Perceptions of oneself and one’s spouse following a stressor discussion predicting attachment insecurity over one year

Abstract

The Attachment Security Enhancement Model (ASEM) posits that attachment anxiety decreases when people revise their negative working model of self, whereas attachment avoidance decreases when people revise their negative working model of others. These revisions are expected to occur in diagnostic situations that provide relevant information about the self and close others. Guided by this theory, the current research assessed whether state perceptions of oneself (state self-efficacy, state self-esteem) and of one’s spouse (state perceived partner responsiveness [PPR]) following a personal stressor discussion provide an impetus for changing trait-level perceptions and decreasing the relevant attachment insecurity dimension over time. In a sample of 164 newly married couples, we found that people who reported greater post-discussion state self-efficacy and state self-esteem reported greater increases in trait self-efficacy and trait self-esteem over the following year. These changes to the model of self in turn predicted greater decreases in attachment anxiety over that year, consistent with the ASEM. Regarding perceptions of one’s spouse, neither state nor trait PPR predicted declines in attachment avoidance, and state PPR did not predict changes in trait level perceptions. Finally, we observed some ASEM-inconsistent changes in attachment insecurity: Increases in trait PPR predicted declines in attachment anxiety, and increases in trait self-esteem and self-efficacy predicted declines in attachment avoidance. These findings suggest that personal stressor discussions may catalyze changes in trait perceptions and attachment change. Additionally, this work shows that improving the model of self and the model of others promotes attachment security but that the processes for reducing attachment anxiety and avoidance may not be entirely separable.

Keywords

attachment attachment security enhancement model (ASEM)close relationships perceived partner responsiveness self-esteem self-efficacy

Introduction

People with more insecure attachment orientations report poorer physical health (e.g., Pietromonaco & Beck, 2019), more frequent and destructive conflict with romantic partners (e.g., J. Feeney & Fitzgerald, 2019), and lower overall relationship satisfaction (e.g., Candel & Turliuc, 2019) than people who report less attachment insecurity (i.e., more attachment security). Given these disparities in personal and relational functioning, relationship researchers are motivated to understand how the two forms of attachment insecurity—attachment anxiety and attachment avoidance—decrease.

Attachment anxiety is characterized by persistent worries about relationships (e.g., doubts about one’s own lovability; concerns about being abandoned) and efforts to mitigate perceived relationship threats, whereas attachment avoidance is characterized by discomfort depending on others and efforts to limit intimacy and remain self-reliant (e.g., Fraley & Shaver, 2000; Mikulincer & Shaver, 2016). The recently proposed Attachment Security Enhancement Model (ASEM; Arriaga et al., 2018) specifies distinct pathways through which attachment anxiety and attachment avoidance can decline over time to foster attachment security. The current research tests propositions of the ASEM and adds to a growing list of work supporting and extending its theoretically-grounded framework of attachment change (e.g., Arriaga et al., 2021; Bayraktaroglu et al., 2022; Gazder & Stanton, 2022; Jakubiak et al., 2022; Rholes et al., 2021).

The Attachment Security Enhancement Model

Central to the ASEM is the idea that enhancing attachment security requires revising the insecure (i.e., negative) working models that underlie each form of attachment insecurity (Arriaga et al., 2018). These working models (i.e., schemas and scripts that guide expectations and beliefs about close relationships) form in early relationships with attachment figures and are continually—though very gradually—revised over time (Bowlby, 1980). The ASEM proposes separate pathways to reduce attachment anxiety and attachment avoidance that involve altering the negative beliefs people hold about themselves and about others, respectively.

Attachment anxiety is based on beliefs that one is unlovable and incapable of receiving others’ love and care (insecure model of self; Bowlby, 1980). These beliefs motivate the hyperactivating strategies of worrying, reassurance-seeking, relationship monitoring, and over-dependence on one’s partner (all to prevent a feared abandonment) that characterize attachment anxiety (e.g., Arriaga et al., 2018; B. C. Feeney et al., 2015). Therefore, the ASEM stipulates that attachment anxiety will decline when the insecure model of self abates and people come to perceive themselves as capable and valuable in personal domains.

If this proposition is accurate, we should observe that attachment anxiety decreases when one develops 1) self-efficacy (positive evaluations of one’s own general competence; Mischel & Shoda, 1995) and 2) self-esteem (positive evaluations of the self in general; Rosenberg, 1965). A person who feels efficacious and develops a positive self-evaluation will gradually have less reason to worry about abandonment and to protectively monitor close relationships. Specifically, people who feel capable of bringing about the outcomes that they desire and who recognize their value should be more confident that they can retain their partner’s interest and care without hypervigilance. There is initial support for this model of attachment anxiety change: new parents experienced greater decreases in attachment anxiety at times when they felt more self-efficacious in their parental role (i.e., felt more competent and accomplished as parents; Arriaga et al., 2021). To our knowledge, the link between increases in self-esteem and declines in attachment anxiety has not been tested.

Attachment avoidance is based on the belief that others will be unreliable, unavailable, and unresponsive to one’s needs (insecure model of others; Bowlby, 1980). This belief motivates the hypo-activating strategy of excessive self-reliance that characterizes attachment avoidance (e.g., Arriaga et al., 2018; B. C. Feeney et al., 2015). Therefore, the ASEM proposes that attachment avoidance will decline when the insecure model of others abates and people come to perceive close others are reliable and responsive. If this proposition is accurate, we should observe that attachment avoidance decreases when one develops perceived partner responsiveness (PPR; the perception that others are caring, validating, and understanding; Crasta et al., 2021). People who evaluate close others (e.g., their spouse) as reliable and responsive should gradually become more comfortable opening up to and relying on others.

There is some extant support for this model of attachment avoidance change as well. People who trusted their romantic partner more (expected them to be predictable, dependable, and faithful) experienced greater declines in attachment avoidance over time than people who trusted their partners less (Arriaga et al., 2014). Additionally, people experienced greater declines in attachment avoidance at times when they perceived more support than usual and at times when they themselves behaved in ways that were more responsive (Bayraktaroglu et al., 2022; Rholes et al., 2021). To our knowledge, no extant research has assessed the link between increases in PPR and declines in attachment avoidance directly.

The current research aims to provide additional support for the ASEM by assessing whether change in these three key variables (self-efficacy, self-esteem, and PPR) predicts declines in the relevant form of attachment insecurity over time.

Shifting Perceptions of Self and Others

Despite the promise of the ASEM and initial support for its propositions, exactly how people’s trait-like perceptions of themselves and others change remains unknown. In other words, when (in what contexts) do trait self-efficacy, self-esteem, and perceived partner responsiveness increase so that attachment insecurity can decrease? In describing the ASEM, Arriaga and colleagues (2018) proposed that insecure models of self and others are revised (or reinforced) in diagnostic situations. These are situations that a) activate the attachment system (i.e., are threatening or distressing) and b) provide diagnostic information about oneself (e.g., one’s capability or value) or others (e.g., a partners’ true motives or reliability). One diagnostic situation that has the potential to catalyze changes in self- and partner-perceptions—and the focus of the current investigation—is a discussion about a personal stressor.

