and each other. However, the MARQ is objective in the sense that couples may be formally
compared to one another, and to other couples, on numerous aspects of marital satisfaction.
Although it is less well known, the MARQ is similar to many other measures of marital sat-
isfaction to the extent that it has been used to benchmark associations between spousal per-
ceptions and relationship satisfaction (e.g., Lucas et al., 2004). However, the MARQ is
distinguished by the extent to which it defines marital satisfaction multidimensionally.
Specifically, more than a dozen scales assessing various aspects of marital relationships are
formally and informally derived from the MARQ. The MARQ was advantageous in the pre-
sent study because of this highly multidimensional operationalization.
We selected two subscales of the MARQ for the present study (see the appendix). The
nine-item Love scale was designed to measure emotional or romantic attachment to one’s
spouse (e.g., “Do you find your spouse attractive?”). The nine-item Partnership scale
assessed perceived communication and support of one’s spouse (e.g., “Is there enough give
and take in your relationship?”). All items were answered using a 5-point Likert-type scale
with individually appropriate labels. Cronbach’s alpha values are reported Table 2 for hus-
bands and wives in each of our four cultures.

STATISTICAL ANALYSES

Although more stringent criteria may be used (i.e., Little, 2000), measures are most
often considered invariant across groups when they demonstrate structural equivalency.
Configural equivalency, a less restrictive form of invariance, is defined as a uniform factor
structure across groups and comprises the first requirement of structural equivalency
(Cheung & Rensvold, 1999; Vandenberg & Lance, 2000). Configural invariance exists
when parallel items load significantly onto the same constructs across all assessed groups.
 Lucas et al. / MARITAL SATISFACTION INVARIANCE 115

TABLE 2
MARQ Scale Internal Consistencies

Scale Items Group Alpha

Love 9 American husbands .91

American wives .91
British husbands .89
British wives .91
Chinese husbands .87
Chinese wives .86
Turkish husbands .87
Turkish wives .89
Partnership 9 American husbands .90
American wives .91
British husbands .87
British wives .90
Chinese husbands .88
Chinese wives .88
Turkish husbands .86
Turkish wives .86

NOTE: MARQ = Marriage and Relationships Questionnaire.

Usually, configural equivalency determines a baseline model of invariance by providing a

platform for next assessing metric invariance, the second requirement of structural equiv-
alency (Vandenberg & Lance, 2000). Metric invariance is generally equated with strong
factorial invariance and is more restrictive than configural invariance in requiring equiva-
lent item loadings (i.e., factor scores for a given item are equivalent across groups).
Historically, researchers have suggested that the means of the latent variables can be
compared across groups once metric invariance has been established (Cheung & Rensvold,
1999). However, more recent recommendations (e.g., Byrne, 2001) have suggested that
intercept invariance may also be required for structural equivalency. We therefore
employed a third requirement of structural equivalency by assessing not only configural
and metric invariance but also the equivalency of item intercepts. In cases where all three
requirements of structural equivalency were met, we examined one final form of invari-
ance by examining the equivalency of latent variable means. Thus, although our primary
analyses related to the equivalency of structure in our measures of marital satisfaction, we
also examined equivalency of level in cases where structure was invariant.
We examined the invariance of each of our four marital satisfaction measures using
covariance structure analyses. All analyses were performed using LISREL 8.30 (Jöreskog
& Sörbom, 1999) and maximum likelihood estimation. In all instances, scale restriction
was imposed by constraining the latent variable variances to one. Because of our large
sample sizes, we expected that the normal weighted theory chi-square statistic would be
significant (Marsh, Balla, & McDonald, 1988). We therefore examined several additional
fit indices, including the nonnormed fit index (NNFI; Bentler & Bonett, 1980), the com-
parative fit index (CFI; Bentler, 1990; Satorra & Bentler, 1994), and the root mean square
error of approximation (RMSEA; Browne & Cudeck, 1993). Values of .90 or higher are
thought to indicate good overall fit for the NNFI and the CFI, whereas values up to .08 are
generally considered acceptable for the RMSEA.
Invariance testing requires an assessment of changes in overall model fit that occur
when restricted models impose additional equivalency constraints on items. Generally,
116 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

