ArticlePDF Available

Willingness to sacrifice in close relationship

Authors:
Article

Willingness to sacrifice in close relationship

Abstract

The authors advance an interdependence analysis of willingness to sacrifice. Support for model predictions was revealed in 6 studies (3 cross-sectional survey studies, 1 simulation experiment, 2 longitudinal studies) that used a novel self-report measure and a behavioral measure of willingness to sacrifice. Willingness to sacrifice was associated with strong commitment, high satisfaction, poor alternatives, and high investments; feelings of commitment largely mediated the associations of these variables with willingness to sacrifice. Moreover, willingness to sacrifice was associated with superior couple functioning, operationalized in terms of level of dyadic adjustment and probability of couple persistence. In predicting adjustment, willingness to sacrifice accounted for significant variance beyond commitment, partially mediating the link between commitment and adjustment; such mediation was not significant for persistence.
Journal of Personality ami Social Psychology
1997,
Vol. 72, No. 6, 1373-1395
Copyright 1997 by the American Psychological Association, Inc.
0022-3514/97/$3.00
Willingness to Sacrifice in Close Relationships
Paul A. M. Van Lange
firee University, Amsterdam
Caryl E. Rusbult
University of North Carolina at Chapel Hill
Stephen M. Drigotas
Southern Methodist University
Ximena B. Arriaga and Betty S. Witcher
University of North Carolina at Chapel Hill
Chante L. Cox
Mississippi State University
The authors advance an interdependence analysis of willingness to sacrifice. Support for model
predictions was revealed in 6 studies (3 cross-sectional survey studies, 1 simulation experiment, 2
longitudinal studies) that used a novel self-report measure and a behavioral measure of willingness
to sacrifice. Willingness to sacrifice was associated with strong commitment, high satisfaction, poor
alternatives, and high investments; feelings of commitment largely mediated the associations of these
variables with willingness to sacrifice. Moreover, willingness to sacrifice was associated with superior
couple functioning, operationalized in terms of level of dyadic adjustment and probability of couple
persistence. In predicting adjustment, willingness to sacrifice accounted for significant variance
beyond commitment, partially mediating me link between commitment and adjustment; such media-
tion was not significant for persistence.
Sometimes it is easy for close partners to coordinate their
behavior in such a manner as to achieve good outcomes—when
two individuals* interests align, achievement of desirable out-
comes is a relatively simple matter. Unfortunately, partners' in-
terests sometimes are at odds—what is good for one partner
is not good for the other. When partners' preferences do not
correspond, one or both individuals may find it necessary or
desirable to sacrifice their needs for their partner's needs. How
do individuals resolve the conflict between that which is best
for them and that which is best for their partner or relationship?
What makes individuals willing to sacrifice, and is this willing-
ness associated with healthy couple functioning?
Editor's Note. John Holmes served as action editor for this article.—NM
Paul A. M. Van Lange, Department of Social Psychology, Free Univer-
sity, Amsterdam, The Netherlands; Caryl E. Rusbult, Ximena B. Arriaga,
and Betty S. Witcher, Department of Psychology, University of North
Carolina at Chapel Hill; Stephen M. Drigotas, Department of Psychol-
ogy, Southern Methodist University; Chante L. Cox, Department of Psy-
chology, Mississippi State University.
Ximena
B.
Arriaga is now at the Center for Organizational and Behav-
ioral Sciences, Claremont Graduate School.
This research was supported by Netherlands Organization for Scien-
tific Research Grant R-57-178, National Institute of Mental Health
Grants BSR-1-R01-MH-45417 and 1-F3-MH-10305-01, and National
Science Foundation Grant BNS-9023817. Preparation of this article was
supported by a Reynolds Research Leave from the University of North
Carolina at Chapel Hill and by the Netherlands Organization for Scien-
tific Research.
