THE JOURNAL OF SEX RESEARCH, 55(2), 191–201, 2018

Copyright © The Society for the Scientific Study of Sexuality

ISSN: 0022-4499 print/1559-8519 online
DOI: https://doi.org/10.1080/00224499.2017.1318817

The Influence of Attachment Styles on

Sexual Communication Behavior
Julia McNeil, Uzma S. Rehman and Erin Fallis
Department of Psychology, University of Waterloo

Attachment theory is one of the major theoretical frameworks for understanding romantic relation-
ships. Attachment styles are formed through interactions with caregivers and shape an individual’s
expectations of subsequent interpersonal relationships. In this study, we examined how attachment
styles influence participants’ ability to communicate with their partners about problems in their sexual
relationship. A community sample of 81 couples engaged in two video-recorded discussions, one
representing an aspect of the couple’s sexual relationship where the male partner wanted change and
the second representing an aspect of the sexual relationship where the female partner wanted change.
Conversations were then coded, with each person being rated on three positive communication
dimensions (positive affect, offering solutions, and responsiveness) and three negative communication
dimensions (hostility, negative affect, and unskilled communication behaviors). As predicted, attach-
ment avoidance was related to more negative and less positive communication for both the individual
and his or her partner. Our observational data did not reveal any significant effects of attachment
anxiety on sexual communication. These results can be contrasted with findings from self-report
studies that do suggest an anxious attachment adversely impacts sexual communication.

Sexual communication is an essential part of developing and issues in committed romantic relationships. While the
maintaining healthy sexual relationships. Sexual self-disclo- attachment and sexuality systems are considered to be
sure promotes understanding of sexual preferences and may distinct components of romantic love (Bowlby, 1982;
allow partners to develop a sexual repertoire that is mutually Shaver, Hazan, & Bradshaw, 1988), sexual behaviors, pre-
satisfying (Simon & Gagnon, 1986). In addition to directly ferences, and motives can serve both attachment and sex-
improving sexual satisfaction, sexual communication ual needs (Davis, Shaver, & Vernon, 2004; Schachner &
enhances relationship quality by fostering closeness, inti- Shaver, 2004; Shaver & Mikulincer, 2006). These two
macy, acceptance, and safety (MacNeil & Byers, 2009). subsystems are also linked due to the role sexual experi-
Across studies, there is robust and consistent support for ence plays in promoting bonding and intimacy in relation-
the role of effective sexual communication in promoting ship (Schachner & Shaver, 2004). Sexual communication,
positive sexual and relationship outcomes (see review by in particular, is likely to be affected by attachment because
Byers & Rehman, 2014). Not only is sexual communication this communication tends to involve disclosures which are
one of the strongest predictors of sexual and relationship highly personal and sensitive and which can make an
satisfaction, it also influences the degree to which other risk individual feel vulnerable. Sexual issues also tend to be
factors, such as negative body image, predict sexual satisfac- avoided by romantic partners, more so than other domains
tion (Rehman, Fallis, & Byers, 2013). Despite these benefits, of communication (Anderson, Kunkel, & Dennis, 2011),
the degree of sexual communication in relationships can be possibly because partners tend to view sexual communica-
less than optimal. Even couples in long-term relationships tion as more risky. Thus, we would expect that individuals
often lack a good understanding of each other’s sexual likes who tend to view their own needs as legitimate and
and dislikes (MacNeil & Byers, 2009). Thus, it is important worthy, as a result of early developmental experiences
to investigate the barriers to such disclosure and to identify with a responsive, trustworthy, and caring caregiver,
factors that promote or hinder sexual communication. would be more likely to communicate those needs
In the current study, we examine how an individual’s effectively.
attachment style relates to communication about sexual
Attachment Theory and Adult Romantic Relationships
Correspondence should be addressed to Uzma S. Rehman, University of
Waterloo, Department of Psychology, 200 University Ave. West, Waterloo, Over the past few decades, attachment theory has
Ontario N2L 3G1, Canada. E-mail: [email protected] become one of the major theoretical perspectives for
 MCNEIL, REHMAN, AND FALLIS