Personal stressors activate the attachment system and often motivate people to seek comfort from close others (e.g., Collins & Feeney, 2000). Moreover, discussing personal stressors (perhaps uniquely) provides information related to the working models of self and other. By their very definition, personal stressors involve challenges that an individual needs to cope with and/or resolve independently, so they provide an opportunity to develop perceptions of one’s own self-worth and efficacy (e.g., Jayamaha & Overall, 2019). Additionally, sharing a personal stressor with another person provides an opportunity for the other person to demonstrate understanding, validation, and care (PPR). For instance, if Annie is stressed about failing the bar exam for a second time (an actual stressor discussed by a couple in the current research), the resulting discussion with her partner could make her feel a) capable or incapable of managing this stressor (e.g., eventually passing, managing her distress effectively), b) better or worse about herself, and c) more or less that her partner is understanding and caring. These state perceptions—particularly if they challenge typically held beliefs—can provide an impetus for questioning and revising trait perceptions of oneself and others. These state perceptions could also motivate new behavior (e.g., goal pursuit resulting from state self-efficacy, or a willingness to depend on a partner resulting from state PPR) that provides an opportunity for state perceptions to be reinforced and generalized.

Despite the ASEM’s suggestion that immediate outcomes of specific diagnostic situations are important for reducing attachment insecurity, recent attachment insecurity reduction research (e.g., Arriaga et al., 2021; Rholes et al., 2021) has focused on changes in broader trait-level perceptions of self and others without addressing the immediate consequences of specific diagnostic situations. To address this gap, the current research tests whether immediate (i.e., state) self-related and other-related perceptions following stressor discussions between spouses predict declines in attachment anxiety and attachment avoidance, respectively. We expect that state perceptions are influential because they shift overarching trait self-related and other-related perceptions. In other words, we expect attachment insecurity to decline when people experience state self-efficacy, self-esteem, and PPR following specific diagnostic situations because these state perceptions enable people to revise their trait-like working models of self and others (trait self-efficacy, self-esteem, and PPR). Consistent with this idea, Bowlby (1980) wrote that the models of self and others are initially formed based on an accumulation of interactions with close others. Thus, to change insecure working models, one must acquire evidence that counters negative perceptions of the self or of others in specific situations and revise trait-like perceptions of the self and others accordingly.

Overview of the Current Study and Hypotheses

In this study, newly married couples discussed personal (relationship-external) stressors and reported their state self-perceptions (state self-efficacy, self-esteem) and state partner-perceptions (state PPR immediately following those discussions. We assessed whether these state perceptions predict changes in attachment insecurity over the following year. Further, to assess a more comprehensive process of attachment insecurity reduction, we assessed whether state perceptions following a stressor discussion predict declines in attachment insecurity indirectly through changes in trait perceptions. We pre-registered several specific predictions consistent with the ASEM (see https://osf.io/pbzjr).

First, we hypothesized that more positive perceptions of oneself (i.e., greater state self-efficacy and state self-esteem) after the stressor discussion would predict greater decreases in trait attachment anxiety 1 year later (H1a and H2a). We also hypothesized that more positive perceptions of oneself (i.e., greater state self-efficacy and state self-esteem) after the stressor discussion would predict greater increases in trait self-perceptions (i.e., increased trait self-efficacy and trait self-esteem) 1 year later (H1b and H2b). Lastly, we hypothesized that increases in trait self-perceptions (i.e., increased treat self-efficacy and trait self-esteem) would explain associations between state self-perceptions after the stressor discussion and declines in attachment anxiety (H1c and H2c). Inherent in this indirect effect hypothesis is the prediction that increases in trait self-efficacy and trait self-esteem would predict declines in attachment anxiety (b-paths of the indirect effects).

Regarding attachment avoidance, we first hypothesized that more positive perceptions of one’s spouse (i.e., greater state PPR) after the stressor discussion would predict decreases in trait attachment avoidance 1 year later (H3a). We also hypothesized that more positive perceptions of one’s spouse (i.e., greater state PPR) after the stressor discussion would predict greater increases in trait spouse-perceptions (i.e., increased trait PPR) 1 year later (H3b). Lastly, we hypothesized that increases in trait PPR would explain the association between state PPR after the stressor discussion and declines in attachment avoidance (H3c). Inherent in this last hypothesis is the prediction that increases in trait PPR predict declines in attachment avoidance (b-path of the indirect effect).

We also tested alternative models that are inconsistent with the ASEM. Specifically, we assessed whether more positive perceptions of one’s spouse following the stressor discussion (i.e., state PPR) predict declines in attachment anxiety, and we assessed whether more positive perceptions of oneself following the stressor discussion (i.e., state self-efficacy, state self-esteem) predict declines in attachment avoidance. We assessed these alternative possibilities directly and indirectly through changes in trait partner- and self-perceptions. The indirect approach also allowed us to assess whether changes in trait partner- and self-perceptions predict changes in attachment insecurity that are inconsistent with the ASEM (i.e., an increase in trait PPR predicting declines in attachment anxiety; increases in trait self-efficacy and self-esteem predicting declines in attachment avoidance).

Method

Participants

Participants were 229 married couples who responded to solicitations for a research study of newlyweds. To be eligible for this study, participants needed to be married for less than a year, in their first marriage, fluent in English, and younger than 40 years old. Of these couples, 164 couples completed all or most of the year 2 assessments and were included in the current research.¹ We opted to test our hypotheses in a sample of newlywed couples because attachment may be most likely to change during life transitions (Arriaga et al., 2021; Rholes et al., 2021). These data were collected in Pittsburgh, Pennsylvania between 2004 and 2008. Data were collected prior to marriage equality in Pennsylvania, resulting in a sample of all opposite-sex couples (229 men, 229 women).

At baseline (year 1), the mean age of our sample was 27.16 years (SD = 4.66, range = 18–54). On average, couple members had been romantically involved with each other for 3.82 years (SD = 2.33) and had been married for 4.9 months (SD = 3.44). Participants self-identified as White/Caucasian (81.9%), Black/African American (6.9%), Asian (4.6%), or Hispanic/Latino (2.9%), American Indian (1.0%), or something else/multi-racial (2.7%). Most participants had completed a bachelor’s degree (33.1%), some college credits (18.1%), a professional degree (16.3%), or some graduate school (15.4%).

