attachment insecurities and interpersonal processes in spanish

Personal Relationships, (2010). Printed in the United States of America.Copyright © 2010 IARR; DOI: 10.1111/j.1475-6811.2010.01325.x

Attachment insecurities and interpersonal processesin Spanish couples: A dyadic approach

FERNANDO MOLERO,a PHILLIP R. SHAVER,b EMILIO FERRER,b

ISABEL CUADRADO,c AND ITZIAR ALONSO-ARBIOLd

aSpanish Open University (UNED), Spain; bUniversity of California, Davis; cUniversityof Almerıa, Spain; dUniversity of the Basque Country, Spain

AbstractSeveral attachment-related phenomena in Spanish couples using dyadic-level analyses were examined. A sample of295 heterosexual couples completed measures of attachment-related anxiety and avoidance, self-esteem, socialself-efficacy, and relationship satisfaction. Results, analyzed from a dyadic perspective using the actor–partnerinterdependence model (APIM), indicate that (a) there are actor but no partner effects of attachment insecurities onintrapersonal variables such as self-esteem and social self-efficacy, (b) there are actor and partner effects of avoidantattachment on relationship satisfaction, and (c) actor anxiety is associated with partner avoidance, and actoravoidance is associated with partner anxiety. Overall, the results reveal the importance of a dyadic perspective oncouple members’ attachment insecurities and their associations with intrapersonal and interpersonal processes andrelationship adjustment. They also show that attachment variables and correlates studied mainly in English-speakingcountries are useful in understanding Spanish couple dynamics.

Attachment theory has stimulated hundredsof studies in the past two decades, show-ing how attachment insecurities of two kinds,anxiety and avoidance, are related to otherpsychological processes, such as affect reg-ulation, and to dyadic processes, such asconflict and forgiveness, in close interper-sonal relationships (see Mikulincer & Shaver,2007, for a review). Nevertheless, many ofthe studies have involved only individualsrather than members of established couples,and most studies have been conducted inEnglish-speaking countries (i.e., the UnitedKingdom, the United States, Canada, and

Fernando Molero, Department of Social and Organi-zational Psychology, Spanish Open University (Uni-versidad Nacional de Educacion a Distancia), Madrid,Spain; Phillip R. Shaver and Emilio Ferrer, Departmentof Psychology, University of California, Davis; IsabelCuadrado, Department of Social Psychology, Universityof Almería, Almería, Spain; Itziar Alonso-Arbiol, Fac-ulty of Psychology, University of the Basque Country,San Sebastian, Spain.

Correspondence should be addressed to FernandoMolero, Departamento de Psicología Social y de lasOrganizaciones, Universidad Nacional de Educacion aDistancia, Madrid, Spain, e-mail: [email protected].

Australia). These are the important limita-tions of research on couple relationships,because, first, couple relationships cannot beunderstood solely in terms of individual-level factors, and second, it cannot simplybe assumed that relationship processes stud-ied in one sociocultural context are universal.Although adult attachment studies conductedin different countries have so far suggestedconsiderable universality, some cross-culturaldifferences have been reported (see Shaver,Mikulincer, Alonso-Arbiol, & Lavy, 2010, fora review and commentary). More germane tothis study are recent findings showing thatSpanish university students (Alonso-Arbiol,Balluerka, Shaver, & Gillath, 2008) andAmerican students of Hispanic ethnicity (Wei,Russell, Mallinckrodt, & Zakalik, 2004) scorehigher than the other groups examined sofar on attachment anxiety. If higher levels ofattachment anxiety are normative in Spanishpopulations, it may reduce the negative effectsof this kind of insecurity on couple dynamics.

Understanding the universalities and cul-tural differences in adult attachment processes

1

2 F. Molero et al.

depends on increasing the number of studiesconducted in non-English-speaking societies.This study addresses this need by examining,from a dyadic perspective, attachment-relatedphenomena in couples residing in Spain.

Attachment in romantic relationships

Attachment theory was initially formulatedby Bowlby (1969/1982) and tested empiri-cally by Ainsworth, Blehar, Waters, and Wall(1978), who identified three major patterns ofinfant attachment to primary caregivers (usu-ally the infant’s mother), which here we willlabel secure, anxious, and avoidant attach-ment. Some years later, Hazan and Shaver(1987) proposed applying Bowlby’s theoryand Ainsworth and colleagues’ research find-ings to late adolescent and adult romanticlove and couple relationships. Subsequently,Bartholomew and Horowitz (1991) arguedthat what Hazan and Shaver had called“romantic attachment” could be understoodin terms of two conceptually independentdimensions—model of self and model of oth-ers—which together defined four attachmentstyles: secure, preoccupied, fearful, and dis-missing. Empirical analyses (e.g., Brennan,Clark, & Shaver, 1998; Fraley & Waller,1998; Griffin & Bartholomew, 1994) sup-ported this conception of attachment patterns,and most recent studies have been based onmeasures of the two main dimensions thatunderlie the four-category typology: anxiousattachment (characterized by concerns aboutrejection and abandonment) and avoidantattachment (characterized by discomfort withcloseness and interdependence and a pref-erence for self-reliance). Numerous studiesbased on these two dimensions have beenpublished, and the majority of them have beenreviewed by Mikulincer and Shaver (2007).

Recently, the measure used in most ofthese studies, the Experiences in Close Rela-tionships inventory (ECR; Brennan et al.,1998), was translated into Spanish and wasshown to have good reliability and preliminaryconstruct validity (Alonso-Arbiol, Balluerka,& Shaver, 2007). These early validity studiesdid not, however, include multilevel analysesto explore couple dynamics, self-esteem, and

social self-efficacy, which are the focus of thisstudy.