Correspondence concerning this article should be addressed to Paul
A. M. Van Lange, Department of Social Psychology, Free University,
Van der Boechorststraat 1, 1081 BT Amsterdam, The Netherlands. Elec-
tronic mail may be sent via Internet to pam.van.lange@psy.vu.nl.
The answers to such questions are indirect at best. Some
researchers have addressed the phenomenon of sacrifice in the
context of nonclose relationships, for example, studies of orga-
nizational behavior, experimental games, or justice phenomena
(e.g., Baefsky & Berger, 1974; MacCrimmon & Messick, 1976;
Meyer, Allen, & Gellatly, 1990; Schwartz, 1975). Also, several
theoretical accounts of behavior in close relationships suggest
that sacrifice may be a determinant of healthy couple function-
ing (e.g., Berscheid, 1985; Clark & Mills, 1979; Holmes &
Boon, 1990). However, remarkably little empirical attention has
been directed toward understanding the determinants and
consequences of willingness to sacrifice in ongoing, close
relationships.
In this article, we adopt an interdependence analysis of willing-
ness to sacrifice (Kelley & Thibaut, 1978), proposing that situa-
tions of conflicting interest are potentially disruptive to the health
and vitality of a relationship. To deal with such situations requires
some inclination toward prorelationship transformation of moti-
vation, yielding increased willingness to sacrifice. Through the
model advanced in this article, we identify several features of
interdependence that should increase willingness to sacrifice and
yield enhanced couple functioning. On the basis of this reasoning,
several hypotheses regarding the determinants and consequences
of willingness to sacrifice are advanced and tested.
An Interdependence Analysis of Willingness to Sacrifice
Dictionary definitions of sacrifice include the following:
"giving up one thing for another," "surrender to gain some
other object,*' "devote with loss/* or "foregoing something
valued for the sake of something having a more pressing claim"
(Webster's New School & Office Dictionary, 1960, p. 640; Gur-
alnik, 1970, p. 1252). In the context of ongoing, close relation-
1373
1374
VAN LANGE ET AL.
ships,
willingness to sacrifice is defined as the propensity to
forego immediate self-interest to promote the well-being of a
partner or relationship. Sacrifice may entail the forfeiting of
behaviors that might otherwise be desirable (i.e., passive sacri-
fice
),
enacting of behaviors that might otherwise be undesirable
(i.e.,
active sacrifice), or both. Such experiences may take a
variety of forms, varying from transient, situation-specific, and
often minor sacrifices (e.g., attend a play your partner wants to
see) to more substantial, extended ones (e.g., agree to live in
an undesirable locale for your partner's career).
Under what circumstances is sacrifice called for? What is it
about the structure of interdependence that makes it necessary
to give up one thing for another? Interdependence theory sug-
gests that individuals are forced to choose between self-interest
and sacrifice in situations involving noncorrespondence (Kel-
ley & Thibaut, 1978). Correspondence of outcomes is defined
as the degree to which partners' preferences for various activi-
ties (i.e., combinations of one's own behavior and partner's
behavior) correspond or conflict. Noncorrespondent situations
resemble social dilemmas: Collective interest is better served if
partners engage in prorelationship behavior than if they behave
selfishly; at the same time, the individual's immediate, personal
interests would be better served by acting selfishly (cf. Dawes,
1980;
Messick & Brewer, 1983). Noncorrespondence is quite
common in ongoing relationships (cf. Gottman, 1994; Holmes,
1989;
Kelley et al., 1983; Surra & Longstreth, 1990) and is
potentially disruptive for a variety of reasons—partners may
expend energy in attempting to resolve conflicts of interest,
become annoyed at one another's different preferences, feel be-
trayed in situations where one or both partners ignore the other's
welfare in the pursuit of self-interest, or dissolve the relation-
ship.
Thus, noncorrespondence is an important challenge to the
well-being of a relationship. Sacrifice represents one important
mechanism by which individuals can solve dilemmas involving
noncorrespondent outcomes.