Procedure

Each couple attended an initial laboratory session where they completed a battery of questionnaires, including trait measures of attachment orientation, self-efficacy, self-esteem, and PPR. Approximately 1 week later, each couple returned to the lab to complete a series of activities together. Of relevance to the current investigation, partners took turns discussing the most significant personal stressor (not involving marital conflicts or problems) they were currently facing. The instructions specified that the stressor should be something that they had not already discussed with their spouse at length. Participants selected stressors that were moderately high in severity (M = 3.59, SD = 0.60; scale ranged from 1 [Not at all stressful] to 5 [Extremely stressful]), and they indicated that the stressors happened (or started bothering them) approximately 3 months prior to the laboratory discussion (M = 11.3 weeks, SD = 31.6). The majority of couples selected a stressor related to work (41.2%), a relationship with someone other than their spouse (11.1%), finances (7.4%), or education (6.7%). Couples discussed each participant’s stressor for 6 minutes, and the order of the two discussions was randomized. After each discussion, the experimenter instructed participants to complete measures assessing their current perceptions of themselves and their partner. Finally, couples were invited back to the lab 1 year later to complete the same battery of questionnaires they completed in their initial laboratory session.

Measures

Descriptive statistics, zero-order correlations, and reliability estimates for all measures are provided in Table 1. All measures and items used in the current study are available on OSF: https://osf.io/ehzdy/).

Table 1.

Descriptive Statistics, Correlations, and Reliability Estimates for all Variables.

Variable	M	SD	α	1	2	3	4	5	6	7	8	9	10	11	12
1. Year 1 att. anxiety	3.17	1.08	.88
2. Year 1 att. avoidance	3.17	0.98	.87	.30**
3. Year 1 trait self-efficacy	5.20	1.00	.88	−.33**	−.26**
4. Year 1 trait self-esteem	5.56	0.99	.88	−.46**	−.37**	.57**
5. Year 1 trait PPR	4.16	0.60	.84	−.11*	−.21**	.21**	.28**
6. Post-disc. state self-efficacy	5.42	1.07	.88	−.20**	−.14**	.32**	.38**	.22**
7. Post-disc. state self-esteem	5.97	1.02	.89	−.20**	−.17**	.21**	.28**	.32**	.71**
8. Post-disc. state PPR	3.95	0.75	.89	−.08	−.13**	.06	.03	.36**	.25**	.44**
9. Year 2 att. anxiety	2.82	1.11	.89	.64**	.25**	−.22**	−.45**	−.15**	−.23**	−.17**	.01
10. Year 2 att. avoidance	2.96	1.02	.87	.25**	.66**	−.18*	−.34**	−.14*	−.09	−.14*	−.06	.42**
11. Year 2 trait self-efficacy	5.08	1.00	.89	−.36**	−.24**	.75**	.52**	.20**	.40**	.23**	.03	−.40**	−.28**
12. Year 2 trait self-esteem	5.53	1.03	.90	−.40**	−.30**	.50**	.73**	.26**	.45**	.31**	.08	−.57**	−.39**	.66**
13. Year 2 trait PPR	4.09	0.65	.87	−.13*	−.16**	.15*	.20**	.60**	.16**	.23**	.28**	−.27**	−.18**	.21**	.32**

Note. Base. = Baseline; Att. = attachment; Post-disc. = post-discussion. * indicates p < .05. ** indicates p < .01.

Baseline (Year 1) and Follow-Up (Year 2) Measures

Attachment Insecurity

Baseline attachment orientation was assessed using an abbreviated (26-item) version of the Experiences in Close Relationships scale (ECR; Brennan et al., 1998). The trait attachment anxiety subscale (e.g., “I worry about being abandoned”) and trait attachment avoidance subscale (e.g., “I prefer not to show people how I feel deep down”) each contained 13 items. Follow-up attachment anxiety and attachment avoidance were assessed with 12 and 11 items from the baseline measure of attachment orientation, respectively. Participants indicated their agreement with each statement was using a 7-point scale (1 = Disagree strongly, 7 = Agree strongly), and we computed participants’ mean responses to items in each subscale (higher scores indicate greater attachment insecurity).²

Trait Self-Efficacy and Self-Esteem

Trait self-efficacy was assessed with the 10-item Sherer Self-Efficacy scale (e.g., “When I make plans, I am certain I can make them work”; Sherer et al., 1982), and trait self-esteem was assessed with the 10-item Rosenberg Self-Esteem scale (e.g., “I feel that I am a person of worth, at least on an equal basis with others”; Rosenberg, 1965). Participants indicated their agreement with each item using a 7-point scale (1 = Disagree strongly, 7 = Agree strongly), and we computed participants’ mean responses to items in each scale (higher scores indicate greater self-efficacy/esteem).

Trait Perceived Partner Responsiveness

Participants responded to a set of 51 items assessing general perceptions of their spouse. We selected seven items that mapped most clearly onto the construct of perceived partner responsiveness, as defined by Crasta et al. (2021; e.g., understanding, caring, considerate). A detailed description of our criteria for selecting these items is available on OSF (see link above). Participants indicated how much each item generally describes their spouse using a 5-point scale (1 = Not at all, 5 = Extremely), and we computed the mean of participants’ responses (higher scores indicate higher PPR).

Post-Stressor Discussion Measures

State Self-Efficacy and Self-Esteem

Participants indicated how they felt about themselves at that moment (following a discussion of their own personal stressor) using 15 pairs of semantic differential items adapted from McFarland and Ross’ (1982) measure. Of these 15 item pairs, we used six to assess state self-efficacy (e.g., useless vs. useful, stupid vs. smart, capable vs. incapable) and six to assess state self-esteem (e.g., rejected vs. accepted, liked vs. disliked, lovable vs. unlovable). Each pair of items was separated by seven scale points, and we computed a mean for each scale (higher scores indicate higher state self-efficacy/esteem).

State Perceived Partner Responsiveness

Participants reported their perceptions of their spouse during the discussion of their own personal stressor by indicating the extent to which 51 adjectives described their spouse (1 = Not at all, 5 = Extremely). Most of these adjectives—though not all—were the same as the trait perceptions assessed above. State PPR was measured as the mean of seven items that best represent the construct of perceived partner responsiveness (e.g., understanding, caring/compassionate, thoughtful/considerate), and it was nearly identical to the measure of trait PPR.

Data Analytic Strategy

Data were analyzed in R studio (see OSM for a list of packages and example syntax). The data were organized so that observations from each couple were spread across two rows (one row for each spouse). For the current study’s purposes, we focused on baseline and year 2 data from each spouse, as well as data provided by each spouse after discussing their own stressor.

Although we did not hypothesize gender differences, we tested for and observed differences in several key variables (described below). Because of the observed gender differences, we conducted analyses that allowed our hypothesized associations to vary for men and women. To test sub-hypotheses a and b, we estimated multilevel linear models where an outcome variable was predicted by the relevant post-discussion variables, control variables, gender (female = −1, male = 1), gender’s interaction with the previous predictors, and a by-couple random intercept to account for non-independence of each spouse’s responses. For example, when testing sub-hypothesis 1a, we regressed year 2 attachment anxiety onto baseline attachment anxiety and avoidance (to estimate residualized change in attachment anxiety and to control for attachment avoidance), state self-efficacy (the hypothesized predictor), gender, and interactions between gender and the predictor variables. Results were consistent with pre-registered analyses that did not include gender moderation and did not control for the other attachment dimension (see Supplemental Tables 2-3).