Associations between attachment insecuritiesand intrapersonal processes

Two intrapersonal variables have receivedconsiderable attention from social and per-sonality psychologists, self-esteem and self-efficacy. Several authors (e.g., Kirkpatrick& Ellis, 2001; Leary, Tambor, Terdal, &Downs, 1995) have characterized self-esteemas the read-out of a “sociometer” that mayhave evolved to monitor a person’s level ofsocial inclusion or rejection by a relation-ship partner or group—a valuable psycho-logical asset for members of a highly socialspecies. Because attachment insecurities arethought to stem from the lack of adequateacceptance and support by caregivers duringchildhood, it is reasonable to expect measuresof these insecurities, especially attachmentanxiety (fear of rejection or abandonment),to be inversely correlated with measures ofself-esteem. Such negative correlations havein fact been reported over the years byresearchers ranging from Bartholomew andHorowitz (1991) to Schmitt and Allik (2005),the latter of whom found a significant negativecorrelation between attachment anxiety andscores on the Rosenberg Self-Esteem Scale(Rosenberg, 1965) in 49 countries includingSpain.

The association between avoidant attach-ment and self-esteem is less clear. Althoughavoidance is a form of attachment insecu-rity, avoidant individuals are highly motivatedto be, and to view themselves as, indepen-dent and self-reliant, which may make themreluctant to acknowledge low or unstableself-esteem. In their review of the researchliterature on adult attachment, Mikulincerand Shaver (2007) found negative correla-tions between self-esteem and avoidance inabout half of the 60 studies they reviewed.These correlations suggest that the defen-sive strategies of avoidant people—whichinclude suppressing thoughts of vulnerabilityand dependence—are not always successful.

Regarding the association between attach-ment insecurities and a slightly different

Attachment insecurities in Spanish couples 3

construct, self-efficacy (Bandura, 1977), theresults also differ with respect to attachmentanxiety and avoidance. The majority of stud-ies have found that attachment anxiety isassociated with relatively low self-efficacyacross all life domains studied (e.g., global,social, academic, and athletic). In contrast,avoidance is often associated positively withperceived self-efficacy in nonsocial domains,such as personal achievement, but associatednegatively with self-efficacy in interpersonaldomains (often referred to as relationshipor social self-efficacy; e.g., Mallinckrodt &Wei, 2005; Taubman-Ben-Ari, Findler, &Mikulincer, 2002; Wei, Russell, & Zakalik,2005). As with other studies of adult attach-ment and couple relationships, most of thestudies published to date have involved onlyEnglish-speaking participants.

Attachment and relationship satisfaction

Many researchers and literature reviewershave concluded that satisfaction or adjustmentin long-term romantic relationships is impor-tant for people’s health and well-being (fora review, see Berscheid & Reis, 1998). InSpain, which is of special interest here, twonational surveys (Centro de InvestigacionesSociologicas [CIS], 2002, 2004) showed thatfor almost all participants family was very(78.5%) or quite (20.4%) important, and mostparticipants were highly satisfied with theirfamilies (averaging 8.2 on a 1–10 scale).In their comprehensive review of attachmentstudies (mostly conducted in English-speakingcountries), Mikulincer and Shaver (2007)found that attachment insecurities (both att-achment anxiety and avoidance) and rela-tionship satisfaction were negatively corre-lated among both women and men. However,although anxiety and avoidance were roughlyequally predictive of women’s dissatisfaction,avoidance was more consistently associatedwith relationship dissatisfaction in men.

Although the majority of studies havefocused on the association between one per-son’s attachment insecurities and his or herown relationship satisfaction, there are somestudies of the joint influence of both partners’insecurities. This research generally shows

that partners of relatively anxious or avoidantindividuals have lower relationship satis-faction than partners of relatively secureindividuals (Banse, 2004; Feeney, 2002;Kachadourian, Fincham, & Davila, 2004;Shaver, Schachner, & Mikulincer, 2005).

Some researchers (e.g., Baldwin, Keelan,Fehr, Enns, & Koh Rangarajoo, 1996; Frazier,Byer, Fischer, Wright, & DeBord, 1996;Klohnen & Luo, 2003) have consideredwhether partners’ attachment styles are relatedto each other, as might be expected basedon the well-researched general hypothesis thatsimilarity enhances attraction (Byrne, 1971).There is some evidence that secure individ-uals are more attracted to secure individu-als, anxious individuals are more attracted toanxious individuals, and avoidant individu-als are more attracted to avoidant individu-als, although this pattern is not always found(see Holmes & Johnson, 2009, for a review).Regarding the effects of different attachment-style pairings on relationship satisfaction, itseems that the combination of an anxiousperson with an avoidant one is detrimentalto relationship satisfaction, as is the combi-nation of two anxious individuals (Allison,Bartholomew, Mayseless, & Dutton, 2008;Feeney, 1994; Roberts & Noller, 1998).

The dyadic perspective on adult attachment

As Kenny, Kashy, and Cook (2006, p. 144)noted: “Virtually all major theories of roman-tic relationships, including theories of equity,commitment, trust, interdependence, and at-tachment, acknowledge the idea that one part-ner’s attributes and behaviors can affect theother partner’s outcomes.” However, most ofthe research on adult attachment has beenconducted at the individual level of analy-sis. Often, only one member of each coupleis studied. Nevertheless, in the last few yearsseveral studies have been conducted from adyadic perspective. For example, Brassard,Shaver, and Lussier (2007) and Butzer andCampbell (2008) documented the influenceof both members’ attachment insecurities onboth partners’ sexual and marital satisfac-tion. Millings and Walsh (2009) explored therelation between attachment and caregiving in

4 F. Molero et al.

long-term couples, and Simpson, Winterhedl,Rholes, and Minda Orina (2007) exploredthe relation between attachment and the reac-tions to different forms of caregiving fromromantic partners. These studies all involvedEnglish- or French-speaking couples in theUnited States and Canada.