Given that acts of sacrifice may involve negative conse-
quences, such as opportunity costs or effort expenditure, it is
important to distinguish between the concepts of sacrifice and
cost. Whereas sacrifice refers to behavior (i.e., behavior that
departs from direct self-interest), cost refers to psychological
experience (i.e., feeling that an event is unpleasant). A variety
of events may be experienced as costly, but many such events
do not involve acts of sacrifice. For example, the experience of
cost may have its origins in a partner's behavior (e.g., he insults
her haircut), broader traits or attributes (e.g., he is a careless
driver),
or experiences external to the relationship (e.g., mutual
friends do not get along). In contrast, sacrifice refers to the
individual's own forfeiting of self-interest. Moreover, acts of
sacrifice may or may not be consciously experienced as costly
and may or may not be described as distasteful. Indeed, to
forfeit self-interest "without grudge" is a very generous form
of sacrifice. Thus, whereas the concept of cost is inherently
linked to the experience of dissatisfaction, acts of sacrifice are
intended to further positive goals—to promote the well-being
of a partner or relationship. It is interesting that, in measuring
costs,
researchers have tended to use broadly defined variables
and measures that may have tapped costs, sacrifice, or both.
This may explain why measures of costs have been shown to
exhibit negative, positive, or no links with relationship function-
ing (e.g., Hays, 1985; Rusbult, 1983).
Clearly, acts of sacrifice involve departures from the individu-
al's underlying, a priori, self-interested preferences, termed
given preferences. The means by which individuals depart from
given preferences is termed transformation of motivation—a.
process that may lead individuals to relinquish immediate
self-
interest and act based on broader goals (Kelley & Thibaut,
1978).
That is, individuals may take broader considerations into
account, such as a partner's interests or one's desire to maintain
a long-term relationship. (Of course, such considerations are
not always positive; individuals may sometimes seek to harm
their partners or relationships.) Transformation of motivation
yields a reconceptualized, effective constellation of preferences,
which are assumed to directly guide overt behavior.
Determinants of Willingness to Sacrifice:
Role of Commitment Level
We assume that specific, noncorrespondent situations initially
are experienced as unique problems to be solved. The transfor-
mation process through which individuals resolve novel dilem-
mas may involve conscious thought; for example, individuals
may consider the available options, interpret the situation in
light of surrounding circumstances, review, feelings for a partner
and goals for a relationship, and decide whether to behave
selfishly or to sacrifice. Alternatively, the transformation process
may involve little conscious thought; for example, individuals
may impulsively act based on the prevailing emotional tone
accompanying an interaction. In either event, the unique problem
has been dealt with, and experience has been acquired.
Over time in a relationship, specific problems of noncorre-
spondence are encountered with regularity, and a relatively sta-
ble orientation to such situations may develop. Through the
process of adaptation to repeatedly encountered patterns of inter-
dependence, individuals develop habitual tendencies to react to
specific patterns in specific ways, such that the transformation
process occurs quite rapidly, with little or no conscious thought.
Of course, at critical choice points, individuals may continue
to engage in transformation-relevant information seeking and
rational decision making; but, just as often, transformation of
motivation may be guided by habitual tendencies. In some rela-
tionships, partners may routinely engage in prorelationship
transformation; whereas, in other relationships, partners may
typically behave selfishly. According to Holmes (1981), such
stable transformational tendencies are guided by relatively en-
during, relationship-specific motives.
We suggest that commitment is a central motive in ongoing
relationships and propose that feelings of commitment promote
prorelationship transformation and willingness to sacrifice.
