Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2021 Jan 25;47(3):682–697. doi: 10.1111/jmft.12477

Satisfying and stable couple relationships: Attachment similarity across partners can partially buffer the negative effects of attachment insecurity

Henk Jan Conradi 1,✉, Arjen Noordhof 1, Jan H Kamphuis 1
PMCID: PMC8359179  PMID: 33493361

Abstract

Insecure attachment in couples is negatively associated with relationship functioning. Similarity of partner attachment on the other hand might attenuate such relationship outcomes. We tested the opposing insecurity and similarity hypotheses by examining associations of attachment with relationship satisfaction and instability in a representative community sample of couples (N = 1,014). We expected: (a) negative effects of insecure attachment (main effects of actor and partner avoidance and anxiety, and interaction effects consistent with fearful‐avoidance and demand‐withdraw patterns); and (b) positive effects of similarity in avoidance and anxiety. Actor‐Partner‐Interdependence Models showed clear support for the insecurity hypothesis. Main effects of avoidance and anxiety, explained 46.2% of the variance of satisfaction, and 17.9% of instability. We conclude that reducing insecurity of attachment, in particular avoidance, must be a central target in couple therapy. Interestingly, similarity of attachment can partially buffer the negative effects of attachment insecurity. Clinical implications are discussed.

Keywords: APIM, attachment insecurity, attachment similarity, demand‐withdraw, fearful‐avoidant attachment, relationship instability, relationship satisfaction, representative community sample

INTRODUCTION

Satisfying relationships are a source of psychological and somatic well‐being and can serve as a buffer against life's stresses. Dissatisfying relationships on the other hand, induce and amplify distress (Lebow et al., 2012). Over time this can add to relationship instability, defined as thoughts and actions related to breaking up the relationship (e.g., Yeh et al., 2006). Relationship satisfaction and stability are largely dependent on how partners coregulate their emotions, i.e., regulate their own and their partner's emotions (Levenson et al., 2015). A principal and enduring factor in emotion coregulation are attachment strategies of partners (Fraley & Brumbaugh, 2004). Attachment strategies determine whether and how people organize validation and support from their partners when they feel distressed (Mikulincer & Shaver, 2016). Insecurely attached partners, scoring high on avoidance and/or anxiety, are less adequate in emotion regulation than securely attached partners, and they consequently report lower relationship satisfaction (Li & Chan, 2012) and higher instability (Davila & Bradbury, 2001; Duemmler & Kobak, 2001).

The adverse effects of attachment insecurity may be attenuated however by partner similarity, a protective factor emphasized by Gottman in the context of conflict resolution (1999). He hypothesized that partners who are similar in conflict strategies, irrespective of their (dys‐)functionality, have relatively favorable relationship outcomes as they recognize and understand each other's way of regulating emotions. By extension, it can be hypothesized that partners similar in attachment strategies, irrespective of degree of attachment insecurity, have better relationship outcomes than dissimilar partners because they recognize and understand the other's way of regulating emotions. In other words: insecure attachment may be less problematic in partners similar in attachment.

We posit two hypotheses: (a) the insecurity hypothesis predicting that attachment insecurity in a couple determines the level of relationship dissatisfaction and instability; (b) the similarity hypothesis predicting that similarity of attachment between partners has the effect of promoting relationship satisfaction and stability above and beyond the impact of attachment insecurity.

The insecurity hypothesis and its empirical support

The insecurity hypothesis stresses the importance of insecure attachment in couples as the foundation of relationship dysfunction. The more insecurely attached, the less satisfied both partners will be. Securely attached people have the confidence rooted in past experience that attachment needs for validation (being loved), support (receiving help when needed) and consolation (being comforted when distressed) can be satisfied by seeking proximity to their partner and openly expressing these needs. This primary attachment strategy leads to a functional coregulation of emotions that promotes satisfying and stable relationships (Mikulincer & Shaver, 2016). Conversely, insecure attachment, characterized by use of the secondary attachment strategies of hyper‐ and/or deactivation of the attachment system, is associated with relationship dissatisfaction and instability. Hyperactivation develops in individuals who are anxious about rejection and abandonment, and are uncertain whether their attachment needs will be satisfied. In order to minimize the chance of being rejected they employ hyperactivation strategies like persistent appeals for emotional closeness, clinging, and other coercive behavior directed at their partners. When partners do not meet attachment needs, anger may surface, which may lead to relational tension, relationship dissatisfaction, and instability (Mikulincer & Shaver, 2016). Deactivation of attachment behavior develops in individuals who are convinced that partners will neglect or reject proximity bids and will not fulfill attachment needs. Presumably for self‐protection, these individuals tend to deny attachment needs and avoid intimacy with their partners by distancing from them. Consequently, avoidantly attached people prefer to rely on themselves for emotion regulation, resulting in impeded connectedness between partners and dissatisfying and unstable relationships (Mikulincer & Shaver, 2016).

Empirical support for the insecurity hypothesis concerning relationship satisfaction is robust. A large meta‐analysis revealed negative correlations between the anxiety about rejection and avoidance of intimacy dimensions on the one hand, and relationship satisfaction on the other (Li & Chan, 2012). The negative effect of avoidance was observed to be significantly larger than that of anxiety. This meta‐analysis concerned actor effects or the effects of attachment on relationship satisfaction within one individual. Partner effects of attachment, or the effects of the partner's avoidance and anxiety on the actor's relationship satisfaction, are also negative although weaker than actor effects as a recent meta‐analysis shows (Candel & Turliuc, 2019).

Empirical support for the insecurity hypothesis concerning relationship stability is relatively scant and predominantly based on categorical analyses of attachment styles, i.e., participants being assigned to mutually exclusive secure, preoccupied, or dismissing styles. Findings suggest that preoccupied and dismissing styles are associated with more instability as expressed by shorter relationships and a higher likelihood to be divorced (Birnbaum et al., 1997; Duemmler & Kobak, 2001; Feeney & Noller, 1990; Hazan & Shaver, 1987; Kirkpatrick & Davis, 1994; Kirkpatrick & Hazan, 1994). Categorical style measures, however, are problematic in that they are inherently based on somewhat arbitrarily cutoffs, and sacrifice statistical power relative to dimensional analyses.

Interaction effects of avoidance and anxiety within partners (fearful‐avoidance) and between partners (associated with demand‐withdraw dynamics) may worsen relationship outcomes further. Fearful‐avoidance is found in individuals who score high on both avoidance and anxiety, meaning they alternate between hyper‐ and deactivation (Bartholomew & Horowitz, 1991). Fearful‐avoidantly attached individuals have shown vulnerability for many adverse outcomes (e.g., representations of others, psychopathology in general etc.), and are more likely to end up in a violent relationship (Mikulincer & Shaver, 2016). Such results would already be expected on the basis of the summation of the main effects of anxiety and avoidance, but this specific interaction of strategies may amplify negative outcomes.