studying romantic relationships (Feeney & Noller, 2004). Attachment and Sexual Communication
According to attachment theory, warm and responsive care-
One reason why people with insecure attachments have
givers promote the perspective that one’s needs are legiti-
worse sexual experiences is that they have more difficulty
mate and that others can be relied on for support and
communicating their needs to their partners (e.g., Collins &
caregiving (Bowlby, 1969, 1980). In contrast, caregivers
Feeney, 2000; Roberts & Noller, 1998). Research has shown
who are either unresponsive or inconsistent in responding
that insecure attachment is related to a variety of negative
to expressed needs promote the development of insecure
communication patterns in relationships. For example, both
attachment styles.
anxious and avoidant attachment have been shown to pre-
Research has shown there are two primary dimensions
dict lower levels of support-seeking behaviors (Collins &
of insecure attachment: an anxious dimension and an
Feeney, 2000). Insecure attachment is also related to more
avoidant dimension (Shaver & Mikulincer, 2002).
difficulty constructively dealing with conflicts in relation-
Anxious attachment can be conceptualized as a hyperacti-
ships (Creasey, 2002; Creasey & Hesson-McInnis, 2001). In
vation of the attachment system (Mikulincer & Shaver,
a study examining conflict in close relationships, anxious
2003, 2007), which results in individuals constantly seek-
attachment was associated with more stress and anxiety
ing closeness to others, while continuously monitoring
during discussions and more negative communication beha-
their relationships for signs of rejection (Cassidy &
viors (Simpson, Rholes, & Phillips, 1996). In the same
Berlin, 1994; Simpson, Ickes, & Grich, 1999). In contrast,
study, avoidant attachment was found to be associated
the avoidant dimension can be conceptualized as a deac-
with fewer positive behaviors, fewer displays of warmth,
tivation of the attachment system, with individuals coping
and fewer signs of support.
with an unresponsive environment by becoming more
Past research has focused primarily on examining how
self-reliant and not seeking out support from others during
attachment styles relate to self-reported sexual communica-
times of distress (e.g., Simpson, Rholes, & Nelligan,
tion (e.g., Davis et al., 2006; Feeney, Peterson, Gallois, &
1992). Avoidant attachment is also associated with dis-
Terry, 2000; Khoury & Findlay, 2014). For example, Timm
comfort with intimacy and closeness (Bartholomew &
and Keiley (2011) found that secure attachment predicted
Horowitz, 1991; Campbell, Simpson, Kashy, & Rholes,
better self-reported sexual communication. Furthermore,
2001).
they found that sexual communication fully mediated the
effect of attachment security on sexual satisfaction and
partially mediated the effect of attachment on relationship
Attachment and Sexuality satisfaction. Davis and her colleagues (2006) examined how