Commitment represents the degree to which an individual expe-
riences long-term orientation toward a relationship, including
intent to persist through both "good and lean times," feelings
of psychological attachment, and implicit recognition that one
' 'needs'' a relationship. From interdependence theory (Kelley &
Thibaut, 1978; Thibaut & Kelley, 1959), commitment is argued
to emerge out of the specific circumstances of interdependence
that characterize a given relationship. In particular, commitment
is argued to develop as a consequence of increasing dependence
as a result of an (a) increase in satisfaction level (i.e., the
relationship gratifies important needs, such as the needs for
intimacy or security) and (b) decline in quality of alternatives
WILLINGNESS TO SACRIFICE
1375
(i.e.,
specific alternative partners, field of eligibles, or nonin-
volvement are relatively unattractive). The investment model
extends interdependence assertions, hypothesizing that commit-
ment is further strengthened as a consequence of (c) increasing
investment size (i.e., resources, such as personal identity, effort,
or material possessions, are linked to a relationship; Rusbult,
1980,
1983). Commitment is an emergent property of depen-
dence, reflecting more than the sum of the components out of
which it arises (e.g., dependence per se does not necessarily
induce psychological attachment). Consistent with these claims,
numerous studies demonstrate that satisfaction level, quality of
alternatives, and investment size contribute unique variance to
predict commitment; moreover, commitment is the strongest pre-
dictor of persistence in a relationship, accounting for significant
variance above and beyond satisfaction, alternatives, and invest-
ments (e.g., Drigotas & Rusbult, 1992; Fclmlee, Sprecher, &
Bassin, 1990; Lund, 1985; RusbuLt, 1983; Rusbult, Johnson, &
Morrow, 1986; Simpson, 1987).
Why should strong commitment promote willingness to sacri-
fice? We suggest four lines of reasoning in support of this pre-
diction, identifying several interrelated mechanisms that
might—individually or in combination—account for such an
association. (Indeed, we suspect that, in the context of ongoing
relationships, several mechanisms are unlikely to occur indepen-
dent of one another; that is, in ongoing relationships, two or
more of the mechanisms may operate in concert.)
First, committed individuals (i.e., committed to a relation-
ship) are dependent on their partners and need their relationships
(e.g., John's relationship provides him with desirable outcomes,
he has invested a great deal in his relationship, and his alterna-
tives are not good). Because committed individuals need their
relationships, they should be more willing to sacrifice direct
self-interest to sustain their relationships—the more you have
to lose, the more you are willing to give up; to hold on to what
you have.
Second, commitment involves long-term orientation; commit-
ted individuals look beyond the here and now, considering not
only the current noncorrespondence problem but also anticipat-
ing future noncorrespondent situations. In a short-term involve-
ment, individuals may achieve relatively good outcomes by be-
having in accordance with immediate self-interest. In contrast,
in long-term involvements, it behooves partners to develop pat-
terns of reciprocal cooperation (e.g., John's long-term well-
being is enhanced if he sacrifices today so that Mary will sacri-
fice next week). Thus, acts of sacrifice may represent a con-
scious or unconscious means to maximize long-term self-interest
(cf. Axelrod, 1984). Moreover, long-term orientation may serve
to ''smooth the bumps," in that any undesirable outcomes that
result from acts of sacrifice are aggregated (a) over a longer
time perspective and (b) in light of the partner's reciprocal
sacrifice (cf. Kelley, 1983).
Third, given that commitment involves psychological attach-
ment to a partner, the person and his or her partner may become
linked, to the extent that a departure from self-interest that
benefits one's partner may not be experienced as a departure
from self-interest (cf. Aron & Aron, 1986). For example, in
John's mind, something good for Mary may be inseparable from
something good for
himself.
Thus, committed individuals may
engage in behaviors that they would otherwise find undesirable
because it makes their partner feel good and accordingly makes
them feel good.
Fourth, strong commitment may bring with it a collectivistic,
communal orientation, including tendencies to respond to a part-
ner's needs in a rather unconditional manner. Indeed, commit-
ment is associated with tendencies to describe one's relationship
in a relatively collectivistic, pluralistic manner (e.g., we, us, our
rather than /. me, mine; Agnew, Van Lange, Rusbult, & Langs-
ton, 1996). In a highly committed, communally oriented rela-
tionship, partners may be willing to exert effort or endure cost,
without counting what they receive in return or calculating
whether their acts will be reciprocated (cf. Clark & Mills,
1979).