Demand‐withdraw dynamics are present in many couples seeking therapy and is, as clinicians know well, hard to modify (c.f. Jacobson & Margolin, 1979). In this pattern, one partner demands the other to change by expressing complaints and anger, while the other withdraws. Demand triggers withdrawal and vice versa. Diverging needs for intimacy underlie this dynamic, with the demanding partner urging for more intimacy and the withdrawing partner defending autonomy (Christensen, 1988). The pattern has been linked to combinations of anxious and avoidant partners (Millwood & Waltz, 2008), with anxious partners urging for closeness and avoidant partners seeking independence. It is to be expected that the main effects of anxiety and avoidance will be predictive of such demand‐withdraw patterns, as each of these strategies may well provoke the opposite reaction in a partner. If this specific combination of strategies in a couple produces worse results than each strategy alone, then one might expect interactions between anxiety of one partner and avoidance of the other to be predictive of negative relationship outcome.

Although the demand‐withdraw dynamic is extensively studied in terms of destructive communication (Christensen, 1988), empirical evidence of attachment‐based demand‐withdraw effects on relationship outcomes is surprisingly scarce and still not very well established. Feeney et al. (1994) found that the combination of female attachment anxiety together with male attachment avoidance was associated with lower relationship satisfaction. No empirical research has specifically addressed the demand‐withdrawal impact on relationship instability. However, Doumas et al. (2008) found an interaction effect of female anxiety and male avoidance on partner violence. Another study, examining interpersonal coping styles resembling attachment, found an interaction effect of actor anxious and partner dismissive coping on relational discord but not closeness (Chow et al., 2014). In sum, the scant extant evidence suggests that the anxiety with avoidance interaction has additional detrimental effects on relationship functioning beyond the main effects of avoidance and anxiety.

The similarity hypothesis and its empirical support

The similarity hypothesis states that partners with similar attachment strategies will have more satisfying and stable relationships than partners who are dissimilar. As mentioned, in the context of research on conflict styles, Gottman (1999) speculated that the salience of similarity derives from shared formative experiences that help to understand each other's emotional reactions and needs. Indeed, Murray et al. (2002) found a positive association between perceptions of similarity between partners and their sense of feeling understood and self‐reported relationship satisfaction. Understanding may subsequently prevent escalation (c.f. Gonzaga et al., 2007). Conversely, dissimilarity may breed misunderstanding and impede connection. Together this research (Gonzaga et al., 2007; Gottman, 1999; Murray et al., 2002) suggests that partners sharing similar attachment strategies will better understand each other's emotion regulation and attachment needs which may promote more functional relationships. For example, partners who are similar in avoidance of intimacy may strive for commensurate levels of autonomy, which will prevent protracted disagreements about the optimal balance of autonomy vs. intimacy in the relationship. Likewise, partners who are similar in anxiety about rejection and have comparable needs for validation and support may experience attenuation of dyadic distress caused by one partner feeling pressed to validate or provide support because the other partner complains about insufficiently met emotional needs.

To our knowledge, the attachment similarity hypothesis has not yet been empirically tested regarding relationship functioning. Positive effects of partner similarity have been documented for related constructs however. For example, similarity in terms of the degree of attractiveness of the partners to each other (Luo & Klohnen, 2005), similarity in values (Acitelli et al., 2001) and similarity in personality were all found to be positively associated with relationship satisfaction (e.g. Gonzaga et al., 2007; Russell & Wells, 1991), although Gattis et al. (2004) found no effects for similarity in personality. On the other hand, no effect was found for similarity in self‐control (Vohs et al., 2011), nor for similarity in conflict styles (Busby & Holman, 2009). In short, extant support for the similarity hypothesis in general is mixed, and no research has specifically tested this hypothesis for attachment.

The current study

In sum, this study replicates and extends research on the impact of attachment on relationship satisfaction and instability by testing two hypotheses. (a) We expected negative main actor and partner effects of avoidance and anxiety, and negative interaction effects of fearful‐avoidance and demand‐withdraw dynamics (insecurity hypothesis). (b) Moreover, we tested whether partner similarity might buffer the negative insecurity effects and yield incremental predictive power for relationship outcomes (similarity hypothesis).

METHOD

Participants and procedure

Couples were recruited by Flycatcher, an internet research company originally affiliated with the University of Maastricht. The Flycatcher panel consists of approximately 10,000 members who live all over the Netherlands in urban as well as rural areas. Panel members participate in scientific studies in exchange for a financial compensation. Inclusion criteria were: (a) being 18 years of age or older and (b) having a heterosexual relationship with a duration of at least half a year. These criteria were applied to focus our research on attachment effects in enduring relationships in which interdependency has developed to some extent, such that interaction patterns have crystallized and relationship issues may have surfaced. Homosexual relationships were excluded because of power considerations. As approximately 5% of the population reports being homosexual, we would expect too few homosexual couples for any of the models to be tested. Members of the Flycatcher panel were sampled in order to obtain a Dutch representative community sample of at least 1,000 couples with regard to age and educational attainment. Panel members were requested to involve their partners. This resulted in the current sample of 1,014 couples (see flow chart in online Figure S1).

Sample characteristics are displayed in online Table S1. Participants were on average 52.6 years old (SD = 14.6), with women significantly younger than men (by 2.9 years). Sample composition was largely in line with the representative demographic statistics collected by the Central Bureau of Statistics (CBS) on the Dutch population of 18 years and older. Not surprisingly however, our sample contained relatively fewer people in the category 18–24 years old (3% vs. 11%) and more in the category 60–64 years old (14% vs. 8%) than in the CBS data. More than likely, this deviation was due to the selection criterion of having had a relationship of at least half a year. The distribution of level of educational attainment (low 29.5%, middle 38.2%, high 32.2%) was fairly representative for the Dutch population as reported by CBS (low 29.6%, middle 42.6% and high 27.8%). Mean relationship duration was about 25 years and 80% of the couples were married. Men scored significantly higher on attachment Avoidance than women, whereas women scored higher on Relationship instability. No gender differences were found for attachment Anxiety and Relationship satisfaction.