To analyze the mutual influences betweentwo couple members, a dyadic perspective isrequired. In this study, we adopt this per-spective and analyze several correlates ofattachment insecurities that have received lit-tle attention in the literature, especially inSpain. Specifically, we examine:

1. the effects of actor and partner attach-ment insecurities (attachment anxietyand avoidance) on intrapersonal pro-cesses such as self-esteem and per-ceived social self-efficacy;

2. the influence of actor and partner attach-ment insecurities on both partners’ rela-tionship satisfaction;

3. the existence of similarity or comple-mentarity between the attachment ori-entations of relationship partners.

Objectives and hypotheses

Our goal was to examine attachment-relatedphenomena within established couples resid-ing in Spain, using both individual-level anddyadic-level analyses. On the basis of adultattachment theory (Mikulincer & Shaver,2007), the findings from previous studies,and advances in dyadic analysis methods,we tested several hypotheses using multilevelmodeling (MLM), as suggested by creatorsof the actor–partner interdependence model(APIM; Kashy & Kenny, 2000; Kenny et al.,2006). The APIM will be explained in moredetail in the Analyses section.

H1 (Model 1): Actors’ attachment inse-curities (anxiety and avoidance) willbe negatively correlated with theirown self-esteem.

Inverse associations between actor attach-ment insecurities and actor self-esteem havebeen reported in the literature (e.g., Schmitt &

Allik, 2005), but little is known about associa-tions between one relationship partner’s inse-curities and the other partner’s self-esteem.This association will be examined in our anal-yses, but we advance no a priori hypothesisabout it.

H2 (Model 2): Actor attachment insecu-rities will be inversely associated withactor social self-efficacy.

As is the case of self-esteem, there are anumber of studies reporting an inverse asso-ciation between attachment insecurities andactor social self-efficacy (e.g., Mallinckrodt& Wei, 2005; Wei et al., 2005). Nevertheless,the possible links between one partner’s inse-curities and the other’s social self-efficacy areunclear. This will be explored in Model 2.

H3 (Model 3): Actor and partner attach-ment insecurities will be inverselyassociated with both partners’ rela-tionship satisfaction.

Several studies have documented thenegative associations between one person’sattachment insecurities and his or her ownrelationship satisfaction. The associations be-tween one partner’s insecurities and the otherpartner’s satisfaction have been less often ex-plored, although a few studies have found theexpected negative associations (e.g., Banse,2004).

H4 (Models 4 and 5): One partner’sattachment anxiety will be associatedwith the other partner’s avoidanceand vice versa, because anxious indi-viduals should cause their partnersto draw away, and avoidant individ-uals should cause their partners tofeel insecure, anxious, and desirous ofgreater closeness.

Method

Participants and procedure

The participants, 295 heterosexual couples,were recruited and interviewed by under-graduate psychology students at the UNED


(National Open University, Spain) who re-ceived practicum credits for conducting theinterviews. Each student was asked to inter-view both members of at least three heterosex-ual couples (regardless of relationship length)using the questionnaire described below. Par-ticipant couples resided in various parts ofSpain, 26% in large cities (with a pop-ulation of more than 1 million), 43% inmiddle size cities (with 50,000 to 1 millioninhabitants), and 31% in small towns (withfewer than 50,000 inhabitants). The age ofthe participants ranged from 17 to 77 years(M = 34.3, SD = 12.05). Relationship lengthranged from 6 months to 60 years, with amean of 6.02 years and a SD of 9.21 years.

Instruments

Attachment

Participants answered a Spanish version(Alonso-Arbiol et al., 2007) of the ECR(Brennan et al., 1998). The ECR containstwo 18-item scales that measure attachment-related anxiety and avoidance. The mea-sure has been used in scores of studies,in a variety of countries and languages,since 1998 (see Mikulincer & Shaver, 2007,for a review). In this study, the internalconsistency reliabilities (Cronbach’s αs) ofthe Spanish versions of the scales were.84 and .85 for anxiety and avoidance,respectively.

Self-esteem

We used a Spanish translation (Exposito &Moya, 1999) of the 10-item Rosenberg Self-Esteem Scale (Rosenberg, 1965), which con-tains items such as “I am able to do things aswell as most other people” and “On the whole,I am satisfied with myself.” Ratings weremade on a scale ranging from 1 (completelydisagree) to 5 (completely agree). In thisstudy, the coefficient α for the scale was .76.

Social self-efficacy

We used the six-item Social Self-EfficacyScale designed by Sherer and colleagues(1982) to measure self-perceived social and

relationship skills. This scale was validated inSpain by Lopez-Torrecillas, García, Canadas,Ucles, and de la Fuente (2006). The fol-lowing items are representative: “I do nothandle myself well in social gatherings”(reversed), “It is difficult for me to make newfriends” (reversed), and “I have acquired myfriends through my personal abilities at mak-ing friends.” Items are rated on a 5-point scaleranging from 1 (completely disagree) to 5(completely agree). In this study, the coeffi-cient α for this scale was .78.

Relationship satisfaction

The participants answered the following ques-tion, “To what extent are you satisfied withyour current relationship?” using a 5-pointscale ranging from 1 (not at all satisfied )to 5 (very satisfied ). Although single-itemscales have psychometric limitations, they canbe useful in the absence of more extensivemeasures and can be adequate alternatives tolonger scales (Barrett & Paltiel, 1996; Robins,Hendin, & Trzesniewski, 2001).

Sociodemographic variables

The participants also provided sociodemo-graphic information (e.g., relationship dura-tion, age, gender, and education level).

Analyses

The analyses were conducted from a dyadicperspective using the APIM. When individualsare members of a couple, their outcomes can-not be assumed to be independent; they arelikely to be a function of both their owncharacteristics (actor effects) and their part-ner’s characteristics (partner effects). Findingpartner effects in the data provides evi-dence of the couple members’ interdepen-dence (Kenny et al., 2006). To statisticallyestimate the actor and partner effects, we usedMLM, a flexible approach that allows for thetesting of interactions (Campbell & Kashy,2002; Kenny et al., 2006). In particular, ourMLM approach treated data from each partneras nested within their dyad. We coded genderas a dummy variable (−1 for women and 1for men).