a chi-square difference test is used, and invariance is established when the imposed equiv-
alencies of metric, intercept, and slope invariance do not significantly reduce the fit of a
previous model (Rensvold & Cheung, 1998; Steiger, Shapiro, & Browne, 1985). However,
because chi-square tests are susceptible to sample size and degrees of freedom, we con-
cluded invariance when changes in the CFI of .01 or less were observed in a more strin-
gently constrained model (Cheung & Rensvold, 1999), whereas additional fit indices
continued to suggest a generally good overall fit (Little, 1997).
We assessed equivalency using two sets of covariance structure equality tests. In the first
set, we assessed invariance between husbands and wives within each of the four cultures (i.e.,
spousal invariance). Similar to time-series covariance analyses, we allowed the residual terms
of husbands and wives to correlate for parallel items. In the second set, we assessed invariance
across same-gendered spouses from all four cultures (i.e., cultural invariance). For both
spousal and cultural invariance, we concluded configural invariance if an initially uncon-
strained model fit well. We then ran additional models in which item loadings, intercepts, and
means were incrementally constrained to be equal across groups (e.g., corresponding item
loadings equal for wives across all four cultures). If overall fit was retained in a more con-
strained model, we concluded that invariance could be ascribed to the prior model.

RESULTS

Our primary hypothesis was that invariance would be established for both love and part-
nership measures. In addition, because of their shared culture, we expected stronger
spousal invariance (i.e., within culture) than cultural invariance (i.e., husbands from all
four cultures) for both of these aspects of marital satisfaction. The results of both sets of
analyses are presented in Tables 3 and 4, respectively. For each measure of marital satis-
faction, the most stringent level of invariance that was obtained is highlighted. In all
instances, tests of invariance concluded once a constrained model demonstrated insuffi-
cient fit with a prior and less constrained model.
Spousal invariance was generally attained in couples from all four cultures. Foremost,
structural invariance was established for scales measuring both love and partnership. In addi-
tion, an equivalent level for both love and partnership was obtained within culture, indicat-
ing that husbands and wives were similar in the amount of both types of satisfaction that they
displayed for one another. As expected, cultural invariance was generally more difficult to
obtain than spousal invariance. At least some forms of structural invariance were obtained for
love and partnership when these measures were assessed cross-culturally. However, and in
all instances, although metric invariance was obtained for both husbands and wives on these
measures, the additional constraint of intercept invariance significantly reduced model fit.

DISCUSSION

Overall, Love and Partnership scales were invariant across couples from all four of our
cultures, thus supporting the possibility that they are an adaptively evolved component of
marital satisfaction. However, cultural invariance of these measures was less robust, indi-
cating that these aspects of marital satisfaction might also be defined uniquely by particu-
lar norms, values, and expectations about marriage. Taken together, these results support
 Lucas et al. / MARITAL SATISFACTION INVARIANCE 117