To test the hypothesized indirect effects, we used the lavaan package to test distinguishable mediation models. We estimated links separately for wives and husbands in the same model and estimated covariance between spouses’ reports and their error terms to account for non-independence. In these models, a post-discussion variable (e.g., state self-efficacy) indirectly predicted residualized change in an attachment variable (e.g., attachment anxiety at year 2, controlling for attachment anxiety and attachment avoidance at year 1) through residualized change in the explanatory variable (e.g., trait self-efficacy at year 2, controlling for trait self-efficacy at year 1). Results were consistent in the pre-registered analyses that did not control for the other attachment dimension (see Supplemental Figures 1-6).

Results

Gender Differences

When testing the average within-couple gender differences in our key variables, we observed significant gender differences such that husbands reported greater trait avoidance at year 1 (b = .09, p = .045), greater trait self-efficacy at year 1 (b = .14, p = .001), lower trait PPR at year 1 (b = −.05, p = .034), lower post-discussion state self-esteem (b = −.13, p = .009), and greater trait avoidance at year 2 (b = .14, p = .011). Men also reported marginally lower trait attachment anxiety at year 1 (b = −.09, p = .054). Because of these gender differences, we treated dyads as distinguishable in all analyses.

Predicting Change in Attachment Anxiety

Hypothesis 1 (Self-Efficacy).

We first tested whether state self-efficacy following one’s stressor discussion predicts residualized change in trait attachment anxiety from year 1 to year 2 (H1a). Consistent with our hypothesis, greater post-discussion state self-efficacy predicted greater declines in attachment anxiety at follow-up (see Table 2). The strength of this association did not differ for husbands and wives, and this association remained significant when controlling for baseline trait self-efficacy (to control for general self-perceptions unrelated to the discussion; see Supplemental Table 4).

We also found support for the hypothesis that greater post-discussion state self-efficacy predicts residualized change in trait self-efficacy from year 1 to year 2 (H1b). Greater post-discussion state self-efficacy was significantly associated with increases in trait self-efficacy at follow-up, and the strength of this association did not differ for husbands and wives (see Table 2).

Finally, we tested the full mediation model to test whether residualized change in trait self-efficacy indirectly links post-stressor discussion state self-efficacy and residualized change in attachment anxiety (H1c). Estimates of this model for wives and husbands are presented in Figure 1, and key statistics from these models are provided in-text below. For wives, greater post-stressor discussion state self-efficacy predicted greater residualized change in trait self-efficacy from year 1 to year 2 (a-path; B = .11, p = .027, 95% CI [.01, .21]), and these changes in trait self-efficacy predicted residualized decreases in attachment anxiety at follow-up (b-path; B = −.48, p < .001, 95% CI [-.72, −.23]). The predicted indirect effect (H1c) approached significance and did not overlap with zero (B = −0.05, p = .058, 95% CI [-.11, −.001]), suggesting that change in trait self-efficacy partially explains the association between greater post-stressor discussion self-efficacy and change in trait attachment anxiety for wives.

The results were similar for husbands. Greater post-stressor discussion state self-efficacy predicted residualized change in trait self-efficacy at follow-up (a-path; B = .19, p = .009, 95% CI [.05, .33]), and these changes in trait self-efficacy predicted residualized decreases in attachment anxiety at follow-up (b-path; B = −.42, p = .002, 95% CI [-.68, −.14]). Finally, the predicted indirect effect (H1c) was significant (B = −0.08, p = .025, 95% CI [−.15, −.02]), suggesting that increased trait self-efficacy partially explains the association between greater post-stressor discussion state self-efficacy and change in trait attachment anxiety for husbands.

Table 2.

Models Predicting Residualized Change in Attachment Anxiety and Trait Self-Efficacy at Year 2.

Predictor	b	SE	95% CI
Attachment anxiety at year 2
(Intercept)	2.81***	0.05	[2.72, 2.92]
Baseline attachment anxiety	0.66***	0.05	[0.56, 0.76]
Baseline attachment avoidance	0.06	0.05	[−0.05, .17]
State self-efficacy	−0.10*	0.05	[−0.20, -0.01]
Gender	0.00	0.05	[−0.09, 0.10]
Baseline attachment anxiety × Gender	0.01	0.05	[−0.09, 0.11]
Baseline attachment avoidance × Gender	−0.06	0.05	[−0.17, 0.04]
State self-efficacy × Gender	−0.01	0.05	[−0.10, 0.08]
Trait self-efficacy at year 2
Predictor	b	SE	95% CI
(Intercept)	5.09***	0.04	[5.01, 5.16]
Baseline self-efficacy	0.73***	0.06	[0.66, .80]
State self-efficacy	0.16***	0.03	[0.09, 0.23]
Gender	−0.10	0.03	[−0.16, −0.03]
Baseline self-efficacy × Gender	−0.00	0.04	[−0.08, 0.07]
State self-efficacy × Gender	0.05	0.03	[−0.02, 0.12]

Notes. Gender was coded female = −1, male = 1. The predictor of interest is shown in bold. Estimates are unstandardized. *indicates p < .05. **indicates p < .01. ***indicates p < .001.

Figure 1.

Wives’/Husbands’ estimates (standardized) from the confirmatory model (H1c) testing whether increased trait self-efficacy indirectly explains any association between post-discussion state self-efficacy and change in attachment anxiety. * indicates p < .05; ** indicates p < .01; *** indicates p < .001.

Hypothesis 2 (Self-Esteem).

Next, we tested the hypothesis that post-discussion state self-esteem predicts residualized change in attachment anxiety from year 1 to year 2 (H2a). This hypothesis was not supported. Post-discussion state self-esteem was not associated with decreases in attachment anxiety (see Table 3), and the association remained nonsignificant in analyses controlling for baseline trait self-esteem (see Supplemental Table 4).

We then tested whether greater post-discussion self-esteem predicts residualized change in trait self-esteem from year 1 to year 2 (H2b). Consistent with our hypothesis, greater post-discussion state self-esteem predicted greater increases in trait self-esteem at follow-up, and the strength of this association did not differ for husbands and wives (see Table 3).

Next, we tested the full mediation model designed to test whether residualized change in trait self-esteem indirectly links post-stressor discussion state self-esteem and residualized change in attachment anxiety (H2c; see Figure 2). For wives, post-discussion state self-esteem did not predict residualized change in trait self-esteem at follow-up (a-path; B = .08, p = .151, 95% CI [−.02, .20]), but residualized change in trait self-esteem predicted lower attachment anxiety at follow-up (b-path; B = −.51, p < .001, 95% CI [-.70, −.33]). The predicted indirect effect (H2c) was not significant (B = −.04, p = .181, 95% CI [−.11, .01]), suggesting that change in trait self-esteem did not explain the association between post-stressor discussion state self-esteem and change in trait attachment anxiety for wives.