Measures

Adult attachment

The Experiences in Close Relationships questionnaire (ECR; Brennan et al., 1998) measures adult attachment dimensions in romantic relationships. It comprises two subscales of 18 items each: Anxiety about rejection and abandonment, i.e., the expectation of being perceived by partners as unacceptable or unlovable (example item “I worry about being abandoned”), and Avoidance of intimacy, i.e., the expectation of inaccessibility and unresponsiveness of partners to one's attachment needs (reverse scored example item “I turn to my partner for many things, including comfort and reassurance”). Items are scored on a 7‐point Likert scale ranging from 1 (disagree strongly) through a middle position 4 (neutral/mixed) to 7 (agree strongly). The Dutch translation of the original ECR was found to have favorable psychometric properties in both a community and a student sample (Conradi et al., 2006). Another large representative population sample of Dutch individuals (N = 1,019; Conradi et al., 2018), not to be confused with the sample of couples used in this study, yielded mean item scores of 2.59 for Avoidance and 3.18 for Anxiety, and norms for identification of the top 40% scorers on Avoidance (≥2.78) and Anxiety (≥3.50) needed for the sensitivity analyses described below. High internal consistencies were observed in the present sample, with Cronbach's α = .92 for men and .91 for women for Avoidance, and .88 for both genders for Anxiety.

Attachment dissimilarity scores were calculated by taking the absolute difference between actor and partner Avoidance and actor and partner Anxiety, respectively. Partners are similar on, for example, their anxiety scores, when the difference between their anxiety scores is small irrespective of the magnitude of the individual anxiety scores. Low dissimilarity scores indicate high resemblance between partners and high dissimilarity scores indicate low resemblance (Vohs et al., 2011).

Relationship satisfaction

The Dyadic Adjustment Scale (DAS; Spanier, 1976) was used to assess relationship satisfaction. The total score of the 32‐item DAS measures multiple aspects of relationship satisfaction, including cohesion, affectional expression, consensus/conflict, and satisfaction (example item “Do you confide in your mate?”). Responses are indicated on Likert scales with variable response options, such that a higher total score is indicative of higher satisfaction with the relationship. Scores of 97 or below indicate relationship distress. Construct validity is strong (Spanier, 1976). In the present sample, internal consistencies were high (Cronbach's α = .92, for men and women).

Relationship instability

The Marital Status Inventory (MSI; Weiss & Cerreto, 1980) measures relationship instability in terms of concrete steps taken towards dissolution of the relationship by means of 14 true‐false items (example item “I have made specific plans to discuss separation or divorce with my spouse. I have considered what I would say, etc.”). Research has shown that the MSI is a reliable measure with strong construct and discriminant validity in couples (Crane et al., 1984). Cutoff scores are ≥4 for men and ≥5 for women (Whiting & Crane, 2003). In the present sample, Cronbach's α = .81 for men and .80 for women.

Statistical analyses

Because measurements between partners are not independent, we applied a multi‐level Actor‐Partner Interdependence Model (APIM; Cook & Kenny, 2005) with couple as the unit of analysis, and partners nested within these couples. The multi‐level models were estimated using linear mixed models (for using normal distributions) and generalized linear mixed models (for using negative binomial distributions) as implemented by SPSS version 26. We corrected for within‐dyad interdependence of partner scores by including a random intercept at the couple's level.

First, in order to test the insecurity hypothesis, main effects of actor Avoidance, actor Anxiety, partner Avoidance, and partner Anxiety were included. Second, the interaction terms actor Avoidance*actor Anxiety and partner Avoidance*partner Anxiety were added to explore the specific importance of fearful‐avoidance. Third, the interaction terms actor Avoidance*partner Anxiety and actor Anxiety*partner Avoidance were added to explore demand‐withdraw patterns. In order to test similarity effects, dissimilarity scores were added for both Avoidance and Anxiety. Individual Relationship satisfaction and Relationship instability were the dependent variables in all models. Inspection of the residuals of Relationship satisfaction revealed a normal distribution. Inspection of the residuals of Relationship instability revealed substantial deviance from normality due to a preponderance of zeros (i.e., 76.3% scored zero, indicating they did not take concrete steps towards relationship dissolution). Hence, we assumed a negative binomial distribution rather than a normal distribution for this variable as it can better accommodate highly skewed distributions (see Sellbom et al., 2014). There were no missing data.

To ascertain comparability of coefficients, all independent variables were transformed into z‐scores, as well as the dependent variable Relationship satisfaction (Hox, 2010). As Relationship instability was not normally distributed, we did not standardize these scores. Note however that the relative strength of coefficients is interpretable within the analyses of Relationship instability due to standardization of the independent variables. The interaction terms were based on the z‐scores of Avoidance and Anxiety. The latter are by definition mean‐centered. Explained variance of each model was computed by dividing the variance of the fixed predicted value by the sum of the variances of the fixed predicted value, the random effects and the residuals (Nakagawa & Schielzeth, 2013).

In dyadic analysis distinguishability within dyads needs to be tested. Within couples partners can be distinguished in terms of gender. Hence, we tested whether gender did moderate the influence of each predictor on the outcomes. We evaluated model fit of the tested models (main effects, fearful‐avoidance, demand‐withdraw and similarity) with Akaike's Information Criterion (AIC) and Schwarz's Bayesian Criterion (BIC). Smaller values indicate better fitting models provided that the difference between models is greater than two points BIC (see Raftery, 1995). Adding the main effect of gender, and the interactions of gender with each of the independent variables to the models was evaluated likewise. We tested six models. In model 1 we estimated main actor effects (actor Avoidance and Anxiety). In model 2 we added main partner effects (partner Avoidance and Anxiety) to model 1 in order to estimate to what extent partner effects added surplus predictive value compared to the main actor effects. In model 3 we added both Dissimilarity indices (dissimilarity of Avoidance and Anxiety) to the main actor effects in order to estimate the surplus predictive value of similarity on top of actor Avoidance and Anxiety. In model 4 we extended model 2 by adding both Dissimilarity indices to test for possible surplus effects on top of main actor and partner effects. In model 5 we extended the basic insecurity model, i.e., the main actor and partner effects of model 2, by adding the interaction terms of Fearful‐Avoidance and Demand‐Withdraw. Finally, in model 6 we added both Dissimilarity indices to the extended insecurity model 5.

These analyses were conducted on our community sample of couples. However, because the large majority (approximately 60%) of individuals in community samples are securely attached (Mickelson et al., 1997), we conducted a sensitivity analysis in which we omitted the couples in which both partners were securely attached. Accordingly, we could determine attachment effects in a less secure, more vulnerable subsample. To identify securely attached individuals we used the above‐mentioned cutoff scores. Secure attachment was found in 68% of the partners of the current couples sample, in line with the 59% found in a US general population sample (Mickelson et al., 1997). Of the couples 56.9% was double securely attached. These couples were excluded leaving a less secure subsample of 437 couples (874 partners). For descriptive statistics on socio‐demographic characteristics and study variables of the less secure subsample, see Table S1 (online).