In noncorrespondent situations, communal orientation
could enhance the probability of sacrifice rather than the pursuit
of self-interest (John is likely to sacrifice his direct self-interest
simply because Mary needs him to).
Thus,
several lines of reasoning support the hypothesis that
strong commitment may be associated with prorelationship
transformation of motivation and enhanced willingness to sacri-
fice. Indeed, existing research provides indirect support for this
supposition, in that commitment is shown to be associated with
prorelationship behaviors, such as derogation of tempting alter-
natives (e.g., Johnson & Rusbult, 1989; Simpson, Gangestad, &
Lerma, 1990) and willingness to accommodate rather than retal-
iate when a partner behaves poorly (e.g., Rusbult, Bissonnette,
Arriaga, & Cox, in press; Rusbult, Verette, Whitney, Slovik, &
Lipkus, 1991).
Consequences of Willingness to Sacrifice: Links With
Quality of Couple Functioning
As noted earlier, noncorrespondent situations are potentially
disruptive for a variety of
reasons.
Such interdependence dilem-
mas are complex and may arouse negative emotions, such as
anger or insecurity, and partners may respond to persistent non-
correspondence by avoiding or terminating a relationship. More-
over, we have suggested that acts of sacrifice may involve nega-
tive consequences, such as opportunity costs or effort expendi-
ture.
In light of the negative features of dilemmas that call for
sacrificial behavior, it is not implausible that individuals might
experience considerable ambivalence about self-sacrifice and
that high levels of sacrifice might be indicative of poor couple
functioning. Indeed, relationships may be strained by the exis-
tence of numerous or extreme noncorrespondent dilemmas, and
the necessity of frequent or excessive acts of sacrifice may take
their toll on long-term partners.
Nevertheless, our analysis suggests that willingness to sacri-
fice exhibits positive links with healthy couple functioning, oper-
ationally defined as the level of dyadic adjustment and probabil-
ity of a person persisting in a relationship. How and why might
acts of sacrifice bring about enhanced couple functioning? At
least four complementary mechanisms seem plausible—none
of which necessarily assume conscious intent on the part of
either individual.
First, an act of sacrifice typically enhances the probability
that one's partner will reciprocate such acts in future noncorre-
spondent situations. In the context of a generally loving and
committed relationship, few individuals are likely to "take a
free ride," responding to a partner's sacrifice with exploitation.
Indeed, if they were to take a free ride, the partner would proba-
1376 VAN LANGE ET AL.
bly discontinue acts of sacrifice (i.e., defection yields reciprocal
defection). For many forms of noncorrespondence, both individ-
uals achieve better outcomes over the course of an extended
interaction if both behave in a prosocial manner than in accor-
dance with immediate self-interest (cf. Axelrod, 1984; Pruitt &
Kimmel, 1977). Thus, in the context of an ongoing relationship,
for example, John and Mary may in fact achieve the highest
possible benefit by developing patterns of reciprocal sacrifice.
Second, when an individual reacts to a noncorrespondent di-
lemma by sacrificing immediate self-interest, the individual liter-
ally solves the problem on the partner's
behalf.
Specifically, the
individual eliminates one or more undesirable response options
with which the partner was previously confronted (e.g., elimi-
nating the partner's need to sacrifice) and creates a more conge-
nial set of options for the partner (e.g., allowing the partner
to enjoy good outcomes and exhibit loving gratitude for the
individual's gesture of good will; cf. Kelley, 1984). Thus, acts
of sacrifice frequently do more than simply solve the problem
at hand; such behavior may exert more global beneficial effects
on the general circumstances of interdependence within which
the partners operate.
Third, reliable tendencies toward sacrifice may create a gen-
eral "climate" of trust and cooperation in which other prorela-
tionship events become increasingly probable. When partners
develop generalized habits of prorelationship transformation,
they are more likely to seek out and identify patterns of interac-
tion for which little or no sacrifice is called. For example, al-
though problems of noncorrespondence may appear to be insur-
mountable, partners with reliable prorelationship habits may
find it easier to locate congenial solutions, whereby both parties
benefit, such as logrolling or integrative solutions (cf. Peterson,
1983;
Pruitt & Lewis, 1975).