For husbands, greater post-discussion state self-esteem did predict residualized increases in trait self-esteem at follow-up (a-path; B = .14, p = .019, 95% CI [.03, .27]), and these changes in trait self-esteem also predicted residualized decreases in attachment anxiety at follow-up (b-path; B = −.43, p = .001, 95% CI [-.68, −.17]). Here, the predicted indirect effect (H2c) was approaching significance and the confidence interval did not contain zero (B = −.06, p = .054 95% CI [-.13, −.01]) suggesting that change in trait self-esteem partially explains the association between post-discussion state self-esteem and change in trait attachment anxiety for husbands.

Table 3.

Models Predicting Residualized Change in Attachment Anxiety and Trait Self-Esteem at Year 2.

Predictor	b	SE	95% CI
Attachment anxiety at year 2
(Intercept)	2.81***	0.05	[2.72, 2.92]
Baseline attachment anxiety	0.67***	0.05	[0.57, 0.78]
Baseline attachment avoidance	0.07	0.06	[-0.04, .17]
State self-esteem	−0.03	0.05	[-0.13, 0.07]
Gender	0.003	0.05	[-0.20, 0.09]
Baseline attachment anxiety × Gender	0.009	0.05	[-0.10, 0.11]
Baseline attachment avoidance × Gender	−0.06	0.05	[-0.17, 0.04]
State self-esteem × Gender	−0.04	0.05	[-0.14, 0.06]
Trait self-efficacy at year 2
Predictor	b	SE	95% CI
(Intercept)	5.54***	0.04	[5.47, 5.62]
Baseline self-esteem	0.75***	0.04	[0.67, 0.84]
State self-esteem	0.12**	0.04	[0.03, 0.19]
Gender	−0.01	0.04	[-0.09, 0.07]
Baseline self-esteem × Gender	−0.06	0.04	[-0.14, 0.02]
State self-esteem × Gender	0.05	0.04	[-0.03, 0.13]

Notes. Gender was coded female = −1, male = 1. The predictor of interest is shown in bold. Estimates are unstandardized. * indicates p < .05. ** indicates p < .01. *** indicates p < .001.

Figure 2.

Wives’/Husbands’ estimates (standardized) from the confirmatory model (H2c) testing whether increased trait self-esteem indirectly explains any association between post-discussion state self-esteem and change in attachment anxiety. * indicates p < .05; ** indicates p < .01; *** indicates p < .001.

Alternative Model (Perceived Partner Responsiveness)

Consistent with the ASEM, we did not observe evidence that state PPR following one’s stressor discussion predicts residualized change in trait attachment anxiety from year 1 to year 2 (b = 0.05, p = .407, 95% CI [−0.09, 0.18]; see Supplemental Table 5 for full results). As shown in Figure 3, an indirect effects model did not show significant indirect effects linking state PPR to declines in attachment anxiety from year 1 to year 2 for wives (B = −0.02, p = .425, 95% CI [-.08, .02]) or for husbands (Indirect effect for husbands, B = −0.05, p = .182, 95% CI [−.13, .02]). However, the indirect effects model showed significant b-paths for wives and husbands. In other words, change in trait PPR predicted residualized change in attachment anxiety from year 1 to year 2 for wives (B = −0.39, p = .009, 95% CI [−.67, −.10]) and for husbands (B = −0.49, p = .002, 95% CI [-.80, −.18]). These significant b-paths suggest that—counter to the ASEM—increasingly positive trait spouse-perceptions predict declines in attachment anxiety.

Figure 3.

Wives’/Husbands’ estimates (standardized) from the exploratory model testing whether increased trait PPR indirectly explains any association between post-discussion state PPR and change in attachment anxiety. * indicates p < .05; ** indicates p < .01; *** indicates p < .001.

Predicting Change in Attachment Avoidance

Hypothesis 3 (PPR).

The hypothesis that state PPR following one’s stressor discussion predicts residualized change in trait attachment avoidance from year 1 to year 2 (H3a) was not supported. Greater post-stressor discussion state PPR was not significantly associated with decreases in attachment avoidance at follow-up (see Table 4) and remained nonsignificant controlling for baseline trait PPR (see Supplemental Table 4). Only baseline attachment avoidance and gender (men reported greater attachment avoidance than women) predicted follow-up attachment avoidance. We also did not find support for the hypothesis that greater post-discussion state PPR predicts residualized change in trait PPR (H3b). Greater post-discussion state PPR was not associated with residualized change in trait PPR (see Table 4).

Finally, we tested the full mediation model to assess whether residualized change in trait PPR indirectly links post-stressor discussion PPR and residualized change in attachment avoidance (see Figure 4). For wives, state PPR after the stressor discussion did not predict residualized change in trait PPR at follow-up (a-path; B = .05, p = .375, 95% CI [−.06, .17]), and change in trait PPR did not predict residualized change in attachment avoidance at follow-up (b-path; B = −.05, p = .719, 95% CI [−.33, .21]). The indirect effect was also not significant (B = −0.003, p = .818, 95% CI [−.03, .02]). Results were similar for husbands. Greater state PPR did not predict residualized change in trait PPR at follow-up (a-path; B = .10, p = .130, 95% CI [−.04, .22]), and change in PPR did predict residualized change in attachment avoidance at follow-up (b-path; B = −.24, p = .074, 95% CI [−.50, .01]), though the latter link was approaching significance. The indirect effect was not significant (B = −0.02, p = .226, 95% CI [−.06, .01]). These results suggest that neither wives’ nor husbands’ post-stressor discussion PPR is linked with change in trait attachment avoidance indirectly through changes in trait PPR.

Table 4.

Models Predicting Residualized Change in Attachment Avoidance and Trait PPR at Year 2.

Predictor	b	SE	95% CI
Attachment anxiety at year 2
(Intercept)	3.01***	0.05	[2.92, 3.10]
Baseline attachment avoidance	0.67***	0.05	[0.58, 0.76]
Baseline attachment anxiety	0.06	0.04	[−0.03, 0.15]
State PPR	0.01	0.06	[−0.10, 0.13]
Gender	0.10*	0.04	[0.03, 0.18]
Baseline attachment avoidance × Gender	−0.09	0.05	[−0.19, 0.001]
Baseline attachment anxiety × Gender	0.02	0.04	[−0.06, 0.11]
State PPR × Gender	0.02	0.06	[−0.09, 0.13]
Trait PPR at year 2
Predictor	b	SE	95% CI
(Intercept)	4.08***	0.03	[4.02, 4.13]
Baseline PPR	0.65***	0.06	[0.55, 0.76]
State PPR	0.07	0.04	[−0.02, 0.16]
Gender	0.02	0.03	[−0.04, 0.08]
Baseline PPR × Gender	0.01	0.05	[−0.10, 0.12]
State PPR × Gender	0.01	0.04	[−0.07, 0.10]

Note. Gender was coded female = −1, male = 1. The predictor of interest is shown in bold. Estimates are unstandardized. *indicates p < .05. ** indicates p < .01. *** indicates p < .001.