RESULTS

Zero‐order correlations of study variables

Zero‐order correlations between study variables are displayed in Table S2 (online). As was expected, actor and partner Anxiety and Avoidance, the Fearful‐Avoidance and Demand‐Withdraw interaction terms, and both Dissimilarity indices were negatively associated with Relationship satisfaction and positively or nonsignificant with Relationship instability. Correlations between the Fearful‐Avoidance and Demand‐Withdraw interaction terms on the one hand, and both Dissimilarity indices on the other hand, were nonsignificant.

Relationship satisfaction

Table 1 shows the model fit indices and explained variances of all tested models. In general, including gender as main effect resulted in better fitting models (indicating that women reported slightly lower Relationship satisfaction and higher on Relationship instability than men), whereas including the interactions of gender with the independent variables in the models tested resulted in a decrease of model fit. Thus, it was established that dyads could be analyzed as indistinguishable. In all reported models main effects of gender were included, but no gender interaction terms.

TABLE 1.

Relationship Satisfaction and Instability: Model fit criteria and explained variances of tested models

Relationship satisfaction
Total sample Less secure subsample
AIC BIC R 2 AIC BIC R 2
model 1: Actor (actor Avoidance, actor Anxiety) 4330.562 4341.787 35.1% 2110.705 2120.242 23.6%
model 2: Actor‐Partner (model 1 + partner Avoidance, partner Anxiety) 4189.577 4200.801 46.2% 2032.533 2042.066 40.6%
model 3: Actor & Dissimilarity (model 1 + dissimilarity Avoidance, dissimilarity Anxiety) 4323.177 4334.400 36.3% 2115.726 2125.258 24.1%
model 4: Actor‐Partner & Dissimilarity (model 2 + dissimilarity Avoidance, dissimilarity Anxiety) 4193.658 4204.880 46.4% 2030.075 2039.603 41.7%
model 5: Actor‐Partner, Fearful‐Avoidance & Demand‐Withdraw 4213.970 4225.190 46.1% 2053.292 2062.815 40.6%
model 6: Actor‐Partner, Fearful‐Avoidance, Demand‐Withdraw & Dissimilarity 4217.312 4228.530 46.3% 2051.115 2060.634 41.6%
Relationship instability
Total sample Less secure sample
AIC BIC R 2 AIC BIC R 2
model 1: Actor (actor Avoidance, actor Anxiety) 4513,353 4518,964 15.0% 2099.948 2104.712 8.1%
model 2: Actor‐Partner (model 1 + partner Avoidance, partner Anxiety) 4464.575 4470.185 17.9% 2082.643 2087.404 12.0%
model 3: Actor & Dissimilarity (model 1 + dissimilarity Avoidance, dissimilarity Anxiety) 4501.685 4507.295 16.9% 2096.275 2101.037 11.2%
model 4: Actor‐Partner & Dissimilarity (model 2 + dissimilarity Avoidance, dissimilarity Anxiety) 4461.329 4466.938 19.1% 2073.993 2078.752 15.9%
model 5: Actor‐Partner, Fearful‐Avoidance & Demand‐Withdraw 4487.397 4493.005 19.1% 2102.073 2106.830 12.5%
model 6: Actor‐Partner, Fearful‐Avoidance, Demand‐Withdraw & Dissimilarity 4483.678 4489.285 21.0% 2093.337 2098.092 16.0%
Open in a new tab

Actor Avoidance and Anxiety (model 1) explained 35.1% of variance of Relationship satisfaction (Table 1). Adding main partner effects to actor effects (model 2) resulted in a better fitting model, with 46.2% explained variance (ΔR 2 = 11.1%). Adding Dissimilarity effects instead of partner effects resulted in a slight improvement in comparison to model 1 (ΔR 2 = 1.2%), but worse fit than model 2. Adding predictors in subsequent models in all cases resulted in a worsening of model fit (BIC and AIC) and did not improve explained variance substantially (i.e., max ΔR 2 = 0.2%, Table 1). It was concluded that model 2, with only main effects of actor and partner, showed best fit to the data. In this model actor Avoidance showed the largest effect (β = −.44), followed by actor Anxiety (β = −.13), partner Avoidance (β = −.18) and partner Anxiety (β = −.06) (Table 2). An increase of 1 SD on actor Avoidance (i.e., 0.87 points) and actor Anxiety (i.e., 0.96 points) indicates a decrease in Relationship satisfaction (DAS) of 6.89 (i.e., −0.44*15.95 (=SD DAS)) and 2.03 points (i.e., −0.13*15.95), respectively.

TABLE 2.

Relationship Satisfaction and Instability: Actor and Partner Avoidance and Anxiety, Fearful‐Avoidant, Demand‐Withdraw and Similarity Effects of the best fitting models

Relationship satisfaction Relationship instability

Total Sample

model 2

Actor, Partner

Less Secure Subsample

model 4

Actor, Partner

Dissimilarity

Total Sample

model 4

Actor, Partner

Dissimilarity

Less Secure Subsample

model 4

Actor, Partner

Dissimilarity
AIC/BIC 4189.577/4200.801 2030.075/2039.603 4461.329/4466.938 2073.993/2078.752
R2 46.2% 41.7% 19.1% 15.9%
F p β (SE) F p β (SE) t p be (SE) t p be (SE)
Sex 8.69 .003 .07 (0.02) 3.90 .049 .08 (0.04) −4.135 <.001 −.15 (0.04) −2.473 .014 −.14 (0.06)
Actor Avoidance 533.85 <.001 −.44 (0.02) 305.09 <.001 −.46 (0.03) 6.024 <.001 .13 (0.02) 4.941 <.001 .14 (0.03)
Actor Anxiety 49.78 <.001 −.13 (0.02) 31.15 <.001 −.17 (0.03) 3.771 <.001 .08 (0.02) 3.479 .001 .11 (0.03)
Partner Avoidance 94.95 <.001 −.18 (0.02) 70.42 <.001 −.22 (0.03) 3.182 .001 .07 (0.02) 2.738 .006 .08 (0.03)
Partner Anxiety 11.49 .001 −.06 (0.02) 14.69 <.001 −.11 (0.03) 3.534 <.001 .08 (0.02) 3.305 .001 .10 (0.03)
Actor Fearful‐Avoidancea
Partner Fearful‐Avoidanceb
Actor Withdraw‐Partner Demandc
Actor Demand‐Partner Withdrawd
Dissimilarity Avoidance 1.34 .247 −.05 (0.05) 2.410 .016 .07 (0.03) 2.541 <.011 .10 (0.04)
Dissimilarity Anxiety 9.68 .002 −.13 (0.04) 3.222 .001 .09 (0.03) 3.320 <.001 .13 (0.04)
Open in a new tab
a

Actor Avoidance*Actor Anxiety.

b

Partner Avoidance*Partner Anxiety.

c

Actor Avoidance*Partner Anxiety.

d

Actor Anxiety*Partner Avoidance.

e

These coefficients are only standardized with regard to the independent variable.