TABLE 3
Spousal Invariance of Marital Satisfaction in Four Cultures

Scale and Model df χ2 RMSEA NNFI CFI ∆df ∆χ2 ∆CFI

American (N = 322)
Love
Configural 125 444.24*** .088 .90 .92 — — —
Metric 134 454.21*** .085 .91 .92 9 9.81 .00
Intercept 142 477.44*** .084 .91 .92 8 23.23** .00
Mean 143 477.11*** .084 .91 .92 1 0.33 .00
Partnership
Configural 125 189.01*** .039 .97 .98 — — —
Metric 134 195.70*** .037 .98 .98 9 6.69 .00
Intercept 142 215.52*** .040 .97 .97 8 19.82* .01
Mean 143 215.53*** .039 .97 .97 1 0.01 .01
British (N = 1,031)
Love
Configural 125 740.88*** .069 .93 .94 — — —
Metric 134 790.56*** .069 .93 .94 9 49.68*** .00
Intercept 142 979.38*** .071 .92 .93 8 188.82*** .01
Mean 143 1,011.97*** .077 .91 .93 1 32.59*** .00
Partnership
Configural 125 563.56*** .058 .94 .95 — — —
Metric 134 629.33*** .060 .93 .94 9 65.77*** .01
Intercept 142 815.85*** .068 .92 .93 8 186.52*** .01
Mean 143 821.73*** .068 .92 .93 1 5.88* .00
Chinese (N = 232)
Love
Configural 125 246.55*** .065 .91 .92 — — —
Metric 134 253.59*** .062 .91 .92 9 7.04 .00
Intercept 142 309.47*** .071 .90 .91 8 56.08*** .01
Mean 143 314.52*** .072 .90 .91 1 5.05* .00
Partnership
Configural 125 213.38*** .055 .94 .95 — — —
Metric 134 231.55*** .056 .94 .95 9 18.17* .00
Intercept 142 253.56*** .058 .93 .94 8 22.01** .01
Mean 143 256.72*** .059 .93 .94 1 3.16 .00
Turkish (N = 350)
Love
Configural 125 293.07*** .062 .93 .94 — — —
Metric 134 307.39*** .061 .93 .94 9 14.32 .00
Intercept 142 340.42*** .063 .93 .93 8 33.03*** .01
Mean 143 344.02*** .063 .93 .93 1 3.60 .00
Partnership
Configural 125 220.15*** .047 .95 .96 — — —
Metric 134 247.61*** .049 .95 .95 9 27.46** .01
Intercept 142 261.65*** .049 .95 .95 8 14.04 .00
Mean 143 280.35*** .052 .94 .94 1 18.70*** .01

NOTE: RMSEA = root mean square error of approximation; NNFI = nonnormed fit index; CFI = comparative
fit index. Bold lines indicate highest level of invariance attained for each analysis.
*p < .05. **p < .01. ***p < .001.

a multidimensional view of marital satisfaction in suggesting that both evolutionarily

invariant and culturally unique criteria may be reflected in marital satisfaction ratings of
husbands and wives.
118 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

TABLE 4
Cultural Invariance of Marital Satisfaction for Husbands and Wives

Scale and Model df χ2 RMSEA NNFI CFI ∆df ∆χ2 ∆CFI

Husbands (N = 1,945)
Love
Configural 108 619.35*** .093 .93 .95 — — —
Metric 135 706.80*** .088 .94 .94 27 87.45*** .01
Intercept 159 1,240.37*** .120 .89 .88 24 533.57*** .06
Partnership
Configural 108 421.92*** .073 .95 .96 — — —
Metric 135 560.72*** .076 .93 .95 27 138.82*** .01
Intercept 159 1,044.95*** .110 .89 .88 24 484.23*** .07
Wives (N = 1,945)
Love
Configural 108 666.86*** .097 .93 .95 — — —
Metric 135 759.41*** .091 .93 .94 27 92.55*** .01
Intercept 159 1,162.73*** .110 .90 .89 24 403.32*** .05
Partnership
Configural 108 513.06*** .082 .94 .96 — — —
Metric 135 631.46*** .082 .93 .95 27 118.40*** .01
Intercept 159 1,302.06*** .120 .88 .86 24 670.60*** .09

NOTE: RMSEA = root mean square error of approximation; NNFI = nonnormed fit index; CFI = comparative
fit index. Bold lines indicate highest level of invariance attained for each analysis.
***p < .001.

Although romantic love is sometimes seen as characteristic of marriage in Western cul-

tures, evolutionary researchers have historically emphasized its ubiquitous role in mar-
riage. For example, Jankowiak and Fischer (1992) have suggested that there exists a
universal amorous component to marriage. Others have noted that even within traditional
cultures in which there are arranged marriages, most couples eventually fall in love (for
reviews, see Lucas et al., 2004; Weisfeld & Weisfeld, 2002). Although there are many plau-
sible interpretations, one possibility is that romantic love may be an adaptively evolved
component of marital satisfaction to the extent that it facilitates both procreation and care-
giving. The results of this study lend support to this possibility in that spouses were simi-
lar to one another in romantic love across all four of our cultures.
In addition to romantic love, there are numerous evolutionary arguments that support
ubiquitous partnership arrangements in marriage. Most notably, communication and sup-
port between husbands and wives may enhance caregiving. Couples who are unable to
coordinate complex support routines that are demanded by a child, or who are generally
unsupportive of one another, may perform less successfully as parents. Accordingly, selec-
tion pressures should favor at least some universally adaptive forms of communication
between husbands and wives for the purpose of supporting children. Thus, in addition to
romantic love, marriage may require at least some unequivocal partnership behaviors that
are generally reflected in couples’ sense of satisfaction with one another.
Although evolutionary explanations of marital satisfaction are at least plausible according to
spousal invariance, a lack of full cultural invariance in the present study also suggests that mar-
ital satisfaction criteria may be determined by culture. In this vein, theory and research support
that husbands and wives from different cultures may define their satisfaction with one another
in unique ways. For example, religion is thought to play an important role in spouses’ satisfac-
tion with one another in many cultures. In China, however, there is a tenuous connection
 Lucas et al. / MARITAL SATISFACTION INVARIANCE 119