Figure 4.

Wives’/Husbands’ estimates (standardized) from the confirmatory model (H3c) testing whether increased trait PPR indirectly explains any association between post-discussion state PPR and change in attachment avoidance. * indicates p < .05; ** indicates p < .01; *** indicates p < .001.

Alternative Models

Consistent with the ASEM, we did not observe that state self-efficacy or state self-esteem following one’s stressor discussion predicts residualized change in trait attachment avoidance over time (b = −0.01, p = .793, 95% CI [−.09, .07], b = −0.03, p = .527, 95% CI [−.11, .06], respectively; see Supplemental Table 6 for full results). However, we did observe some findings that are inconsistent with the ASEM in the indirect effect models.

First (as shown in Figure 5), husbands’ state self-efficacy following the stressor discussion predicted residualized change in trait self-efficacy (a-path, B = .19, p = .009, 95% CI [.04, .32]), and change in husbands’ trait self-efficacy predicted residualized change in attachment avoidance (b-path, B = −.39, p < .001, 95% CI [−.59, −.20]), with a significant indirect effect (B = −0.07, p = .035, 95% CI [−.15, −.01]). A comparable indirect effect linking state self-efficacy and change in attachment avoidance was not observed for wives (B = −0.02, p = .208, 95% CI [−.05, .01]), and the b-path was also nonsignificant for wives (b = −0.16, p = .111). In other words, increasing trait self-efficacy (following a stressor discussion) predicted declines in attachment avoidance for husbands.

Figure 5.

Wives’/Husbands’ estimates (standardized) from the exploratory model testing whether increased state self-efficacy indirectly explains any association between post-discussion state self-efficacy and change in attachment avoidance. * indicates p < .05; ** indicates p < .01; *** indicates p < .001.

The indirect effects model linking state self-esteem to declines in attachment avoidance (see Figure 6) did not show significant indirect effects for wives (B = −0.02, p = .248, 95% CI [−.07, .01]) or for husbands (B = −0.04, p = .103, 95% CI [-.10, −.005]). However, we observed significant b-paths for wives and husbands. In other words, a change in trait self-esteem predicted residualized change in attachment avoidance from year 1 to year 2 for wives (B = −0.26, p = .011, 95% CI [−.48, −.07]) and for husbands (B = −0.28, p = .013, 95% CI [−.51, −.08]). Collectively, these findings provide initial evidence that increasingly positive self-perceptions may contribute to declines in attachment avoidance, in contradiction to the ASEM.

Figure 6.

Wives’/Husbands’ estimates (standardized) from the exploratory model testing whether increased state self-esteem indirectly explains any association between post-discussion state self-esteem and change in attachment avoidance. * indicates p < .05; ** indicates p < .01; *** indicates p < .001.

Supplemental Analyses

Given the unanticipated b-paths that are inconsistent with the ASEM (i.e., change in trait PPR predicting residualized change in attachment anxiety; change in trait self-efficacy and self-esteem predicting residualized change in attachment avoidance) and the one unanticipated indirect effect (i.e., state self-efficacy indirectly predicting change in attachment avoidance through increases in trait self-efficacy, for husbands), we conducted additional analyses to confirm that these results were not an artifact of our analysis approach.

First, we ran additional indirect effect models with indirect effects tested simultaneously to assess each b-path controlling for other b-paths (see Supplemental Figures 7-10 for full results). Because trait self-efficacy and trait self-esteem are highly correlated, we assessed them in separate models to avoid suppression effects. In a model combining indirect effects for self-efficacy and PPR (Supplemental Figure 7), we continued to observe significant b-paths for both change in self-efficacy and PPR for wives and husbands. In a second model combining indirect effects for self-esteem and PPR (Supplemental Figure 8), we also continued to observe significant b-paths for change in self-esteem (for wives and husbands) and PPR (for husbands only). Overall, these results suggest that changes in trait self-efficacy, trait self-esteem, and trait PPR all predict declines in attachment anxiety, even when modeled simultaneously.³

In a combined indirect effects model predicting attachment avoidance, we continued to observe a significant b-path for change in trait self-efficacy predicting declines in attachment avoidance (for husbands only) and, we continued to observe no b-paths for change in trait PPR (Supplemental Figure 9). Similarly, we observed a significant b-path for change in trait self-esteem predicting declines in attachment avoidance (for husbands only) and nonsignificant b-paths for husbands’ and wives’ change in trait PPR (Supplemental Figure 10) ⁴

As one final approach to determine the consistency of these findings across analysis approaches, we computed change scores for trait self-efficacy, trait self-esteem, and trait PPR from year 1 to year 2 (see Supplemental Table 7 for descriptives and zero-order correlations), and we used those change scores to predict residualized change in attachment anxiety and residualized changed in attachment avoidance (controlling for the other attachment dimension). As shown in Supplemental Table 8, increases in trait self-efficacy (b = −.26, p < .001, 95% CI [−0.40, −0.12]), trait self-esteem (b = −. 19, p = 005, 95% CI [−0.32, −0.96]), and trait PPR (b = −.22, p = .007, 95% CI [−0.38, −0.06]) each independently predicted declines in attachment anxiety. Regarding change in attachment avoidance, increases in trait self-esteem predicted declines in attachment avoidance (b = −.15, p = 019, 95% CI [−0.29, −0.03]), and neither change in trait self-esteem (b = −.10, p = .099, 95% CI [−0.12, 0.02]) or trait PRR (b = −.10, p = .185, 95% CI [−0.25, 0.05]) predicted declines in attachment avoidance (see Supplemental Table 9). None of the effects in either model were moderated by gender.

Discussion

Guided by the ASEM (Arriaga et al., 2018), the current research assessed whether improving the model of self (i.e., developing self-efficacy and self-esteem) and improving the model of others (i.e., developing PPR toward one’s spouse) predict declines in attachment anxiety and attachment avoidance, respectively. We were particularly interested in investigating how (in what contexts) people can develop these positive (trait) perceptions of themselves and others. We posited that discussing a personal stressor with one’s spouse provides diagnostic information about oneself and one’s spouse so that immediate (state) perceptions following these discussions are consequential for shifting trait perceptions and attachment orientation. We tested ASEM-consistent models and alternative models that deviate from the ASEM.

Findings from Confirmatory (ASEM -Consistent) Models

We observed that increases in trait self-efficacy and trait self-esteem over 1 year predict residualized declines in attachment anxiety over the same year. These findings lend additional support for the ASEM’s proposition that bolstering the model of self reduces attachment anxiety and provide novel evidence that self-esteem, like self-efficacy, is a key self-component involved in attachment anxiety reduction (e.g., Arriaga et al., 2014; Arriaga et al., 2021).