anxious and avoidant attachment relate to sexual commu-
Attachment styles have been linked to a variety of
nication in a large community sample. The results of their
sexual outcomes, including an individual’s sexual beha-
study showed that both anxious and avoidant attachment
vior, sexual identity, and sexual satisfaction (Fricker &
were associated with more inhibited sexual communication.
Moore, 2002; Morrison, Goodlin-Jones, & Urquiza,
In their follow-up study, Khoury and Findlay (2014) repli-
1997). Secure attachment is generally associated with posi-
cated these effects, showing significant correlations between
tive sexual outcomes, such as comfort with sexuality,
both anxious and avoidant attachment and self-reported
openness to sexual activities, and more positive emotions
sexual communication.
about sexual relationships (e.g., Birnbaum, Reis,
In the current study, we wanted to build on this work by
Mikulincer, Gillath, & Orpaz, 2006). In contrast, insecure
examining how attachment styles relate to observable com-
attachment has been linked to a variety of sexual problems.
munication behavior. Self-reports about sexual communica-
Specifically, both anxious and avoidant attachment predict
tion are informative for understanding how a person
lower levels of sexual satisfaction, less sexual arousal, and
perceives the effectiveness of his or her communication.
less sexual pleasure (e.g., Butzer & Campbell, 2008;
However, a number of factors, such as the individual’s
Khoury & Findlay, 2014).
self-awareness and introspection, influence these reports,
Although both dimensions of insecure attachment are
making self-reports more appropriately suited for examining
related to negative sexual outcomes, evidence suggests
an individual’s perceptions of communication. Furthermore,
each dimension has unique effects (Davis et al., 2006). For
it is well established that insecure attachment, the anxious
example, sex can be seen as a sign of love and desire (Davis
dimension specifically, is related to hypervigilance in rela-
et al., 2004), which may be validating for individuals high
tionships (Alexander, Feeney, Hohaus, & Noller, 2001; Ein-
in anxious attachment, while simultaneously making things
Dor, Mikulincer, & Shaver, 2011). This excessive monitor-
more uncomfortable for people who report higher avoidant
ing may result in an overreporting of communication pro-
attachment. These dimensions may also have different
blems or exaggeration of normal communication
effects on partners. For example, Butzer and Campbell
difficulties. Thus, it may be difficult to determine whether
(2008) found that the avoidant attachment dimension was
communication difficulties reported by individuals high in
a negative predictor of a partner’s sexual satisfaction, but
anxious or avoidant attachment reflect this hypervigilance to
anxious attachment was unrelated to partner ratings of sex-
relationship problems, or whether anxiously attached
ual satisfaction.