Fourth, sacrifice serves a communicative function, providing
relatively unambiguous evidence of the individual's prorelation-
ship orientation. Such acts have "surplus value," yielding posi-
tive consequences for the partner above and beyond any direct
impact on experienced outcomes. Acts of sacrifice are likely to
enhance one's partner's trust and conviction that the individual
can be relied on to behave in a prorelationship manner (cf.
Holmes, 1989; Holmes & Rempel, 1989). In turn, enhanced
trust is likely to increase one's partner's willingness to commit
to the relationship and to engage in reciprocal prorelationship
acts.
In fact, through dissonance reduction or self-perception
processes, for example, John's acts of sacrifice may serve to
enhance his feelings of commitment (e.g., "I deviated from
what would seem to be in my self-interest, thereby benefitting
my partner, so I must be committed to my relationship"; Aron-
son & Mills, 1959; D. J. Bern, 1972).
Thus,
several lines of reasoning support the hypothesis that
willingness to sacrifice may be associated with enhanced couple
functioning. Research provides indirect support for this claim,
in that levels of dyadic adjustment and probability of persistence
are shown to be associated with prorelationship maintenance
behaviors, such as willingness to accommodate rather than retal-
iate when a partner behaves poorly (e.g., Gottman, 1979; Jacob-
son & Margolin, 1979; Rusbult et al., in press).
A Model of the Determinants and Consequences of
Willingness to Sacrifice
Given that our work represents an initial empirical examina-
tion of sacrifice in the context of close relationships, we have
provided a rather detailed description of the process by which
sacrifice comes about and the process by which such acts affect
quality of couple functioning. At the same time, that this is an
initial empirical examination means that it is important to keep
our preliminary explorations relatively focused. In this research,
we examine two broad issues regarding sacrifice by testing
hypotheses concerning (a) the determinants of and (b) the con-
sequences of willingness to sacrifice.
We examined willingness to sacrifice rather than level or
frequency of actual sacrifice because the prorelationship motiva-
tion embodied in willingness to sacrifice is of most direct rele-
vance to our interdependence analysis. To understand prorela-
tionship motivation and behavior, a key issue is whether depar-
tures from immediate self-interest are evident when the
necessity for such departures arises. Thus, it is important to
measure sacrifice in such a manner as to "unconfound" prorela-
tionship motivation from degree of couple correspondence. For
example, Mary may be ready and willing to sacrifice when the
necessity arises. But if Mary's relationship is characterized by
exceptionally high correspondence, she may exhibit low levels
of actual sacrifice because the necessity for sacrifice seldom
arises.
In the initial explorations reported in this article, a rela-
tively straightforward means of assessing sacrifice, independent
of degree of correspondence, was to examine willingness to
sacrifice. (In addition, in one of our studies, we used a behav-
ioral measure to examine level of sacrifice in a context in which
degree of noncorrespondence was controlled.)
Relevant to the understanding of the determinants of sacrifice,
the most important hypothesis guiding our work is the claim
that commitment is a central relationship-specific motive that is
associated with enhanced willingness to sacrifice (Hypothesis
1).
Of lesser importance, we examine the well-documented
finding that commitment level is positively associated with satis-
faction level, negatively associated with perceived quality of
alternatives, and positively associated with investment size (Hy-
pothesis la). Also, extending earlier logic regarding the role of
dependence in the promotion of willingness to sacrifice, we also
predict that greater satisfaction, poorer alternatives, and greater
investments are associated with greater willingness to sacrifice
(Hypothesis lb). At the same time, we assume that commitment
is a broad, relationship-specific motive that emerges out of rela-
tively more specific circumstances of interdependence, a motive
that represents more than the sum of
the
components from which
it arises. Thus, we assume that commitment mediates the effects
of other features of interdependence; that is, when commitment
is included in a model, along with satisfaction, alternatives, and
investments, these variables contribute little variance above and
beyond commitment (Hypothesis lc).