Next, we conducted a sensitivity analysis. We examined whether the effects we found for the total sample replicated in the less secure subsample. Results were largely similar to those in the complete sample (Table 1). Again, model 2 (actor and partner main effects) showed a better fit and larger explained variance than model 1 with actor effects only (R 2 = 40.6%; ΔR 2 = 17.0%), and most alternative models showed worse fit indices (Table 1). However, model 4 (actor and partner main effects plus the dissimilarity effects) showed a slightly better fit and improved explained variance (R 2 = 41.7%; ΔR 2 = 1.1%). In this model the coefficients of the main effects were very comparable to the complete sample, i.e., actor Avoidance (β = −.46), actor Anxiety (β = −.17), partner Avoidance (β = −.22) and partner Anxiety (β = −.11). Additionally, a significant effect was found for Dissimilarity of Anxiety (β = −.13), while Dissimilarity of Avoidance was nonsignificant (Table 2).

Relationship instability

The main effects of actor attachment (model 1) explained 15.0% variance of Relationship instability (Table 1), which was notably lower than what was found for Relationship satisfaction. Consistent with findings for Relationship satisfaction, adding main partner effects to actor effects (model 2) resulted in a better fitting model with increased explained variance (R 2 = 17.9%; ΔR 2 = 2.9%). However, adding the Dissimilarity indices (model 4) to the main actor and partner effects model (model 2) improved fit and explained variance (R 2 = 19.1%; ΔR 2 = 1.2%) for the prediction of Relationship instability. Adding the four Anxiety*Avoidance interaction terms resulted in a worse model fit, but increased explained variance (model 6; R 2 = 21.0%; ΔR 2 = 1.9%). Inspection of the model showed that none of the interaction terms were significantly predictive. Therefore, model 4 was judged to be the most adequate model (Table 1). In this model, actor Avoidance showed the largest effect (b = .13), followed by actor Anxiety (b = .08), partner Avoidance (b = .07), partner Anxiety (b = .08) Dissimilarity of Avoidance (b = .07) and Dissimilarity of Anxiety (b = .09) (Table 2).

As before, we conducted a sensitivity analysis in the subsample of less secure partners. The results for main actor and partner effects and dissimilarity were comparable to those in the full sample. Main actor effects of attachment (model 1) explained 8.1% of the variance (Table 1). Adding main partner effects (model 2) resulted in a better fit and explained variance (R 2 = 12.0%; ΔR 2 = 3.9%). Adding Dissimilarity indices also improved model fit and explained variance (model 4; R 2 = 15.9%; ΔR 2 vs. model 2 = 3.9%; ΔR 2 vs. model 3 = 4.7%). Adding the four Anxiety*Avoidance interaction terms to this model clearly worsened model fit and did not improve explained variance (ΔR 2 = 0.1%, Table 1). In the best fitting model (model 4) actor Avoidance had a substantial effect (b = .14), and other significant effects were observed for actor Anxiety (b = .11), partner Avoidance (b = .08), partner Anxiety (b = .10), Dissimilarity of Avoidance (b = .10), and Dissimilarity of Anxiety (b = .13) (Table 2).

DISCUSSION

In a large representative sample of Dutch couples, we examined whether attachment similarity predicted relationship satisfaction and instability beyond the impact of joint attachment insecurity. The main findings can be summarized as follows. Clear support was found for the insecurity hypothesis, with the main actor and partner effects of avoidance and anxiety replicating over both samples and outcomes. Findings with regard to the similarity hypothesis concerning with relationship satisfaction were inconsistent. Interactions between anxiety and avoidance did not influence relationship satisfaction. Attachment similarity did influence instability of the relationship.

Insecurity effects

In line with earlier studies, insecure actor and partner attachment were negatively associated with relationship satisfaction (Candel & Turliuc, 2019; Li & Chan, 2012) and positively with relationship instability (e.g., Birnbaum et al., 1997; Duemmler & Kobak, 2001). Of note, actor and partner insecurity effects were predominant after fearful‐avoidance, demand‐withdraw, and attachment similarity were taken into account. Another noteworthy finding was that attachment insecurity explained approximately 2.5 times more variance of relationship satisfaction than of relationship instability. This stands to reason as thinking about actually breaking up of the relationship will also involve considerations other than attachment‐related determinants, such as consequences for children, finances, housing arrangements, etc. Actor avoidance had the strongest negative association with relationship satisfaction. The negative impact of avoidance on relationship satisfaction and stability likely hinges on the progressive erosion of connectedness between partners caused by denial of attachment needs (Mikulincer & Shaver, 2016). Actor anxiety may be less detrimental than actor avoidance because anxiously attached partners, in contrast to avoidant partners, try to satisfy their attachment needs in connection with their partners, albeit in suboptimal ways (e.g., by clinging, claiming, and sometimes blaming their partner).

Whereas actor effects of attachment concern one's own way of asking for support, validation and consolation, partner effects of attachment concern whether and how partners help to satisfy the actors’ attachment needs. Partner avoidance showed negative associations with both relationship outcomes. Avoidant partners tend to be unavailable and unresponsive to the attachment needs of the other, enhancing dissatisfaction in the actor and adding to destabilization of the relationship. Again, partner anxiety was less detrimental than partner avoidance, presumably because they tend to be available to their partner. However, their responsiveness to their partner's needs is suboptimal because they are preoccupied with their own attachment needs which adds to the actor's dissatisfaction and instability of the relationship (Mikulincer & Shaver, 2016).

Interaction effects between anxiety and avoidance were absent for satisfaction and instability. It warrants mentioning that one cannot conclude from these null findings that fearful‐avoidance and demand‐withdraw patterns are irrelevant to relationship dynamics or outcomes. For example, main effects of anxiety and avoidance may predict demand‐withdraw patterns in relations that are harmful to satisfaction and stability. The absence of interaction effects merely indicates that there are no additional multiplicative effects (i.e., actor anxiety*actor avoidance, partner anxiety*partner avoidance and actor avoidance*partner anxiety).

Similarity effects

Attachment similarity predicted relationship stability, and inconsistently relationship satisfaction. Apparently, mutual understanding of a partner's blaming or withdrawal does not help much to feel more satisfied with relationship functioning, but it may help to accept insecure attachment behavior. It bears mentioning that these findings concerning instability were consistent across the full sample and less secure subsample, and that these attenuating similarity effects were found on top of the main effects of attachment insecurity, i.e., after the destructive main effects of actor and partner avoidance and anxiety were already accounted for. As was mentioned before, partners who resemble each other may share similar formative experiences and may better understand “where the other is coming from” (c.f. Gonzaga et al., 2007), which may facilitate acceptance. Additionally, partners similar in avoidance and anxiety probably share similar relational aspirations. Partners similar in anxiety strive for comparable levels of closeness, thus foregoing relationship instability because of escalating dynamics related to strong claiming behaviors in one partner that elicit enhanced autonomy needs in the less anxious partner, and vice versa. Likewise, when two partners are similar on avoidance or preferred degree of autonomy, relationship instability will not be fueled by heated autonomy vs. closeness discussions, not even when both partners opt for parallel lives. On the other hand, the strong main effect of actor avoidance implies that such distancing will come at the cost of weaker connectedness between partners and lower relationship satisfaction (Mikulincer & Shaver, 2016). In other words, our findings are in line with the notion that in some cases avoidant partners might form relatively stable, but less happy relationships.