Moreover, and consistent with hypotheses, state self-perceptions immediately following a personal stressor discussion predicted changes in trait self-perceptions and declines in attachment anxiety. Specifically, we observed that greater post-discussion state self-efficacy (but not self-esteem) predicted greater declines in attachment anxiety over 1 year. For self-efficacy, we also observed the theorized indirect effect: People who reported greater post-discussion state self-efficacy showed greater increases in trait self-efficacy over the year, which in turn predicted greater decreases in attachment anxiety over the same year. The results for the theorized self-esteem indirect effect were similar but less robust. State self-esteem following the stressor discussions predicted increases in trait self-esteem and increases in trait self-esteem predicted declines in attachment anxiety, but the indirect effect only approached significance for husbands and was not observed for wives. Evaluations of one’s competence (trait self-efficacy) may change more quickly than evaluations of one’s self-worth (trait self-esteem), or personal stressor discussions may provide more affordances for fostering state self-efficacy than state self-esteem, either of which could explain the disparate findings.

In any case, these novel findings regarding state self-perceptions suggest that attachment anxiety change may be initiated in everyday moments—such as personal stressor discussions—that provide diagnostic information relevant to the working model of self (Arriaga et al., 2018). Partners likely serve an important role in fostering positive state self-perceptions during stressor discussions: people feel efficacious and develop self-esteem when their partners emphasize their abilities, provide encouragement, and compliment their efforts and progress during stressor discussions (Jayamaha & Overall, 2019).

Regarding attachment avoidance, we did not find support for any part of our theorized model. First, state PPR following the personal stressor discussion did not predict changes in attachment avoidance or changes in trait PPR. It is unclear why state PPR did not predict increases in trait PPR. Perhaps we could have observed this association with a validated measure of trait PPR that uses statements instead of adjectives (e.g., Crasta et al., 2021), or perhaps perceptions of a partner’s trait responsiveness shift more in other diagnostic situations (e.g., strain-tests; Simpson, 2007). Second, and perhaps more surprisingly, changes in trait PPR from baseline to follow-up did not predict changes in attachment avoidance over the same year. One explanation for this null finding is that the PPR measures we used assessed perceptions of the spouse’s responsiveness, whereas the attachment avoidance measure assessed attachment avoidance in close relationships globally. Measuring relationship-specific attachment (towards one’s spouse) would avoid this mismatch in future work. Relationship-specific attachment and global attachment do change in concert (Dugan et al., 2022), but PPR is likely more closely linked to changes in relationship-specific attachment than global attachment.

Findings from Alternative (ASEM -Inconsistent) Models

Although an increase in trait PPR did not predict a decline in attachment avoidance, it did predict a decline in attachment anxiety. In other words, people who came to see their spouses as more responsive over the year experienced greater declines in global attachment anxiety than people who did not develop trait PPR. This finding is inconsistent with the ASEM, but it does make some sense given the nature of attachment anxiety. People with high attachment anxiety do not only hold negative beliefs about themselves; they also hold “ambivalent expectations about others” including “chronic fears of being abandoned” (Arriaga et al., 2018, p. 73). Theoretically, these ambivalent expectations of others are rooted in negative self-perceptions and should be eased when people believe that they are worthy of others’ care. However, it is also sensible that attachment anxiety could decline when people perceive that important close others are responsive and therefore unlikely to abandon or reject them. Attachment theory describes responsiveness as the core feature underlying attachment security, so it is reasonable that PPR could predict a decline in attachment anxiety as well as avoidance.

Also inconsistent with the ASEM, we observed that people (especially men) who reported greater increases in self-efficacy or self-esteem also reported greater decreases in attachment avoidance. People with high attachment avoidance are not considered to have a negative model of self; instead, they can have “inflated” self-perceptions as a “defensive strategy for being self-reliant” (Arriaga et al., 2018, p. 73). Thus, it is not immediately clear why strengthening the model of self would predict declines in attachment avoidance. One possibility is that people (particularly men) who develop self-efficacy and self-esteem report lower attachment avoidance because the meaning of some attachment avoidance items varies depending on one’s self-perceptions. For instance, “I prefer not to show people how I feel deep down” has a different meaning depending on how people feel “deep down.” People who develop self-efficacy/esteem may stop endorsing this item simply because sharing their positive self-perceptions is less vulnerable than sharing feelings of incapability or worthlessness. Alternatively, enhancing one’s model of self may genuinely contribute to declines in attachment avoidance if questioning and revising self-perceptions encourages people to question the validity of their distrust in others as well. We suggest caution in interpreting this unexpected finding since past research has not found that improvements to the model of self predict declines in attachment avoidance (Arriaga et al., 2021).

Limitations and Future Directions

The current work has several limitations that provide avenues for future research. First, the nature of our state post-discussion measures limits our ability to establish conclusively that personal stressor discussions are a diagnostic context relevant to attachment change. Future research could more directly assess changes in self- and partner-perceptions during stressor discussions to ensure that these perceptions are directly tied to that context. Our state perception measures were only modestly correlated with corresponding baseline trait measures, and links between state perceptions and attachment held controlling for trait perceptions, demonstrating that state post-discussion reports are distinct from general perceptions. Nonetheless, we cannot be certain that our state measures assessed changes tied only to the stressor discussions.

The current work is also limited in that it does not describe how state perceptions generalize to trait perceptions. There may be behavioral mediators (e.g., enhanced goal pursuit resulting from state self-efficacy, willingness to depend on a partner resulting from state PPR) that provide opportunities for brief state experiences to generalize. Alternatively, the state perceptions we observed after the personal stressor discussion may represent a snapshot of how participants typically perceive themselves and one another in stressful and stress-related contexts, in which case an accumulation of similar experiences may shift a state perception into a trait-like perception. A related limitation is that the current work does not clarify which characteristics of the stressor discussions (e.g., stressor topic, partner behaviors) facilitate the development of positive state self- and partner-perceptions. Future research can build on the current research by delving into the specifics of the stressor discussions themselves. Finally, these data were collected 15–19 years ago (between 2004 and 2008), and at the time we collected no information about participants’ gender identities (only sex), sexual orientation, or disability status. Future research should collect these characteristics to assess whether they are relevant to attachment change processes.

In conclusion, this work makes two important contributions to the growing literature on attachment change in adulthood. First, we showed that state self-perceptions following a personal stressor discussion predict changes in trait self-perceptions and declines in attachment insecurity (attachment anxiety, as expected, but also attachment avoidance). Second, findings from the ASEM-consistent and alternative models suggest that enhancing the model of self and the model of others enhances attachment security, though the processes for reducing attachment anxiety and avoidance may not be completely separable. Future research can build on these findings by evaluating how state perceptions generalize to shift trait perceptions, by identifying other interpersonal contexts that catalyze attachment changes in adulthood, and by continuing to test precursors of changes in attachment avoidance to resolve discrepant findings.