Relevant to the understanding of the consequences of sacri-
fice, our most important hypothesis is that willingness to sacri-
fice is associated with enhanced couple functioning, that is,
greater dyadic adjustment and greater probability of persistence
(Hypothesis 2). Of lesser importance, we hypothesize that com-
mitment level exhibits parallel links with couple functioning;
that is, strong commitment is also likely to be associated with
healthy functioning (Hypothesis 2a). Finally, we examine the
role of willingness to sacrifice in the mediation of
the
association
between commitment and quality of functioning. In this respect,
note that sacrifice
is
just one of numerous prorelationship mech-
anisms exhibited by committed individuals by which commit-
WILLINGNESS TO SACRIFICE
1377
ment enhances couple well-being. (Presumably, if we could
identify all of the prorelationship mechanisms that are motivated
by commitment, collectively, these mechanisms would wholly
mediate the link between commitment and couple functioning.)
Moreover, sacrifice may "feed back on" commitment. For ex-
ample, an act of sacrifice may be experienced as an investment
in one's relationship, which in turn may strengthen feelings of
commitment. Accordingly, we hypothesize that commitment as
well as willingness to sacrifice independently contribute to pre-
dict quality of couple functioning, anticipating that willingness
to sacrifice partially mediates the effects of commitment (Hy-
pothesis 2b). Specifically, we predict that (a) commitment ex-
erts direct effects on couple functioning that extend beyond that
which is mediated by willingness to sacrifice (i.e., when the
effects of sacrifice are taken into consideration, commitment
accounts for significant unique variance in couple functioning);
(b) willingness to sacrifice exerts direct effects on couple func-
tioning that extend beyond the variance attributable to commit-
ment (i.e., when the effects of commitment are taken into con-
sideration, willingness to sacrifice accounts for significant
unique variance in couple functioning); and (c) willingness
to sacrifice "explains" a portion of the association between
commitment and couple functioning, significantly mediating this
link (i.e., in comparison to the simple link between commitment
and couple functioning, the commitment-functioning link is
reduced when shared variance with sacrifice is taken into
consideration).
Six studies provide preliminary evidence relevant to the
model outlined above. Studies 1, 2, and 4 are cross-sectional
survey studies, Study 3 is a simulation experiment, and Studies
5 and 6 are longitudinal studies. In studies 1-5, we examine
dating relationships; whereas in Study 6, we examine marital
relationships. All six studies include self-report measures of
willingness to sacrifice; Study 4 also includes a behavioral mea-
sure of sacrifice, providing evidence regarding convergence of
findings across measurement techniques. All six studies include
self-report measures of commitment level and the three pre-
dictors of commitment (i.e., satisfaction, alternatives, and in-
vestments). Studies 3-6 also include measures of quality of
couple functioning, assessing level of dyadic adjustment, persis-
tence in a relationship, or both. Thus, in these studies, we make
use of the concept of converging operations by testing parallel
hypotheses with a variety of methods, participant populations,
and measurement techniques.
Studies 1 and 2
Studies 1 and 2 are cross-sectional survey studies. Both stud-
ies include measures of satisfaction, alternatives, investments,
commitment, and willingness to sacrifice, allowing us to test
hypotheses concerning the predictors of sacrifice. Because both
studies were conducted in the Netherlands, Study 1 was partially
exploratory, designed not only to test model predictions but also
to examine the reliability of (a) our measure of willingness
to sacrifice and (b) Dutch translations of investment model
measures.
Method
Participants. Study 1 participants were 105 individuals (57 women,
47 men, and 1 individual who did not report gender) who were recruited
at a variety of locations (e.g., in the cafeteria and library) on the campus
of the Free University at Amsterdam. Participants were 23.95 years old
on average, and their relationships were about 32 months in duration.