6 F. Molero et al.

Table 1. Descriptive statistics for all variables, broken down by gender

Women Men

Variables M SD M SD t Cohen’s d

Anxiety 3.07 .59 2.99 .62 1.93∗ 0.16Avoidance 2.01 .52 2.12 .54 −2.53∗ 0.21Self-esteem 3.92 .58 4.03 .58 −2.31∗ 0.19Social self-efficacy 3.39 .79 3.47 .85 −1.24 0.10Relationship satisfaction 4.21 .83 4.40 .75 −3.16∗∗ 0.26

Note. Scores could range from 1 to 5.∗p < .05. ∗∗p < .01.

Results

Descriptive statistics and correlations

The means and standard deviations for allvariables are shown in Table 1, separatelyfor men and women. Because the samplesize was large, there were significant genderdifferences on all variables, with the excep-tion of social self-efficacy. Nevertheless, theeffect sizes were small in all cases. Men wereslightly less anxious, more avoidant, higher inself-esteem, and more satisfied with their rela-tionships than women. These results fit withgender-role stereotypes (see, e.g., Deaux &Kite, 1993; Spence & Buckner, 2000). Table 2displays the zero-order correlations amongthe variables, separately for men and women.Most of the correlation coefficients are in thesame directions and of roughly similar sizesfor men and women, but there are some signif-icant differences in the correlations involvingavoidance. The correlation between avoidanceand self-esteem is more strongly negative formen than for women (z = 1.96, p = .05, two-tailed), and the correlation between avoidance

and relationship satisfaction is more stronglynegative for women than for men (z = 2.23,p < .05, two-tailed).

Table 3 presents the zero-order correlationsbetween men’s and women’s variables. In thediagonal of this table, it can be seen that thereare significant correlations between men’s andwomen’s avoidance, self-esteem, and relation-ship satisfaction scores. These correlationsindicate that the couple members’ scores arenot independent, which indicates the valueof a dyad-level analyses (Kenny et al., 2006,p. 28).

Hypothesis tests based on the APIM

To simplify the presentation of results wereport only significant effects.

Model 1 (Hypothesis 1)

This model tests the actor and partner eff-ects of attachment insecurities (anxiety andavoidance) on self-esteem, and the result wasexpected to be a negative relation between the

Table 2. Pearson correlations among variables

Variables 1 2 3 4 5

1. Anxiety — −.02 −.23∗∗ −.19∗∗ −.12∗2. Avoidance −.01 — −.21∗∗ −.09 −.45∗∗3. Self-esteem −.17∗∗ −.36∗∗ — .30∗∗ .12∗4. Social self-efficacy −.14∗ −.24∗∗ .38∗∗ — −.045. Relationship satisfaction −.04 −.29∗∗ .19∗∗ .04 —

Note. Women’s coefficients are above the diagonal, whereas men’s coefficients are below the diagonal.∗p < .05. ∗∗p < .01.


Table 3. Pearson correlations between men’s and women’s variables

Variables 1 (W) 2 (W) 3 (W) 4 (W) 5 (W)

1. Anxiety (M) .07 .24∗∗ −.01 −.04 −.22∗∗

2. Avoidance (M) .21∗∗ .27∗∗ −.13∗∗ −.04 −.24∗∗

3. Self-esteem (M) −.04 −.09 .18∗∗ .10 .14∗∗

4. Social self-efficacy (M) −.03 −.11∗ .06 .03 .025. Relationship satisfaction (M) −.11∗ −.23∗∗ .11∗ .03 .41∗∗

Note. W = women; M = men.∗p < .05. ∗∗p < .01.

insecurity scores and self-esteem. The pre-dictor variables in the model included gen-der and both actor and partner scores onthe anxiety and avoidance scales. As can beseen in Table 4, actors’ attachment anxietyand avoidance scores were negatively associ-ated with their own self-esteem, as predicted,but there were no partner or gender effects,suggesting that couple members’ self-esteemwas not affected by their partners’ attachmentinsecurities.

Table 4. Actor and partner effects of anxietyand avoidance on self-esteem and social self-efficacy (Models 1 and 2)

Self-esteemSocial

self-efficacyPredictorvariables b SE b SE

Intercept 5.03∗∗ .19 4.59∗∗ .26Gender .08 .16 .17 .25

AttachmentanxietyActor effect −.20∗∗ .03 −.22∗∗ .06Partner effect .06 .03 .02 .06

AttachmentavoidanceActor effect −.33∗∗ .04 −.28∗∗ .07Partner effect .01 .04 .01 .07

Note. b represents an unstandardized regressioncoefficient.∗∗p < .001.


This model tests the actor and partner effectsof attachment insecurities (anxiety and avoid-ance) on social self-efficacy, and negativeassociations were expected. The predictorvariables in the model included gender andboth actor and partner scores on the anxietyand avoidance scales. As can be seen inTable 4, actors’ attachment anxiety and avoid-ance scores were negatively related to theirown social self-efficacy, but as in the test ofHypothesis 1, there were no partner or gendereffects.


This model tests the actor and partner effectsof attachment insecurities (anxiety and avoid-ance) on relationship satisfaction. The predic-tor variables in the model included gender andactor and partner scores on the anxiety andavoidance scales. As can be seen in Table 5,there were actor and partner effects of avoid-ance on relationship satisfaction. But in thecase of attachment anxiety, there was only aneffect of the partner; with the other variablesin the model, actor attachment anxiety didnot affect actor’s own relationship satisfac-tion. There was also a significant main effectof gender, as already seen in Table 1, indi-cating that men were, on average, more satis-fied with their relationships than were women.Moreover, there was a significant interac-tion between actor avoidance and gender:Avoidant women were less satisfied thanavoidant men.