192
 ATTACHMENT AND SEXUAL COMMUNICATION

individuals tend to have relationships characterized by Table 1. Sample Demographics

poorer communication.
Demographics Men Women
The overall goal of the current study was to revisit the
association between attachment style and sexual communi- Age (years)
cation using a methodology that allowed us to directly M 40.3 38.2
observe couples’ sexual communication. To our knowledge, Range 23–67 23–68
SD 11.3 11.3
this study will be the first to examine how an individual’s Relationship length (years)
attachment style affects his or her partner’s sexual commu- M 13 12.5
nication. Previous research on sexual communication and Range 3.6–41.1 2.4–41.1
attachment has almost exclusively examined only one mem- SD 8.9 8.09
ber of a couple. Yet one individual’s attachment style may Education (years)
M 16.1 16.2
elicit certain communication behaviors from that person’s Range 9–24 9–26
partner. For example, if individuals high in anxious attach- SD 2.8 3.1
ment frequently engage in reassurance seeking during a
conversation, this may lead their partners to withdraw
(Christensen & Heavey, 1990). The dyadic design of the English at a minimum of an eighth grade level (to ensure
present study makes it possible to explore this question to they were able to respond accurately to the questions used in
see how attachment styles impact both the individual and the study); and (d) both members had to agree to participate.
the partner’s communication. See Table 1 for the demographic characteristics of the
current sample.
Hypotheses
In this study, we offered two hypotheses: Measures
Background Questionnaire. The background
H1: Based on consistent findings from studies of self- questionnaire gathered information about paricipants’
reported communication, we expected that higher general demographic characteristics (e.g., age, income,
levels of avoidant and anxious attachment would be educational achievement), relationship histories (e.g.,
associated with more negative and fewer positive com- marital status, relationship length), and sexual histories.
munication behaviors.
Experience in Close Relationship Scale—Short
H2: Based on the previous studies showing negative effects
Form. Both anxious and avoidant attachment were
for partners, it was predicted that both anxious and
measured using the 12-item version of the Experiences in
avoidant attachment would be associated with more
Close Relationship Scale—Short Form (ECR-S; Wei,
negative communication and fewer positive communi-
Russell, Mallinckrodt, & Vogel, 2007). Participants
cation from partners.
responded on a 7-point Likert scale from 1 (Strongly
disagree) to 7 (Strongly agree). Attachment avoidance was
Method measured by six items, including “I want to get close to my
partner, but I keep pulling back.” Attachment anxiety was
Participants measured by six items, including “My desire to be very
A convenience sample of 81 couples, in committed rela- close sometimes scares people away.” For females, the
tionships, participated in the current study (N = 162 indivi- anxiety items (Cronbach’s α = .75) and the avoidance
duals). All participating couples identified as heterosexual. items (Cronbach’s α = .84) were found to have good
Data for the current study were gathered as part of a larger, internal consistency. For males, for the anxiety items had
longitudinal study (described in the next section). The cur- good internal consistency (Cronbach’s α = .85), while the
rent study used data from the third wave of data collection avoidance items demonstrated modest but acceptable
(T3), which took place two years after their initial participa- internal consistency (Cronbach’s α = .69).
tion. Couples were originally recruited from the southwes-
tern area of Ontario. The study was advertised with posters Sexual Problems Questionnaire. The Sexual
placed in local businesses, offices of physicians, and offices Problems Questionnaire (SPQ) was developed for the
of mental health professionals. Advertisements were also current study and was used to identify topics for the
placed in local newspapers and in online classified ads discussion task described in the next section. The measure
(e.g., Kijiji). Couples therapists and sex therapists referred lists 25 sexual issues about which partners might disagree
approximately 2% of the couples that participated in the (e.g., “Frequency of sexual relations,” “Showing interest in
study. Couples were originally required to meet several having sex”). Participants were asked to rate the extent to
criteria: (a) either married or living together for a minimum which issues were problems in his or her relationship on a
of two years; (b) aged 21 to 65; (c) able to speak and read scale from 1 (Not at all) to 7 (Very much so). Items for the