Supplemental Material

Supplemental Material - Perceptions of Oneself and One’s Spouse Following a Stressor Discussion Predicting Attachment Insecurity Over One Year

Supplemental Material for Perceptions of Oneself and One’s Spouse Following a Stressor Discussion Predicting Attachment Insecurity Over One Year by Brett K. Jakubiak, Julian D. Fuentes, and Brooke C. Feeney in Journal of Social and Personal Relationships

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a grant from the National Science Foundation (BCS0424579) awarded to Dr. Brooke Feeney

Author Note

Brett K. Jakubiak is at Assistant Professor at Syracuse University; Julian D. Fuentes is a graduate student in the Social Psychology PhD program at Syracuse University; and Brooke C. Feeney is a Professor of Psychology at Carnegie Mellon University.

Open Research Statement

As part of IARR’s encouragement of open research practices, the author(s) have provided the following information: This research was pre-registered. The aspects of the research that were pre-registered were the hypotheses and data analysis plan. The registration was submitted to OSF: .

Data Availability

The data used in the research are available and can be obtained by emailing bkjakubi@syr.edu. The materials and measures are available and can be accessed at: .

ORCID iDs

Brett K. Jakubiak

Julian D. Fuentes

Supplemental Material

Supplemental material for this article is available online

Notes

References

Arriaga

X. B.

Eller

Kumashiro

Rholes

W. S.

Simpson

J. A.

(2021). Self-efficacy and declines over time in attachment anxiety during the transition to parenthood. Social Psychological and Personality Science, 12(5), 658–666. https://doi.org/10.1177/1948550620933411

Arriaga

X. B.

Kumashiro

Finkel

E. J.

VanderDrift

L. E.

Luchies

L. B.

(2014). Filling the void: Bolstering attachment security in committed relationships. Social Psychological and Personality Science, 5(4), 398–406. https://doi.org/10.1177/1948550613509287

Arriaga

X. B.

Kumashiro

Simpson

J. A.

Overall

N. C.

(2018). Revising working models across time: Relationship situations that enhance attachment security. Personality and Social Psychology Review: An Official Journal of the Society for Personality and Social Psychology, Inc, 22(1), 71–96. https://doi.org/10.1177/1088868317705257

Bayraktaroglu

Gunaydin

Selcuk

Besken

Karakitapoglu-Aygun

(2022). The role of positive relationship events in romantic attachment avoidance. Journal of personality and social Psychology. Advance online publication. https://doi.org/10.1037/pspi0000406

Bowlby

(1980). Attachment and loss. Basic Books.

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). Guilford Press.

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. https://doi.org/10.1016/J.PAID.2019.04.037

Collins

N. L.

Feeney

B. C.

(2000). A safe haven: An attachment theory perspective on support seeking and caregiving in intimate relationships. Journal of Personality and Social Psychology, 78(6), 1053–1073. https://doi.org/10.1037/0022-3514.78.6.1053

Crasta

Rogge

R. D.

Maniaci

M. R.

Reis

H. T.

Crasta

Rogge

R. D.

Maniaci

M. R.

Reis

H. T.

(2021). Toward an optimized measure of perceived partner responsiveness: Development and validation of the perceived responsiveness and insensitivity scale. Psychological Assessment, 33(4), 338–355. https://doi.org/10.1037/pas0000986

10.

Dugan

K. A.

Fraley

R. C.

Gillath

Deboeck

P. R.

(2022). Changes in global and relationship-specific attachment working models. Journal of Social and Personal Relationships, 39(10), 3015–3043. https://doi.org/10.1177/02654075211051408

11.

Feeney

B. C.

Van Vleet

Jakubiak

B. K.

(2015). An attachment-theoretical perspective on optimal dependence in close relationships. In Simpson

J. A.

Rholes

S. W.

(Eds), Attachment theory and research: New directions and emerging themes (pp. 195–233). The Guilford Press.

12.

Feeney

Fitzgerald

(2019). Attachment, conflict and relationship quality: Laboratory-based and clinical insights. Current Opinion in Psychology, 25, 127–131. https://doi.org/10.1016/j.copsyc.2018.04.002

13.

Fraley

R. C.

Shaver

P. R.

(2000). Adult romantic attachment: Theoretical developments, emerging controversies, and unanswered questions. Review of General Psychology, 4(2), 132–154. https://doi.org/10.1037//1089-2680.4.2.132

14.

Gazder

Stanton

S. C. E.

(2022). Longitudinal associations between mindfulness and change in attachment orientations in couples: The role of relationship preoccupation and empathy. Journal of Social and Personal Relationships, 026540752211396. https://doi.org/10.1177/02654075221139654

15.

Jakubiak

B. K.

Fuentes

J. D.

Sun

E. R.

Feeney

B. C.

(2022). Social support is a balancing act: Mitigating attachment anxiety by supporting independence and dependence. Journal of personal and social relationships. Advance online publication. https://doi.org/10.1177/02654075221142000

16.

Jayamaha

S. D.

Overall

N. C.

(2019). The dyadic nature of self-evaluations: Self-esteem and efficacy shape and are shaped by support processes in relationships. Social Psychological and Personality Science, 10(2), 244–256. https://doi.org/10.1177/1948550617750734

17.

McFarland

Ross

(1982). Impact of causal attributions on affective reactions to success and failure. Journal of Personality and Social Psychology, 43(5), 937–946. https://doi.org/10.1037/0022-3514.43.5.937

18.

Mikulincer

Shaver

P. R.

(2016). Attachment in adulthood: Structure, dynamics, and change (2nd ed.). Guilford Press.

19.

Mischel

Shoda

(1995). A cognitive-affective system theory of personality: Reconceptualizing situations, dispositions, dynamics, and invariance in personality structure. Psychological Review, 102(2), 246–268. https://doi.org/10.1037/0033-295X.102.2.246

20.

Pietromonaco

P. R.

Beck

L. A.

(2019). Adult attachment and physical health. Current Opinion In Psychology, 25, 115–120. https://doi.org/10.1016/j.copsyc.2018.04.004

21.

Rholes

W. S.

Eller

Simpson

J. A.

Arriaga

X. B.

(2021). Support processes predict declines in attachment avoidance across the transition to parenthood. Personality and Social Psychology Bulletin, 47(5), 810–825. https://doi.org/10.1177/0146167220948706

22.

Rosenberg

(1965). Society and the adolescent self-image. Princeton University Press.

23.

Sherer

Maddux

J. E.

Mercandante

Prentice-Dunn

Jacobs

Rogers

R. W.

(1982). The self-efficacy scale: Construction and validation. Psychological Reports, 51(2), 663–671. https://doi.org/10.2466/pr0.1982.51.2.663

24.

Simpson

J. A.

(2007). Psychological foundations of trust. Current Directions in Psychological Science, 16(5), 264–268. https://doi.org/10.1111/j.1467-8721.2007.00517.x

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.35 MB