Limitations and research implications

This study is not without limitations. First, inherent to dyadic analyses, both partners need to participate. This prerequisite may have resulted in a selection bias towards lengthier and happier relationships. However, mean relationship satisfaction was fairly similar to the satisfaction reported by married couples in the original validation study of the DAS (Spanier, 1976). Also, the explained variances by the main effects model in the less secure sample were comparable to those of the inclusive representative sample. Nevertheless, we recommend future research incorporate substantial (sub‐)clinical samples. Furthermore, this study focused on the relationship satisfaction and instability outcomes. Other relationship outcomes, such as relationship maintenance behaviors, communication patterns also deserve investigation because of their inherent clinical relevance. Finally, the cross‐sectional nature of the current study precludes definite conclusions regarding the direction of the associations between attachment and relationship functioning. Attachment insecurity may cause dissatisfaction with the relationship as well as relationship instability, but conversely, relationship problems may amplify avoidance of intimacy or anxiety about rejection just as well. Longitudinal designs are needed to more definitively determine the (bi‐)directionality of associations.

Clinical implications

This study documented that attachment‐based couple dynamics are a predominant determining force in relationship satisfaction (in fact explaining almost half of its variance). Accordingly, it stands to reason that intervening in these crucial patterns constitutes a fertile approach in couple therapy. Emotionally Focused Couple Therapy (EFCT) is aimed at replacing insecure attachment strategies, i.e., de‐ and hyperactivation of the attachment system, with the secure primary attachment strategy (Johnson, 2004). EFCT’s intervention model guides therapists through attachment change processes in several steps (Johnson, 2004). First, couples are urged to de‐escalate. This means that partners go beyond the apparent content of their disagreement, e.g., household tasks, upbringing of the children, and learn that secondary attachment strategies, blaming and claiming or distancing, trigger insecure attachment behavior in the other partner contributing to escalation. We will illustrate these dynamics with a clinical case example.

Consider the case of Mary who is anxiously attached, living with John, who is avoidantly attached. John's proclivity to distance himself may make Mary feel unseen by John, and trigger insecure cognitions like “My partner does not love me”, thus amplifying her anxiety about rejection and abandonment. Mary estimates that openly expressing her fear of rejection and need for approval, i.e., the primary attachment strategy, is too risky because of her previous experiences that proximity bids can be followed by rejection. Therefore, Mary applies a secondary attachment strategy and blames John for not validating her. Although blaming leaves her less vulnerable, the effect it has on John is problematic. John feels criticized, which triggers his core cognition, “I can never live up to Mary's expectations”. Again, instead of sharing this thought and the feelings of insecurity it triggers, i.e., the primary attachment strategy, he reacts by deactivating his attachment system in order to avoid the painful emotions, thus making Mary more anxious. It is important that both partners recognize the destructive effect their insecure attachment strategy has on the other partner: distancing by John fuels attachment fears in Mary which are expressed by blaming, and blaming fuels further distancing, etc. Recognizing and exploring this spiral may function as a vehicle for mutual understanding between partners and assist the couple to stop the escalation.

Next, in order to flip the spiral, partners are encouraged to refrain from using their secondary attachment strategies, and instead to apply the primary attachment strategy. This means the EFCT therapist has to re‐engage the withdrawing partner, in our case example John. As can be seen from our findings this step is essential as actor avoidance explains a large part of relationship dissatisfaction. This step is also challenging because the therapist first has to turn on John's deactivated attachment system and aid him in discovering his attachment‐based fears by upregulating them and to reveal his attachment needs. In case of John this concerns his fear of not living up to Mary's expectations and his need for acceptance by Mary. Hence, the therapist encourages John to disclose these fears and concerns to Mary. It is essential for Mary to understand that John's withdrawal is caused by his thoughts of not living up to her expectations instead of a plain desire to avoid intimacy which would immediately trigger her fear of rejection (Johnson, 2004). This task becomes easier when Mary is able to acknowledge that her blaming triggers his withdrawal: “When I blame you, you think you have failed and you want to avoid the feelings this triggers inside of you.” Only when Mary acts more secure, John will not relapse immediately into withdrawal.

In similar vein, it is essential for John to understand that Mary's blaming is a disguised expression of her need for connection. With this in mind, it will be easier for John to stay available and responsive, the two building blocks of secure attachment (Ainsworth et al.,1978), to Mary's attempts to connect (Johnson, 2004). For Mary, developing trust that John will be available is crucial to downregulate her hyperactivated attachment system. Subsequently, she can exercise the primary attachment strategy by seeking proximity and openly expressing her anxiety about being rejected and her need for his approval. In short, only when the withdrawer re‐engages (John) can the blamer be expected to soften (Mary), and vice versa. These specific change processes are identified and explained to the couple to enhance secure attachment and relationship satisfaction (Dalgleish et al., 2015; Johnson & Greenberg, 1988).

EFCT’s core intervention of replacing secondary attachment strategies by the primary strategy, will work in favor of the two effects we found in this study. First and foremost, EFCT targets the negative main effects of actor and partner avoidance and anxiety. Second, the reduction of insecurity may often result in more similarity between both partners, and we found similarity to partially buffer the negative main effects of attachment insecurity on relationship instability and, for less secure couples, relationship satisfaction.

In clinical practice, the balance between the buffering effect of attachment similarity and the detrimental effect of insecure attachment presumably depends on the couple's specific combination and severity of their attachment strategies and the therapeutic modification thereof over the course of therapy. In couples with a dissimilar attachment constellation (i.e., pairings of an anxiously attached partner with an avoidantly attached partner, an anxiously or avoidantly attached partner with a securely attached partner, or two anxiously or avoidantly attached partners with dissimilar severity levels) there will be no buffering similarity effect. Reduction of insecurity will be the focus of treatment. This is also the case with partners displaying similar severe insecure attachment. However, in couples with similar mild insecure attachment, which may be achieved over the course of therapy, the negative effects of residual insecure attachment may be buffered by the positive effects of attachment similarity. Such couples may be stable and relatively satisfied. Their emotion coregulation may include limited hyperactivation and reassurance seeking, or alternatively limited deactivating emotional dynamics, but these patterns may be experienced as sufficiently functional. In these cases the therapist and the couple may decide to re‐focus treatment from a sustained attachment‐focused approach to other relational issues, or to end therapy altogether. Finally, it should be mentioned that instability is only modestly associated with insecure attachment, implying that therapists need to explore other potential causes of break up and other targets for interventions, such as discussing conflicts about the upbringing of children, spending more time together to discover (new) shared interests, the sexual relation, etc.