193
 MCNEIL, REHMAN, AND FALLIS

SPQ were selected based on lists of sexual “concerns” and One of these discussions focused on an aspect of the
“problems” identified by Frank, Anderson, and Rubinstein couple’s sexual relationship in which the male partner
(1978) and Byers and MacNeil (1997), as well as wanted change and the other discussion focused on an
suggestions from therapists and clinical psychologists who aspect in which the female partner wanted change
provide couples therapy. The scale also included five blank (referred here onward as the “male topic discussion” and
spaces for couples to identify sexual problems that were not “female topic discussion”). The topics selected for the two
captured by the other items. In the last part of the SPQ, conflict discussions were based on problems that the male
participants are asked to list the “four most important” and female partners independently identified on the SPQ.
sexual problems in their relationship in order of If the male and female partner identified different topics
importance, with the top listed items representing the issue as their top-rated issue, then these two topics were
that is viewed by the participants as the biggest sexual selected for the two discussions. If both partners identified
problem in the relationship. the same issue as the top-rated problem, then the research
assistant considered the second highest issue identified by
both partners. If the issues ranked as second highest were
Procedure
different for the two partners, these topics were selected as
All study measures and procedures were reviewed and the discussion topic. If they were the same, the research
approved by the university’s research ethics board. assistants reviewed the third ranked issue for both part-
Participants were originally recruited for a prospective, ners, and so on.
two-year longitudinal study. At Time 1, participating cou- This protocol was developed with two goals in mind:
ples completed a lab assessment in which they answered a First, we wanted to ensure that partners were discussing
variety of measures and engaged in observational tasks different issues during their conversations. Second, we
designed to understand the role of interpersonal factors in wanted the topics of the discussions to be of equal ranking
sexual satisfaction and sexual functioning. One year after so that there were no systematic differences in the degree of
this participation, couples completed an interview over the importance of the female topic versus the male topic. The
phone (Time 2). One year after this phone conversation order of the two discussions was counterbalanced. Before
(two years after their initial session), couples participated each discussion, each research assistant informed the partner
in a second in-lab assessment (Time 3). Data from Time 3 with whom he or she was working about the topic of
were used for the current study. discussion to determine if the participant was willing to
A research assistant contacted the couples by telephone discuss that issue.
or e-mail to schedule the Time 3 assessment. Of the 113
couples that completed the Time 1 assessment, 84 couples
participated in the Time 3 assessment. In regard to those Discussion Task. Partners were seated across from
who did not participate at Time 3, eight couples did not each other (rather than adjacent) so that video cameras
participate because their relationship had ended, and 21 did mounted on the wall above each partner’s head could
not participate because they were not interested or could not capture the interaction using split-screen technology. One
be reached. Excluding the couples who were no longer in research assistant provided instructions for the discussion
the same relationship, the retention rate was 80%. Of the 84 task, while the other research assistant prepared to start the
couples who participated in the Time 3 assessment, two recording in a different room. Participants were given the
couples did not engage in the videotaped interaction, and following instructions:
one couple’s data were lost due to a technical issue. Thus,
the final sample of the current study was 81 couples. As you know you are going to be discussing (topic). You
will have eight minutes to discuss (topic). In a moment, I’m
During the Time 3 lab assessment, two trained research
going to leave you alone to start your discussion. We won’t
assistants worked individually with each couple. When a hear what you are saying while you have your discussion.
couple arrived at the lab, the research assistants reviewed After eight minutes have passed (research assistant’s name)
the information letter and consent forms with them. One will knock on the door to let you know that the time is up
research assistant was randomly assigned to work with each and you may end the discussion. Please try to discuss the
partner from that point forward. The male and female part- issue as naturally as possible, exactly as you would at home.
ners were separated into two rooms where they completed Do you have any questions or concerns?
study measures individually. Participants completed a series
of questionnaires, including the background questionnaire, The research assistant addressed any questions and then
the SPQ, and the ECR-S. Participants also completed addi- instructed the couple to begin their discussion when the
tional questionnaires unrelated to the current study, while door closed. The recording began when the door closed.
the research assistants used their responses to the SPQ to At the eight-minute mark, a research assistant knocked on
select the topics for the two sexual conflict discussions. For the door and ended the interaction. The procedure was then
the observational component of the study, partners were repeated for the second discussion topic.
brought together to engage in two discussions about sexual After completing the discussion tasks, participants com-
problems in their relationships. pleted questionnaires unrelated to the current study and