between religion and marriage, as religion has been discouraged politically for most of the
past century. Marital satisfaction may vary in other less formal ways as well. For example,
Campos, Keltner, Beck, Gonzaga, and John (2007) recently suggested that teasing, a social
interaction that benefits relational bonds at the expense of the self, may be viewed more
positively by Eastern than Western cultures. Thus, playful teasing within the context of
marriage may be viewed differently by couples from different cultures. Perhaps most
importantly, our results suggest that romantic love and partnership, which compose two
seemingly universal qualities of a satisfying marriage, may also reflect cultural criteria.

LIMITATIONS

There are several methodological limitations of our research. Foremost, our results are
limited by the nature of our samples. Samples from all four cultures were nonrepresenta-
tive and predominantly urban, and thus, they do not reflect the entirety of each culture. In
addition, although they were selected to provide diversity in geography, economic devel-
opment, religion, and political system, our samples do not reflect the entirety of culture
across the globe. It is entirely possible that the invariance of marital satisfaction would pre-
sent itself differently not only across a broader range of cultures but also if more repre-
sentative samples were available from within each culture. In addition, our results are
characterized by uneven sample sizes. Although our use of relative fit indices minimizes
any potential influence, this characteristic nevertheless confounds the fit indices that we
obtained for each test of invariance.
A second set of limitations concerns the use of marital satisfaction indices that were
developed in a Western culture. Undoubtedly, our assessments of marital satisfaction were
operationalized using at least some culturally specific perceptions and benchmarks, and this
could differentially affect the fit of these measures in non-Western cultures. In addition, mar-
ital satisfaction may be defined beyond considering only romantic love and partnership.
Thus, researchers might consider replicating our results using alternative and additional mea-
sures of marital satisfaction. Such an undertaking should additionally consider better known
and more commonly used marital satisfaction measures. However, as noted previously, we
believe that the use of the MARQ in the present study was advantageous at least to the extent
that it defines marital satisfaction in a highly multidimensional manner, and this afforded
both evolutionary and cultural considerations of spousal satisfaction.
A final set of limitations concerns our interpretation of invariance results. Specifically,
we interpreted invariance characteristics as strictly supportive of either evolutionary or cul-
turally defined criteria of marital satisfaction. However, these perspectives are not mutu-
ally exclusive. The existence of cross-culturally diverse marital values certainly does not
preclude the influence of evolutionary forces in this domain, nor does the homogeneity of
satisfaction preclude the importance of culture. In addition, tests of invariance do not offer
unequivocal support for either cultural or evolutionary interpretations of marriage. For
example, although spousal invariance of romantic love does not rule out the possibility of
an evolutionary interpretation, there are many other equally viable explanations that also
remain. Similarly, noninvariance could merely indicate cultural differences in ways of
answering survey items, such as unique use of scale midpoints (e.g., Cheung & Rensvold,
2000). Invariance testing is also limited in that it requires an imposed-epic approach to
defining marital satisfaction, which may promote the potential for cross-cultural standard-
ization at the expense of measurement misspecification.
120 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

CONCLUSION

The present study addressed the cross-cultural measurement invariance of two indices of
marital satisfaction. Results generally supported the possibility of universal multidimen-
sional criteria in marital satisfaction. However, results also suggested unique cultural defini-
tions of marital satisfaction. We believe the current study provides an important step toward
establishing both universality and uniqueness in indices of marital satisfaction. In addition,
we suggest the possibility that measures of marital satisfaction may reflect themes that are
emphasized by cultural and evolutionary perspectives on marriage. Although there are limi-
tations, future studies in this area may continue to apply invariance analyses to examine
whether they reflect the components of marital satisfaction proposed on the basis of cultural
and evolutionary perspectives.