To determine whether the effect of ac-tor’s avoidance on relationship satisfaction

8 F. Molero et al.

Table 5. Actor and partner effects of anxi-ety and avoidance on relationship satisfaction(Model 3)

Predictor variables b SE

Intercept 6.24∗∗ .27Gender .12∗∗ .03Attachment anxiety

Actor effect −.05 .06Partner effect −.13∗ .05

Attachment avoidanceActor effect −.46∗∗ .06Partner effect −.21∗∗ .06

InteractionsActor Avoidance ×

Gender.15∗ .06

Note. b represents an unstandardized regressioncoefficient.∗p < .05. ∗∗p < .001.

(b = −.46) was significantly greater than theeffect of actor’s anxiety (b = −.05), as itseems to be, we used a Wald test, which isbased on the estimates of the parameters to becompared in relation to the standard error oftheir difference.1 The result was highly sig-nificant, F(1, 292) = 30.58, p < .001. Theeffects of partner avoidance (b = −.21) and

1. To formally evaluate whether the two coefficients weredifferent from each other, we used a Wald test basedon the estimates of the parameters in question andthe standard error of the difference between them.Specifically, this approach is based on two premises:

1 . β1 = β2; or β1 − β2 = 0, and

2 . σ2x±y = σ2

x + σ2y + 2σxy, from which

σ2β1−β2 = σ2

β1 + σ2β2 + 2σβ1,β2.

An appropriate test statistic for this hypothesis is asfollows:

F1,N−K−1 =

⎛⎜⎝ (b1 − b2) − (β1 − β2)0√

σ2b1 + σ2

b2 − 2σb1,b2

⎞⎟⎠

2

=

⎛⎜⎝ (b1 − b2)√

σ2b1 + σ2

b2 − 2σb1,b2

⎞⎟⎠

2

.

In other words, the observed value minus the predictedvalue (from the null hypothesis β1 − β2 = 0) is dividedby the standard error of the difference.

anxiety (b = −.13) on satisfaction were com-pared in the same way, and the difference wasnot significant, F(1, 292) = 1.20, p > .05.

To explore the possible influence of actorand partner self-esteem and social self-efficacyon relationship satisfaction, we tested anothermodel including these two variables togetherwith the attachment insecurity variables (anx-iety and avoidance). None of the four vari-ables—actor self-esteem, partner self-esteem,actor social self-efficacy, and partner socialself-efficacy—was significantly associatedwith relationship satisfaction, and their pres-ence in the model did not change thepattern of associations between the two attach-ment insecurity variables and relationshipsatisfaction.

Models 4 and 5 (Hypothesis 4)

Model 4 examines the extent to which actoranxiety is related to actor or partner avoid-ance when the effects of gender are controlled.As predicted in Hypothesis 4, actor anxietywas positively associated with partner avoid-ance. Model 5 tests the extent to which actoravoidance is affected by actor or partner anx-iety when the effects of gender are controlled.As predicted in Hypothesis 4, actor avoid-ance was positively associated with partneranxiety. Results for both models are reportedin Table 6. These results indicate a potentiallydestructive relationship dynamic in which onepartner’s anxiety increases the other’s avoid-ance, and one partner’s avoidance increasesthe other partner’s anxiety.

Discussion

The main objective of this study was toexamine attachment-related phenomena in asample of Spanish couples. The study hasthree characteristics that distinguish it frommost previous studies of adult attachment andromantic or marital relationships: Analyseswere conducted at both the individual andthe dyadic levels, participants were membersof established couples living outside a col-lege or university context, and the study wasconducted in a non-English-speaking coun-try, Spain. Overall, the results confirm links


Table 6. Actor and partner effects of anxietyon avoidance and vice versa (Models 4 and 5)

Anxiety AvoidancePredictorvariables b SE b SE

Intercept 2.64∗∗∗ .13 1.54∗∗∗ .17Gender −.00 .00 .02 .03Attachment

anxietyActor effect — — −.02 .03Partner effect — — .20∗∗∗ .03

AttachmentavoidanceActor effect −.08 .04 — —Partner effect .27∗∗∗ .05 — —

Note. b represents an unstandardized regression coeffi-cient.∗∗∗p < .001.

between partners’ attachment insecurities andpersonal and relationship well-being, and theyalso confirm the value of the APIM in couplerelationship studies.

Intrapersonal effects of attachmentinsecurities

As predicted by Hypothesis 1, higher levels ofanxiety and avoidance were associated withlower levels of self-esteem at the individuallevel when controlling for partner’s scoreson attachment insecurities. These results areconsistent with the literature on attachmentanxiety (Bartholomew & Horowitz, 1991;Schmitt & Allik, 2005) and are compatiblewith a number of studies that have foundan association between avoidance and lowself-esteem (e.g., Gentzler & Kerns, 2004).The results suggest that attachment insecu-rity and more general forms of self-doubt arerelated, either because of shared genetic influ-ences (Crawford et al., 2007; Donnellan, Burt,Levendosky, & Klump, 2008; Gillath, Shaver,Baek, & Chun, 2008) or because of the sharedinfluences of personal history emphasizedby attachment theory (see reviews in Cas-sidy & Shaver, 2008). Within this constella-tion of insecurities, however, only the attach-ment insecurities were significantly related

to partners’ relationship outcomes in ourstudy.

There was a significant correlation betweenpartners’ self-esteem levels. Longitudinal stu-dies would be needed to determine whetherpeople with similar levels of self-esteem tendto form relationships with each other or,alternatively, one partner’s self-esteem influ-ences the other partner’s self-esteem, perhapsthrough associated personal qualities, such asassertiveness or communication skills, acrosstime. Interestingly, in the APIM analyses,self-esteem did not predict relationship sat-isfaction when the attachment variables wereincluded in the model.

As predicted in Hypothesis 2, attachmentinsecurities were inversely related to per-ceived social self-efficacy at the individuallevel when controlling for partner’s scores onattachment insecurities (see also Mallinckrodt& Wei, 2005; Wei et al., 2005). As with self-esteem, there were no significant associationsbetween one partner’s attachment insecuri-ties and the other partner’s social self-efficacyor relationship satisfaction. Thus, social self-efficacy, at least when measured in gen-eral (rather than intimate relationship-specificterms), seems to be a self-perception that doesnot play a role in dyadic processes.