194
 ATTACHMENT AND SEXUAL COMMUNICATION

were debriefed. At the end of the study, participants were thoughts, and expressing hostile demands for change). The
given a feedback letter and a list of sexual health resources. behaviors coded under “hostile and domineering negative
In appreciation for their participation, each partner received affect” and “reactive/defensive affect” were based on nega-
CAN$50 for their time. The entire procedure took approxi- tive affect codes from the SPAFF (Gottman & Krokoff,
mately three hours. 1989). The dimension of “unskilled communication beha-
viors” was based on the coding system developed by Sillars
Coding. After data collection was complete, research and colleagues (1982), described previously.
assistants independently viewed each couple’s discussions. One research assistant coded all conversations and a
These research assistants were given detailed definitions, second research assistant independently coded 58 randomly
instructions, and training on each of the constructs being selected targets and conversations (35% of total conversa-
coded. During the first viewing, raters watched the tions) to examine interrater reliability. The interrater correla-
conversation in its entirety, paying attention to both tions for the six constructs ranged from .75 to .87
partners and both verbal and nonverbal behaviors. suggesting good overall reliability. When two raters were
Following this initial viewing, the raters then watched the available, ratings were averaged. Given the power con-
video again and rated a specific individual based on the straints in this study, coded variables were averaged across
frequency and intensity of a particular behavior or display. the same domain and gender for two conversations to
These ratings focused on three dimensions of positive improve power. This resulted in three measures of positive
communication and three dimensions of negative communication and three measures of negative communica-
communication and were rated on a 7-point scale ranging tion for each participant. An exploratory factor analysis was
from None to A great deal. used to examine the factor structure of these coded
The three positive dimensions were (a) displays of posi- variables.
tive affect (this dimension included expressions of affection,
warmth, validation, and understanding); (b) offering solu-
tions to the problem; and (c) responsiveness to their partner
Results
(this included behaviors such as soliciting information from
a partner, asking follow-up questions or encouraging a
Models were tested using Mplus version 7. Evaluations
partner to share his or her perspective). We focused on
of the models were conducted using a chi-square test of
positive affect based on the extensive literature demonstrat-
model fit, with a nonsignificant chi-square indicating good
ing that greater displays of positive affect in romantic rela-
overall model fit. In addition, models were evaluated using
tionship predict greater relationship satisfaction and stability
standard criteria of acceptable fit: comparative fit index
(see review by Ramsey & Gentzler, 2015). The specific
(CFI) > .90, and root mean square error of approximation
behaviors we coded under positive affect were based on
(RMSEA) < .06, with a 90% confidence interval (CI)
the Specific Affect Coding System (SPAFF; Gottman &
including .05 (Schermelleh-Engel, Moosbrugger, & Müller,
Krokoff, 1989), a widely used system for coding couple
2003).
interactions. We included the dimension of partner respon-
siveness in light of evidence demonstrating that this con-
struct is critically important to relationship well-being (e.g.,
Factor Analysis of Communication Variables
Reis, 2012). Further, past research has shown that observed
partner responsiveness can be coded in a reliable fashion For the six communication variables that were coded, we
(e.g., Birnbaum et al., 2016). Although not a central focus conducted an exploratory factor analysis (EFA) with a max-
of our research, we wanted to examine whether partners imum likelihood extraction method using Mplus version 7.
would be able to identify and discuss solutions to their We tested one- and two-factor solutions and compared the
existing sexual problems as it has been theorized that sexual models based on fit. A geomin rotation was used, which
problems may be perceived as particularly difficult to solve allowed the factors to correlate with one another. The correla-
(Metts & Cupach, 1989). Our code of “offering solutions” tion table used for this factor analysis is shown in Table 2.
was based on a coding system developed by Sillars, Coletti, The one-factor solution had a significant chi-square,
Parry, and Rogers (1982) that is designed to capture a range indicating poor model fit (χ2 = 56.37, df = 9, p < 0.001),
of conflict management strategies used in interpersonal good fit on CFI, and poor fit on RMSEA (CFI = .91,
communication. RMSEA = 0.18, 90% CI [0.14, 0.23]). The two-factor
The three negative dimensions were (a) hostile and dom- solution had a nonsignificant chi-square, suggesting the
ineering negative affect (this dimension included expressions model fit the data well (χ2 = 8.02, df = 4, p = 0.09). In
of contemptuous, hostile, and domineering behaviors); (b) addition, the two-factor solution had good fit on CFI but
reactive/defensive negative affect (this dimension included relatively poor fit on RMSEA (CFI = .99, RMSEA = 0.08,
expressions of annoyance, frustration, whining, and defen- 90% CI [0.00, 0.16]). Although RMSEA was high, small
siveness); and (c) unskilled communication behaviors (this sample size and low degrees of freedom can result in
included behaviors such as blaming one’s partner, rejecting a RMSEA being an unreliable measure of model fit (Kenny,
partner’s feelings, making assumptions about a partner’s Kaniskan, & McCoach, 2015).