Study 2 participants were 83 individuals (44 women, 39 men) who were
recruited by means of an advertisement placed in a local university
newspaper that invited individuals who were involved in dating relation-
ships to participate in an interpersonal relationships study. Participants
were 22.65 years old on average, and their relationships were about
30 months in duration. Participants were paid 12.50 Dutch guilders
(approximately $7.00 in U.S. currency) for taking part in Study 2.
Procedure. Participants completed questionnaires describing their
current dating relationships, answering questions designed to measure
willingness to sacrifice, commitment level, satisfaction level, quality of
alternatives, and investment size. Three procedural differences distin-
guished Study
1
from Study 2: (a) Study I questionnaires were adminis-
tered individually, whereas Study 2 questionnaires were administered to
groups of 10 to 15 participants; (b) Study 2 participants were paid,
whereas Study 1 participants were not; and (c) given that Dutch scales
to measure model variables had not previously been developed, a few
items developed for Study
1
were modified for use in Study 2 to improve
measure reliability and validity (e.g., minor changes in wording and an
addition of a few items). After completing their questionnaires, partici-
pants were debriefed and thanked for their assistance.
Questionnaires. To measure willingness to sacrifice, we assessed
activities that were relatively central to the individual's well-being by
asking each participant to list, in order, the three "most important activi-
ties in your life, other than your relationship with your partner." Partici-
pants listed life domains, such as parents and siblings, career, education,
religion, friends, or pastimes (e.g., going to the beach and playing soc-
cer).
We pitted personal well-being against relationship well-being, us-
ing the logic of forced-choice methodology: For each activity, the partici-
pant was asked,
'
'Imagine that it was not possible to engage in Activity
1 and maintain your relationship with your partner. To what extent
would you consider ending your relationship with your partner?" (0 =
definitely would not consider ending relationship, 8 = definitely would
consider ending relationship, reverse scored; Study 1 and Study 2 as
= .83 and .82, respectively).'
The measures of commitment and the three investment model vari-
ables were modeled after those used in prior research (Rusbult, 1983;
Rusbult et al., 1991). In Study 1, commitment level was measured
with seven items (e.g., "Do you feel committed to maintaining your
relationship with your partner?"; 0 = not at all committed, 8 = com-
pletely committed; Study I a = .73). An additional item was developed
for Study 2, yielding an eight-item measure with improved reliability
(Study 2 a = .87). In both studies, satisfaction level was measured
with five items (e.g., "All things considered, to what degree do you
feel satisfied with your relationship?"; 0 = not at all satisfied, 8 =
1
The measure of willingness to sacrifice emphasized one's desire to
avoid harming the relationship (i.e., "To what degree would you con-
sider ending your relationship''); three- or four-item versions of this
instrument were used in Studies 1, 2, and 5. Our measure of willingness
to sacrifice emphasizes the sacrificial act per se (i.e., "To what degree
would you consider giving up this activity"; see Appendix); versions
of this instrument were used in Studies 3, 4, and 6 (a three-item version
was used by Van Lange, Agnew, Harinck, & Steemers, in press; the
briefer, three-item version also exhibits good reliability and validity).
Note that our sacrifice instrument does not include a conceptual overlap
with the commitment instrument, which is used to include items tapping
subjective commitment (e.g., "Do you feel committed to maintaining
your relationship with your partner?"), psychological attachment (e.g.,
"Do you feel attached to your relationship with your partner [like
you're 'linked' to your partner, whether or not you're happy with the
relationship]?"), and intent to persist (e.g., "For how much longer do
you want your relationship to last?").
1378
VAN LANGE ET AL.
completely satisfied: Study 1 and Study 2 «s = .90 and .88). In Study
1,
quality of alternatives was measured with four items (e.g., "How
attractive are the people other than your current partner with whom you
could become involved?"; 0
alternatives are not at all appealing, 8
= alternatives are extremely appealing). One item exhibited a low
item-total correlation and was dropped. However, the three-item mea-
sure remained unreliable,