Associations between actor and partnerattachment insecurities and relationshipsatisfaction

As predicted by Hypothesis 3, both actor andpartner avoidance were negatively associatedwith both partners’ relationship satisfaction.That is, avoidant people tended to be lesssatisfied with their relationships, and theirrelationship partners were also less satisfied.For attachment anxiety, however, we foundpartner but not actor effects in the APIManalyses, suggesting that anxious people mayfeel satisfied with their relationships even iftheir attachment-related behavior reduces theirpartner’s satisfaction.

There was also a main effect of gender:Men reported higher levels of relationshipsatisfaction than women. There was also aninteraction between actor avoidance and gen-der, indicating that for women avoidance was

10 F. Molero et al.

more closely related to dissatisfaction. Thisis similar to findings from the United Statesreported by Collins and Read (1990), and itmay be a result of men’s avoidance beingmore congruent than women’s avoidance withgender roles and stereotypes.

The model including self-esteem and socialself-efficacy in addition to the two attach-ment insecurity scores revealed no significanteffects of self-esteem or social self-efficacy onrelationship satisfaction. This is not surprisingin the case of social self-efficacy, because itwas not correlated with satisfaction in Table 2either, but it suggests that the negative cor-relation between self-esteem and satisfactionin Table 2 was a function of the connectionbetween self-esteem and attachment insecuri-ties. (And in fact, in an analysis not reportedin the Results section, we found that the asso-ciation between self-esteem and relationshipsatisfaction disappeared when the attachmentinsecurity variables were partialed.) The fail-ure of social self-efficacy as measured in thisstudy to be related to relationship satisfactionmay be a result of our having measured it ina general rather than in relationship-specificway.

The results suggest that avoidance is moredetrimental than anxiety to relationship satis-faction, at least in Spain, where adults gener-ally report higher attachment anxiety scoresthan adults in the United States (Alonso-Arbiol et al., 2008; Wei et al., 2004). Attach-ment anxiety may be more acceptable inSpain than in the United States—an issuethat should be examined directly in futurestudies.

Joint effects of attachment insecurities inboth members of a couple

The results for Hypothesis 4 indicate thatcouple members’ attachment insecurity scoreswere interrelated. In particular, one partner’sanxiety was related to the other partner’savoidance (see Kirkpatrick & Davis, 1994,and Kirkpatrick & Hazan, 1994, for compa-rable findings in the United States). Becauseour study was cross-sectional, we cannotdetermine whether anxious and avoidant indi-viduals were attracted to each other because

of preexisting attachment orientations, or thetwo people’s insecurities developed over timeas one partner withdrew and the other pur-sued (Bartholomew & Allison, 2006). This isanother important topic for future research.

Although our study contributes to theunderstanding of the role of attachment inse-curities in long-term couple relationships, itis limited in certain respects. First, we usedonly self-report measures rather than behav-ioral observations. Second, although the studywas cross-sectional in design, we conceptual-ized the analysis mainly in terms of the effectsof attachment insecurities on other variables,including both partners’ relationship satisfac-tion. It is also possible that difficulties ina relationship, leading to lower satisfaction,affect attachment anxiety and avoidance overtime, perhaps especially avoidance, whichinvolves backing away from intimacy andinterdependence. Attachment theory viewsattachment insecurities primarily as outcomesof actual relationship experiences, so thecausal model in the case of adults may involvebidirectional forces operating over time. Thatis, all variables in the model are subject tochange and therefore warrant study over time.

Finally, regarding the issue of cultural dif-ferences and similarities, we obtained manyfindings similar to those obtained in priorEnglish-language studies. Attachment insecu-rities were related, at the individual level,to lower self-esteem and lower social self-efficacy, and men’s avoidance scores wereslightly higher than women’s avoidance scor-es, which sometimes occurs in English-lang-uage studies as well. Avoidance was moreassociated than anxiety with low relationshipsatisfaction, which is also common in English-language studies (see Mikulincer & Shaver,2007, for a review). One partner’s anxietywas related negatively to the other partner’ssatisfaction and avoidance, suggesting thatattachment anxiety causes difficulties in arelationship, especially for avoidant partners.

In some English-language studies, wo-men’s attachment anxiety seems to be agreater problem for their male partners’ sat-isfaction than men’s attachment anxiety is aproblem for their female partners’ satisfaction(e.g., Collins & Read, 1990), but this was not


the case in this study. If anything, men’s anx-iety was a greater problem for their femalepartners than women’s anxiety was a prob-lem for their male partners, as indicated bythe correlation coefficients of −.22 and −.11in Table 3, but these coefficients were not sig-nificantly different.

Thus, overall, the pattern of findings forSpanish couples was very consistent with thepattern of findings in most English-languagestudies, suggesting cross-culturally generalpatterns, as have been found in studies ofinfant–parent attachment as well (see reviewby van IJzendoorn & Sagi-Schwartz, 2008). Itremains to be seen if this generality extends tonations and cultures more different than Spainand Anglo-American countries and cultures.

References

Ainsworth, M. D. S., Blehar, M. C., Waters, E., &Wall, S. (1978). Patterns of attachment: Assessed inthe Strange Situation and at home. Hillsdale, NJ: Erl-baum.

Allison, C., Bartholomew, K., Mayseless, O., & Dutton,D. G. (2008). Love as a battlefield: Relationshipdynamics in couples identified for male partner vio-lence. Journal of Family Issues, 29, 125–150.

Alonso-Arbiol, I., Balluerka, N., & Shaver, P. R. (2007).A Spanish version of the Experiences in Close Rela-tionships (ECR) adult attachment questionnaire. Per-sonal Relationships, 14, 45–63.