APPENDIX

Scale Item

Love Do you enjoy your spouse’s company?

Are you happy?
Do you find your spouse attractive?
Do you enjoy doing things together?
Do you enjoy cuddling your spouse?
Do you respect your spouse?
Are you proud of your spouse?
Does your marriage have a romantic side?
How much do you love your spouse?
Partnership Do you find your spouse easy to get along with?
If you are unhappy, can you discuss it with your spouse?
Is your spouse kind to you?
Does your spouse sympathize when you are under pressure?
Does your spouse know what you really think and feel?
Is there enough give and take in your marriage?
Does your spouse understand you?
Does your spouse support you in what you are trying to do?
Does your spouse respect you?

NOTE: All items answered using an appropriately labeled 5-point Likert-type scale.

REFERENCES

Amato, P. R. (1993). Children’s adjustment to divorce: Theories, hypotheses, and empirical support. Journal of
Marriage and Family, 55, 23-38.
Bailey, K. D. (1987). Methods of social research. New York: Free Press.
Bentler, P. M. (1990). Comparative fit indices in structural models. Psychological Bulletin, 107, 238-246.
Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness of fit on the analysis of covariance struc-
tures. Psychological Bulletin, 88, 588-606.
Bernard, J. (1972). The future of marriage. New York: World Books.
Berscheid, E. (1995). Help wanted: A grand theorist of interpersonal relationships, sociologist or anthropologist
preferred. Journal of Social and Personal Relationships, 12, 529-533.
Betzig, L. (1989). Causes of conjugal dissolution: A cross-cultural study. Current Anthropology, 30, 654-676.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen & J. S. Long
(Eds.), Testing structural equation models (pp. 136-162). Newbury Park, CA: Sage.
 Lucas et al. / MARITAL SATISFACTION INVARIANCE 121