Conclusion

Attachment insecurity was shown to be highly detrimental to relationship satisfaction, and to pose a threat to relationship stability, albeit to a lesser extent. Avoidant attachment was found to be especially harmful to relationship outcome. Hence, treatment should prioritize the substitution of insecure secondary attachment strategies for the secure primary attachment strategy. An interesting finding of this study was that attachment similarity may attenuate the negative effects of insecure attachment. This finding suggests that it may not always be necessary to change mild residual but similar attachment problems.

Supporting information

Supplementary Material

Click here for additional data file. (65.2KB, docx)

REFERENCES

  1. Acitelli, L. K., Kenny, D. A., & Weiner, D. (2001). The importance of similarity and understanding of partners’ marital ideals to relationship satisfaction. Personal Relationships, 8(2), 167–185. 10.1111/j.1475-6811.2001.tb00034.x. [DOI] [Google Scholar]
  2. Ainsworth, M.D.S., Blehar, M.C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the strange situation. Lawrence Erlbaum. [Google Scholar]
  3. Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four‐category model. Journal of Personality and Social Psychology, 61(2), 226–244. 10.1037/0022-3514.61.2.226. [DOI] [PubMed] [Google Scholar]
  4. Birnbaum, G. E., Orr, I., Mikulincer, M., & Florian, V. (1997). When marriage breaks up: Does attachment style contribute to coping and mental health? Journal of Social and Personal Relationships, 14(5), 643–654. 10.1177/0265407597145004. [DOI] [Google Scholar]
  5. Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self‐report measurement of adult attachment: an integrative overview. In Simpson J. A. & Rholes W. S. (Eds.), Attachment Theory and Close Relationships (pp. 46–76). The Guilford Press. [Google Scholar]
  6. Busby, D. M., & Holman, T. B. (2009). Perceived match or mismatch on the Gottman conflict styles: Associations with relationship outcome variables. Family Process, 48(4), 531–545. 10.1111/j.1545-5300.2009.01300.x. [DOI] [PubMed] [Google Scholar]
  7. Candel, O. S., & Turliuc, M. N. (2019). Insecure attachment and relationship satisfaction: A meta‐analysis of actor and partner associations. Personality and Individual Differences, 147, 190–199. 10.1016/j.paid.2019.04.037. [DOI] [Google Scholar]
  8. Chow, C. M., Buhrmester, D., & Tan, C. C. (2014). Interpersonal coping styles and couple relationship quality: Similarity versus complementarity hypotheses. European Journal of Social Psychology, 44(2), 175–186. 10.1002/ejsp.2000. [DOI] [Google Scholar]
  9. Christensen, A. (1988). Dysfunctional interaction patterns in couples. In Noller P. & Fitzpatrick M. A. (Eds.), Perspectives on marital interaction (pp. 31–52). Multilingual Matters LTD. [Google Scholar]
  10. Conradi, H. J., Dingemanse, P., Noordhof, A., Finkenauer, C., & Kamphuis, J. H. (2018). Effectiveness of the ‘Hold me Tight’ relationship enhancement program in a self‐referred and a clinician‐referred sample: An Emotionally Focused Couples Therapy‐based approach. Family Process, 57(3), 613–628. 10.1111/famp.12305. [DOI] [PubMed] [Google Scholar]
  11. Conradi, H. J., Gerlsma, C., van Duijn, M., & de Jonge, P. (2006). Internal and external validity of the Experiences in Close Relationships questionnaire in an American and two Dutch samples. European Journal of Psychiatry, 20, 258–269. [Google Scholar]
  12. Cook, W. L., & Kenny, D. A. (2005). The Actor‐Partner Interdependence Model: A model of bidirectional effects in developmental studies. International Journal of Behavioral Development, 29(2), 101–109. 10.1080/01650250444000405. [DOI] [Google Scholar]
  13. Crane, D. R., Newfield, N., & Armstrong, D. (1984). Predicting divorce at marital therapy intake: Wives’ distress and the marital status inventory. Journal of Marital and Family Therapy, 10(3), 305–312. 10.1111/j.1752-0606.1984.tb00021.x. [DOI] [Google Scholar]
  14. Dalgleish, T. L., Johnson, S. M., Moser, M. B., Wiebe, S. A., & Tasca, G. A. (2015). Predicting key change events in emotionally focused couple therapy. Journal of Marital and Family Therapy, 41(3), 260–275. 10.1111/jmft.12101. [DOI] [PubMed] [Google Scholar]
  15. Davila, J., & Bradbury, T. N. (2001). Attachment insecurity and the distinction between unhappy spouses who do and do not divorce. Journal of Family Psychology, 15(3), 371–393. 10.1037/0893-3200.15.3.371. [DOI] [PubMed] [Google Scholar]
  16. Doumas, D. M., Pearson, C. L., Elgin, J. E., & McKinley, J. L. (2008). Adult attachment as a risk factor for intimate partner violence: the "mispairing" of partners’ attachment styles. Journal of Interpersonal Violence, 23(5), 616–634. 10.1177/0886260507313526. [DOI] [PubMed] [Google Scholar]
  17. Duemmler, S., & Kobak, R. (2001). The development of attachment and commitment in dating relationships: Attachment security as a relationship construct. Journal of Adolescence, 24, 401–415. 10.1006/jado.2001.0406. [DOI] [PubMed] [Google Scholar]
  18. Feeney, J. A., & Noller, P. (1990). Attachment style as a predictor of adult romantic relationships. Journal of Personality and Social Psychology, 58, 281–291. 10.1037/0022-3514.58.2.281. [DOI] [Google Scholar]
  19. Feeney, J.A., Noller, P., & Callan, V.J. (1994). Attachment style, communication and satisfaction in the early years of marriage. In Bartholomew K. & Perlman D. (Eds.), Advances in personal relationships, Vol. 5 Attachment processes in adulthood, (pp. 269–308). Jessica Kingsley Publishers. [Google Scholar]
  20. Fraley, R. C., & Brumbaugh, C. C. (2004). A dynamical systems approach to understanding stability and change in attachment security. In Rholes W. S. & Simpson J. A. (Eds.), Adult attachment: Theory, research, and clinical implications (pp. 86–132). Guilford Press. [Google Scholar]