Alonso-Arbiol, I., Balluerka, N., Shaver, P. R., & Gillath,O. (2008). Psychometric properties of the Spanish andAmerican versions of the ECR adult attachment ques-tionnaire: A comparative study. European Journal ofPsychological Assessment, 24, 9–13.

Baldwin, M. W., Keelan, J. P. R., Fehr, B., Enns, V., &Koh Rangarajoo, E. (1996). Social-cognitive concep-tualization of attachment working models: Availabil-ity and accessibility effects. Journal of Personalityand Social Psychology, 71, 94–109.

Bandura, A. (1977). Self-efficacy: Toward a unifyingtheory of behavioural change. Psychological Review,84, 191–215.

Banse, R. (2004). Adult attachment and marital satis-faction: Evidence for dyadic configuration effects.Journal of Social and Personal Relationships, 21,273–282.

Barrett, P. T., & Paltiel, L. (1996). Can a single itemreplace an entire scale? POP vs the OPQ 5.2. Selectionand Development Review, 12, 1–4.

Bartholomew, K., & Allison, C. (2006). An attachmentperspective on abusive dynamics in intimate relation-ships. In M. Mikulincer & G. S. Goodman (Eds.),Dynamics of romantic love: Attachment, caregiving,and sex (pp. 102–127). New York: Guilford.

Bartholomew, K., & Horowitz, L. M. (1991). Attach-ment styles among young adults: A test of a four-category model. Journal of Personality and SocialPsychology, 61, 226–244.

Berscheid, E., & Reis, H. T. (1998). Attraction andclose relationships. In D. T. Gilbert, S. T. Fiske, &G. Lindzey (Eds.), The handbook of social psychology(4th ed., pp. 193–281). New York: McGraw-Hill.

Bowlby, J. (1982). Attachment and loss: Vol. 1. Attach-ment (2nd ed.). New York: Basic Books. (Originalwork published 1969)

Brassard, A., Shaver, P. R., & Lussier, Y. (2007). Attach-ment, sexual experience, and sexual pressure inromantic relationships: A dyadic approach. PersonalRelationships, 14, 475–493.

Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998).Self-report measurement of adult romantic attach-ment: An integrative overview. In J. A. Simpson &W. S. Rholes (Eds.), Attachment theory and closerelationships (pp. 46–76). New York: Guilford.

Butzer, B., & Campbell, L. (2008). Adult attachment,sexual satisfaction, and relationship satisfaction: Astudy of married couples. Personal Relationships, 15,141–154.

Byrne, D. (1971). The attraction paradigm. New York:Academic Press.

Campbell, L., & Kashy, D. A. (2002). Estimating actor,partner, and interaction effects for dyadic data usingPROC MIXED and HLM: A user-friendly guide.Personal Relationships, 9, 327–342.

Cassidy, J., & Shaver, P. R. (Eds.). (2008). Handbookof attachment: Theory, research, and clinical appli-cations (2nd ed.). New York: Guilford.

Centro de Investigaciones Sociologicas. (2002). Datos deOpinion [Opinion Data], 29. Retrieved December 1,2008, from http://www.cis.es/cis/opencms/Boletines/29/BSO 29 Relaciones personales.html#felicidad

Centro de Investigaciones Sociologicas. (2004). Opin-iones y Actitudes sobre la Familia [Opinions andattitudes regarding family]. Retrieved December 1,2008, from http://www.cis.es/cis/opencm/ES/1 encuestas/estudios/ver.jsp?estudio=4556

Collins, N. L., & Read, S. J. (1990). Adult attachment,working models, and relationship quality in datingcouples. Journal of Personality and Social Psychol-ogy, 58, 644–663.

Crawford, T. N., Livesley, W. J., Jang, K. L., Shaver,P. R., Cohen, P., & Ganiban, J. (2007). Insecureattachment and personality disorder: A twin studyof adults. European Journal of Personality, 21,191–208.

Deaux, K., & Kite, M. (1993). Gender stereotypes. InF. L. Denmark & M. A. Paludi (Eds.), Psychologyof women: A handbook of issues and theories (pp.107–139). Westport, CT: Greenwood Press.

Donnellan, M. B., Burt, S. A., Levendosky, A. A., &Klump, K. L. (2008). Genes, personality, and at-tachment in adults: A multivariate behavioral geneticanalysis. Personality and Social Psychology Bulletin,34, 3–16.

12 F. Molero et al.

Exposito, F., & Moya, M. (1999). Soledad y apoyo social[Loneliness and social support]. Revista de PsicologíaSocial, 14, 297–316.

Feeney, J. A. (1994). Attachment style, communicationpatterns, and satisfaction across the life cycle ofmarriage. Personal Relationships, 1, 333–348.

Feeney, J. A. (2002). Attachment, marital interaction,and relationship satisfaction: A diary study. PersonalRelationships, 9, 39–55.

Fraley, R. C., & Waller, N. G. (1998). Adults attach-ment patterns: A test of the typological model. InJ. A. Simpson & W. S. Rholes (Eds.), Attachmenttheory and close relationships (pp. 77–114). NewYork: Guilford.

Frazier, P. A., Byer, A. L., Fischer, A. R., Wright, D. M.,& DeBord, K. A. (1996). Adult attachment style andpartner choice: Correlational and experimental find-ings. Personal Relationships, 3, 117–136.

Gentzler, A. L., & Kerns, K. A. (2004). Associationsbetween insecure attachment and sexual experiences.Personal Relationships, 11, 249–265.

Gillath, O., Shaver, P. R., Baek, J.-M., & Chun, D. S.(2008). Genetic correlates of adult attachment style.Personality and Social Psychology Bulletin, 34,1396–1405.

Griffin, D. W., & Bartholomew, K. (1994). Models of theself and other: Fundamental dimensions underlyingmeasures of adult attachment. Journal of Personalityand Social Psychology, 67, 430–445.