Buss, D. M. (1988). Love acts: The evolutionary biology of love. In R. J. Sternberg & M. L. Barnes (Eds.), The
psychology of love (pp. 100-118). New Haven, CT: Yale University Press.
Buss, D. M. (1989). Sex differences in human mate preferences: Evolutionary hypotheses tested in 37 cultures.
Behavioral & Brain Sciences, 12(1), 1-49.
Buss, D. M., Abbott, M., Angleitner, A., Asherian, A., Biaggio, A., Blanco-Villasenor, A., et al. (1990).
International preferences in selecting mates: A study of 37 cultures. Journal of Cross-Cultural Psychology,
21(1), 5-47.
Byrne, B. M. (2001). Structural equation modeling with AMOS, EQS, and LISREL: Comparative approaches to
testing for the factorial validity of a measuring instrument. International Journal of Testing, 1, 55-86.
Byrne, B. M., & Campbell, T. L. (1999). Cross-cultural comparisons and the presumption of equivalent mea-
surement and theoretical structure: A look beneath the surface. Journal of Cross-Cultural Psychology, 30,
555-574.
Campos, B., Keltner, D., Beck, J., Gonzaga, G. C., & John, O. P. (2007). Culture and teasing: The relational ben-
efits of reduced desire for positive differentiation. Personality and Social Psychology Bulletin, 33, 3-16.
Cheung, G. W., & Rensvold, R. B. (1999). Testing factorial invariance across groups: A reconceptualization and
proposed new method. Journal of Management, 25(1), 1-27.
Cheung, G. W., & Rensvold, R. B. (2000). Assessing extreme and acquiescence response sets in cross-cultural
research using structural equation modeling. Journal of Cross-Cultural Psychology, 31, 187-212.
Daly, M., & Wilson, M. (1988). Homicide. New York: Aldine de Gruyter.
Dion, K. L., & Dion, K. K. (1993). Individualistic and collectivistic perspectives on gender and the cultural con-
text of love and intimacy. Journal of Social Issues, 49, 53-69.
Ember, M., & Ember, C. R. (1979). Male-female bonding: A cross-species study of mammals and birds.
Behaviour Science Research, 14, 37-56.
Fiske, A., Kitayama, S., Markus, H. R., & Nisbett, R. E. (1998). The cultural matrix of social psychology. In
D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (Vol. 2, 4th ed., pp. 915-
981). New York: McGraw-Hill.
Goode, W. J. (1993). World changes in divorce patterns. New Haven, CT: Yale University Press.
Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality
and Social Psychology, 52, 511-524.
Hazan, C., & Zeifman, D. (1994). Sex and the psychological tether. Advances in Personal Relationships, 5, 151-177.
Higgins, L. T., Zheng, M., Liu, Y., & Hui Sun, C. (2002). Attitudes to marriage and sexual behaviors: A survey
of gender and culture differences in China and United Kingdom. Sex Roles, 46(3/4), 75-89.
I⋅ mamoğlu, E. O., & Yasak, Y. (1997). Dimensions of marital relationships as perceived by Turkish husbands and
wives. Genetic, Social and General Psychology Monographs, 123(2), 211-232.
Insel, T. R. (1997). A neurobiological basis of social attachment. American Journal of Psychiatry, 154, 726-735.
Jankowiak, W. R., & Fischer, E. F. (1992). A cross-cultural perspective on romantic love. Ethnology, 31, 149-155.
Jöreskog, K. G., & Sörbom, D. (1999). LISREL, version 8.30. Chicago: Scientific Software International.
Kleiman, D. G. (1977). Monogamy in mammals. Quarterly Review of Biology, 52, 39-69.
Lalonde, R. N., Hynie, M., Pannu, M., & Tatla, S. (2004). The role of culture in interpersonal relationships: Do
second generation South Asian Canadians want a traditional partner? Journal of Cross-Cultural Psychology,
35, 503-524.
Landis, D., & O’Shea, W. A. (2000). Cross-cultural aspects of passionate love: An individual differences analy-
sis. Journal of Cross-Cultural Psychology, 31, 752-777.
Little, T. D. (1997). Mean and covariance structures (MACS) analyses of cross-cultural data: Practical and theo-
retical issues. Multivariate Behavioral Research, 32(1), 53-76.
Little, T. D. (2000). On the comparability of constructs in cross-cultural research: A critique of Cheung and
Rensvold. Journal of Cross-Cultural Psychology, 31(2), 213-219.
Lucas, T., Alexander, S., Firestone, I. J., & LeBreton, J. M. (2007). Development and initial validation of a pro-
cedural and distributive just world measure. Personality and Individual Differences, 43, 71-82.
Lucas, T., Michalopoulou, G., Falzerano, P., Menon, S., & Cunningham, W. (in press). Healthcare provider cul-
tural competency: Initial validation of a patient report measure. Health Psychology.
Lucas, T., Wendorf, C. A., Imamoglu, E. O., Shen, J., Parkhill, M. R., Weisfeld, C. C., et al. (2004). Marital sat-
isfaction in four cultures as a function of homogamy, male dominance and female attractiveness. Sexualities,
Evolution and Gender, 6, 97-130.
Marsh, H. M., Balla, J. R., & McDonald, R. P. (1988). Goodness-of-fit indexes in confirmatory factor analysis:
The effect of sample size. Psychological Bulletin, 103, 391-410.
Money, J., & Ehrhardt, A. A. (1972). Man & woman, boy & girl. Baltimore: The Johns Hopkins University Press.
122 JOURNAL OF CROSS-CULTURAL PSYCHOLOGY