  21. Gattis, K. S., Berns, S., Simpson, L. E., & Christensen, A. (2004). Birds of a feather or strange birds? Ties among personality dimensions, similarity, and marital quality. Journal of Family Psychology, 18(4), 564–574. 10.1037/0893-3200.18.4.564. [DOI] [PubMed] [Google Scholar]
  22. Gonzaga, G. C., Campos, B., & Bradbury, T. (2007). Similarity, convergence, and relationship satisfaction in dating and married couples. Journal of Personality and Social Psychology, 93(1), 34–48. 10.1037/0022-3514.93.1.34. [DOI] [PubMed] [Google Scholar]
  23. Gottman, J. M. (1999). The marriage clinic: A scientifically based marital therapy. W.W. Norton. [Google Scholar]
  24. Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52(3), 511–524. 10.1037/0022-3514.52.3.511. [DOI] [PubMed] [Google Scholar]
  25. Hox, J. (2010). Multilevel analysis: Techniques and applications. Routledge. [Google Scholar]
  26. Jacobson, N. S., & Margolin, G. (1979). Marital therapy: Strategies based on social learning and behavior exchange principles. Brunner/Mazel. [Google Scholar]
  27. Johnson, S. (2004). The practice of emotionally focused couple therapy (1st/2nd edn). Brunner‐Routledge. [Google Scholar]
  28. Johnson, S.M., & Greenberg, L.S. (1988). Relating process to outcome in marital therapy. Journal of Marital and Family Therapy, 14(2), 175–183. [Google Scholar]
  29. Kirkpatrick, L. A., & Davis, K. E. (1994). Attachment style, gender, and relationship stability: A longitudinal analysis. Journal of Personality and Social Psychology, 66(3), 502–512. 10.1037/0022-3514.66.3.502. [DOI] [PubMed] [Google Scholar]
  30. Kirkpatrick, L. A., & Hazan, C. (1994). Attachment styles and close relationships: A four‐year prospective study. Personal Relationships, 1(2), 123–142. 10.1111/j.1475-6811.1994.tb00058.x. [DOI] [Google Scholar]
  31. Lebow, J. L., Chambers, A. L., Christensen, A., & Johnson, S. M. (2012). Research on the treatment of couple distress. Journal of Marital and Family Therapy, 38, 145–168. 10.1111/j.1752-0606.2011.00249.x. [DOI] [PubMed] [Google Scholar]
  32. Levenson, R. W., Haase, C. M., Bloch, L., Holley, S. R., & Seider, B. J. (2015). Emotion regulation in couples. In Gross J. J. (Ed.), Handbook of Emotion Regulation (pp. 267–283). Guilford Press. [Google Scholar]
  33. Li, T., & Chan, D. K. S. (2012). How anxious and avoidant attachment affect romantic relationship quality differently: A meta‐analytic review. European Journal of Social Psychology, 42(4), 406–419. 10.1002/ejsp.1842. [DOI] [Google Scholar]
  34. Luo, S., & Klohnen, E. C. (2005). Assortative mating and marital quality in newlyweds: A couple‐centered approach. Journal of Personality and Social Psychology, 88, 306–326. 10.1037/0022-3514.88.2.304. [DOI] [PubMed] [Google Scholar]
  35. Mickelson, K. D., Kessler, R. C., & Shaver, PHR (1997). Adult attachment in a nationally representative sample. Journal of Personality and Social Psychology, 73(5), 1092–1106. 10.1037/0022-3514.73.5.1092. [DOI] [PubMed] [Google Scholar]
  36. Mikulincer, M., & Shaver, PHR (2016). Attachment in adulthood: Structure, dynamics and change (2nd edn). The Guilford Press. [Google Scholar]
  37. Millwood, M., & Waltz, J. (2008). Demand‐withdraw communication in couples: An attachment perspective. Journal of Couple & Relationship Therapy, 7(4), 297–320. 10.1080/15332690802368287. [DOI] [Google Scholar]
  38. Murray, S. L., Holmes, J. G., Bellavia, G., Griffin, D. W., & Dolderman, D. (2002). Kindred spirits? The benefits of egocentrism in close relationships. Journal of Personality and Social Psychology, 82, 563–581. 10.1037/0022-3514.82.4.563. [DOI] [PubMed] [Google Scholar]
  39. Nakagawa, S., & Schielzeth, H. (2013). A general and simple method for obtaining R 2 from generalized linear mixed‐effects models. Methods in Ecology and Evolution, 4, 133–142. 10.1111/j.2041-210x.2012.00261.x. [DOI] [Google Scholar]
  40. Raftery, A. E. (1995). Bayesian model selection in social research. Sociological Methodology, 25, 111–163. 10.2307/271063 [DOI] [Google Scholar]
  41. Russell, R. J. H., & Wells, P. A. (1991). Personality similarity and quality of marriage. Personality and Individual Differences, 12, 407–412. 10.1016/0191-8869(91)90057-I. [DOI] [Google Scholar]
  42. Sellbom, M., Smid, W., de Saeger, H., Smit, N., & Kamphuis, J. H. (2014). Mapping the Personality Psychopathology Five domains onto DSM‐IV personality disorders in Dutch clinical and forensic samples: implications for DSM‐5. Journal of Personality Assessment, 96(2), 185–191. [DOI] [PubMed] [Google Scholar]
  43. Spanier, G. B. (1976). Measuring dyadic adjustment: New scales for assessing the quality of marriage and similar dyads. Journal of Marriage and the Family, 38(1), 15–28. 10.2307/350547. [DOI] [Google Scholar]
  44. Vohs, K. D., Finkenauer, C., & Baumeister, R. F. (2011). The sum of friends’ and lovers’ self‐control scores predicts relationship quality. Social Psychological and Personality Science, 2(2), 138–145. 10.1177/1948550610385710. [DOI] [Google Scholar]
  45. Weiss, R. L., & Cerreto, M. C. (1980). The Marital Status Inventory: Development of a measure of dissolution potential. American Journal of Family Therapy, 8(2), 80–85. 10.1080/01926188008250358. [DOI] [Google Scholar]
  46. Whiting, J. B., & Crane, D. R. (2003). Distress and divorce: Establishing cutoff scores for the Marital Status Inventory. Contemporary Family Therapy: an International Journal, 25(2), 195–205. 10.1023/A:1023672101157. [DOI] [Google Scholar]
  47. Yeh, H., Lorenz, F., Wickrama, K., Conger, R., & Elder, G. (2006). Relationships among sexual satisfaction, marital quality, and marital instability at midlife. Journal of Family Psychology, 20, 339–343. 10.1037/0893-3200.20.2.339. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material

Click here for additional data file. (65.2KB, docx)

Articles from Journal of Marital and Family Therapy are provided here courtesy of Wiley

RESOURCES