Hazan, C., & Shaver, P. R. (1987). Romantic love con-ceptualized as an attachment process. Journal of Per-sonality and Social Psychology, 52, 511–524.

Holmes, B. M., & Johnson, K. R. (2009). Adult attach-ment and romantic partner preference: A review.Journal of Social and Personal Relationships, 26,833–852.

Kachadourian, L. K., Fincham, F., & Davila, J. (2004).The tendency to forgive in dating and married cou-ples: The role of attachment and relationship satisfac-tion. Personal Relationships, 11, 373–393.

Kashy, D. A., & Kenny, D. A. (2000). The analysisof data from dyads and groups. In H. T. Reis &C. M. Judd (Eds.), Handbook of research methods insocial psychology (pp. 451–477). New York: Cam-bridge University Press.

Kenny, D. A., Kashy, D. A., & Cook, W. L. (2006).Dyadic data analysis. New York: Guilford.

Kirkpatrick, L. A., & Davis, K. E. (1994). Attachmentstyle, gender, and relationship stability: A longitudinalanalysis. Journal of Personality and Social Psychol-ogy, 66, 502–512.

Kirkpatrick, L. A., & Ellis, B. J. (2001). An evolutionaryapproach to self-esteem: Multiple domains and mul-tiple functions. In G. J. O. Fletcher & Clark, M. S.(Eds.), Blackwell handbook of social psychology:interpersonal processes (pp. 411–436). Oxford, UK:Blackwell.

Kirkpatrick, L. A., & Hazan, C. (1994). Attachment styleand close relationships: A four-year prospectivestudy. Personal Relationships, 1, 123–142.

Klohnen, E. C., & Luo, S. (2003). Interpersonal attrac-tion and personality: What is attractive-self similarity,ideal similarity, complementarity or attachment secu-rity. Journal of Personality and Social Psychology,85, 709–723.

Leary, M. R., Tambor, E. S., Terdal, S. K., & Downs,D. L. (1995). Self-esteem as an interpersonal monitor:The sociometer hypothesis. Journal of Personalityand Social Psychology, 68, 518–530.

Lopez-Torrecillas, F., García, J., Canadas, G. A., Ucles,I. R., & de la Fuente, E. I. (2006).Validity of self-efficacy scale scores for a Spanish sample. Psycho-logical Reports, 98, 437–450.

Mallinckrodt, B., & Wei, M. (2005). Attachment, socialcompetencies, social support, and psychological dis-tress. Journal of Counseling Psychology, 52, 358–367.

Mikulincer, M., & Shaver, P. R. (2007). Attachment inadulthood: Structure, dynamics, and change. NewYork: Guilford.

Millings, A., & Walsh, J. (2009). A dyadic explorationof attachment and caregiving in long-term couples.Personal Relationships, 16, 437–453.

Roberts, N., & Noller, P. (1998). The associationsbetween adult attachment and couple violence: Therole of communication patterns and relationship sat-isfaction. In J. A. Simpson & W. S. Rholes (Eds.),Attachment theory and close relationships (pp.317–350). New York: Guilford.

Robins, R. W., Hendin, H. M., & Trzesniewski, K. H.(2001). Measuring global self-esteem: Construct val-idation of a single-item measure and the RosenbergSelf-Esteem Scale. Personality and Social PsychologyBulletin, 27, 151–161.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press.

Schmitt, D. P., & Allik, J. (2005). Simultaneous admin-istration of the Rosenberg Self-Esteem Scale in 53nations: Exploring the universal and culture-specificfeatures of global self-esteem. Journal of Personalityand Social Psychology, 89, 623–642.

Shaver, P. R., Mikulincer, M., Alonso-Arbiol, I., &Lavy, S. (2010). Assessment of adult attachmentacross cultures: Conceptual and methodological con-siderations. In P. Erdman & K. M. Ng (Eds.),Attachment: Expanding the cultural connections (pp.89–108). New York: Routledge.

Shaver, P. R., Schachner, D. A., & Mikulincer, M.(2005). Attachment style, excessive reassurance seek-ing, relationship processes, and depression. Personal-ity and Social Psychology Bulletin, 31, 343–359.

Sherer, M., Maddux, J. E., Mercadante, B., Prentice-Dunn, S., Jacobs, B., & Rogers, R. W. (1982). TheSelf-Efficacy Scale: Construction and validation. Psy-chological Reports, 51, 663–671.

Simpson, J. A., Winterheld, H. A., Rholes, W. S., &Minda Orina, M. (2007). Working models of attach-ment and reactions to different forms of caregivingfrom romantic partners. Journal of Personality andSocial Psychology, 93, 466–477.


Spence, J. T., & Buckner, C. E. (2000). Instrumentaland expressive traits, trait stereotypes, and sex-ist attitudes. Psychology of Women Quarterly, 24,44–62.

Taubman-Ben-Ari, O., Findler, L., & Mikulincer, M.(2002). The effects of mortality salience on relation-ship strivings and beliefs: The moderating role ofattachment style. British Journal of Social Psychol-ogy, 41, 419–441.

van IJzendoorn, M. H., & Sagi-Schwartz, A. (2008).Cross-cultural patterns of attachment: Universal andcontextual dimensions. In J. Cassidy & P. R. Shaver(Eds.), Handbook of attachment: Theory, research,

and clinical applications (2nd ed., pp. 880–905).New York: Guilford.

Wei, M., Russell, D. W., Mallinckrodt, B., & Zakalik,R. A. (2004). Cultural equivalence of adult attach-ment across four ethnic groups: Factor structure,structured means, and associations with negativemoods. Journal of Counseling Psychology, 51,408–417.

Wei, M., Russell, D. W., & Zakalik, R. A. (2005). Adultattachment, social self-efficacy, self-disclosure, lone-liness, and subsequent depression for freshman col-lege students: A longitudinal study. Journal of Coun-seling Psychology, 52, 602–614.

attachment insecurities and interpersonal processes in spanish

Documents