Rensvold, R. B., & Cheung, G. W. (1998). Testing measurement models for factorial invariance: A systematic
approach. Educational and Psychological Measurement, 58(6), 1017-1034.
Rogers, S. J., & White, L. K. (1998). Satisfaction with parenting: The role of marital happiness, family structure,
and parents’ gender. Journal of Marriage and Family, 60, 293-308.
Russell, R. J. H., & Wells, P. A. (1986). Marriage questionnaire. Unpublished booklet.
Russell, R. J. H., & Wells, P. A. (1993). Marriage and relationship questionnaire: MARQ handbook. Kent, UK:
Hodder and Stoughton.
Satorra, A., & Bentler, P. M. (1994). Corrections to test statistics and standard errors in covariance structure
analysis. In A. VonEye & C. C. Clogg (Eds.), Latent variable analysis: Applications for developmental
research (pp. 399-519). Thousand Oaks, CA: Sage.
Shackelford, T. K., & Buss, D. M. (1997). Marital satisfaction in evolutionary psychological perspective. In R. J.
Sternberg & M. Hojjatt (Eds.), Satisfaction in close relationships (pp. 7-25). New York: Guilford.
Sprecher, S., & Toro-Morn, M. (2002). A study of men and women from different sides of Earth to determine if
men are from Mars and women are from Venus in their beliefs about love and romantic relationships. Sex
Roles, 46(5/6), 131-147.
Steiger, J. H., Shapiro, A., & Browne, M. W. (1985). On the multivariate asymptotic distribution of sequential
chi-square statistics. Psychometrika, 50(3), 253-263.
Story, A., Walsh, C., Quinton, R., & Wynne-Edwards, K. (2000). Hormonal correlates of paternal responsiveness
in new and expectant fathers. Evolution and Human Behavior, 21, 79-95.
Teachman, J. D., & Heckert, A. (1985). The impact of age and children on remarriage: Further evidence. Journal
of Family Issues, 6, 185-203.
van Ijzendoorn, M. H., & Sagi, A. (1999). Cross-cultural patterns of attachment. In J. Cassidy & P. H. Shaver
(Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 713-734). New York:
Guilford.
Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the measurement invariance literature:
Suggestions, practices, and recommendations for organizational research. Organizational Research Methods,
3(1), 4-69.
Wang, G. T. (1994). Social development and family formation in China. Family Perspective, 28, 283-301.
Weisfeld, G. E., & Weisfeld, C. C. (2002). Marriage: An evolutionary perspective. Neuroendocrinology Letters,
23(Suppl. 4), 47-54.
Zavalloni, M. (1975). Social identity and the recoding of reality. International Journal of Psychology, 10, 197-217.

Todd Lucas is an assistant professor in the Department of Family Medicine and Public Health Sciences–
Division of Occupational and Environmental Health at Wayne State University. He recently completed a
postdoctoral fellowship at the Center for Behavioral and Decision Sciences in Medicine (University of
Michigan & Ann Arbor Veterans Affairs Hospital). In addition to cross-cultural measurement issues, his
research interests include social justice and psychosocial aspects of stress and health.

Michele R. Parkhill is a postdoctoral fellow at the University of Washington. She received her PhD in
social psychology from Wayne State University. Her research interests include alcohol’s role in sexual
assault and risky sexual behavior.

Craig A. Wendorf is an associate professor of psychology at the University of Wisconsin–Stevens Point.

He received his PhD in social psychology from Wayne State University. In addition to statistical model-
ing, his research interests include social justice and the psychology of teaching, learning, and assessment.

⋅
E. Olcay I mamoğlu is a professor at the Middle East Technical University (METU), in Ankara, Turkey.
She received her BS in psychology from METU, her MA in social psychology from the University of Iowa,
and her PhD in developmental social psychology from Strathclyde University in Scotland. Her research
and theoretical interests include relationships between self, well-being, attachment–exploration orienta-
tions, value orientations, gender and marital relationships, and the social psychology of older adults
across and within cultures.

Carol C. Weisfeld received her PhD from the University of Chicago. She is a professor at the University
of Detroit Mercy. Her interests include female inhibition in mixed-sex competition.
 Lucas et al. / MARITAL SATISFACTION INVARIANCE 123

Glenn E. Weisfeld received his doctorate from the University of Chicago. He is a professor of psychology at
Wayne State University and president of the International Society for Human Ethology. His research and
theoretical interests include humor, kin recognition through olfaction, dominance, and adolescence.

Jiliang Shen directs the Institute of Developmental Psychology at Beijing Normal University. His main
research interests are in scientific creativity and teacher professional development. He is currently con-
ducting a research project on cross-national comparisons of creativity.