195
 MCNEIL, REHMAN, AND FALLIS

Table 2. Zero-Order Correlations Between the Six Coded Communication Behaviors and Attachment, and Factor Loadings for Coded
Communication Variables

Factor 1: Factor 2:
Negative Positive
1 2 3 4 5 6 7 Communication Communication

1. Hostility — 0.88 0.03

2. Negative affect 0.73*** — 0.70 −0.21
3. Unskilled 0.77*** 0.77*** — 0.78 −0.17
communication
4. Positivity −0.55*** −0.66*** −0.63*** — −0.09 0.80
5. Responsiveness −0.43*** −0.48*** −0.56*** 0.63*** — 0.00 0.74
6. Solutions −0.26*** −0.31*** −0.31*** 0.46*** 0.41*** — 0.17 0.66
7. Anxious attachment 0.35*** 0.25** 0.20* −0.18* 0.03n.s. −0.03n.s. —
8 Avoidant attachment 0.41*** 0.39*** 0.34*** −0.38*** −0.19* −0.15n.s. 0.41***

Note. N = 162.
*p < .05; **p < .01; ***p < .001; n.s. = not significant.

Taken together, these results supported a two-factor solu- Model Overview

tion. The two factors suggest that the communication variables
Given the interdependent nature of couples’ data, we
could be divided into positive communication variables (offer-
analyzed our data with a variation of Kashy and Kenny’s
ing solutions, positive affect, and responsiveness) and negative
(1999) actor–partner interdependence model (APIM). The
communication (hostility, negative affect, and unskilled com-
APIM provides two effects estimates: an actor effect, which
munication behavior). The EFA indicated a correlation
estimates the effect of a participant’s predictor variable on
between the two factors of r = −.70. The factor loadings
his or her own dependent variable (e.g., the effect of the
generated for the two-factor solution are shown in Table 2.
wife’s level of anxious attachment on the wife’s commu-
Based on the EFA we created two composite measures, one
nication behavior), and a partner effect, which examines
for positive communication and one for negative communica-
how a partner’s score on a predictor variable affects an
tion behavior. These scores were created for both the male and
actor’s dependent variable (e.g., the effect of husband’s
female partners by averaging across the three positive dimen-
level of anxious attachment on the wife’s communication
sions and averaging across the three negative dimensions.
behavior).
These composite scores were used as a measure of partici-
In this study we used a variation of the APIM model,
pants’ overall communication skills. The positive dimension
shown in Figure 1. The model includes two predictor vari-
was shown to have good overall reliability for both women
ables for each person: anxious attachment and avoidant
(α = .72) and men (α = .74). The negative dimension also had
attachment. Previous research has shown that attachment is
good reliability for both women (α = .91) and men (α = .88).
highly correlated in individuals and within couples (Wei

Actor: Positive
Communication

Actor Anxious
Attachment

Actor: Negative
Actor Avoidant Communication
Attachment

Partner Anxious Partner: Positive

Attachment Communication

Partner Avoidant
Attachment

Partner: Negative
Communication

Figure 1. Hypothesized model linking anxious and avoidant attachment to positive and negative communication behavior.

196
 ATTACHMENT AND SEXUAL COMMUNICATION

Table 3. Zero-Order Correlations Between Attachment and Communication Behavior

1 2 3 4 5 6 7 8

1. Female positive communication —

2. Female negative communication −0.78*** —
3. Male positive communication 0.32** −0.38*** —
4. Male negative communication −0.41*** 0.59*** −0.68*** —
5. Female avoidant attachment −0.27** 0.40*** −0.35*** 0.48*** —
6. Female anxious attachment −0.09n.s. 0.28** −0.32** 0.40*** 0.49*** —
7. Male avoidant attachment −0.21* 0.44*** −0.37*** 0.53*** 0.51*** 0.49*** —
8. Male anxious attachment −0.22* 0.36*** −0.10n.s. 0.22* 0.44*** 0.28** 0.41*** —

Note. N = 81.
*p < .05; **p < .01; ***p < .001; n.s. = not significant.

et al., 2007), so all predictor variables were allowed to dimension were significantly less likely to engage in posi-
freely correlate. The current model also included two depen- tive communication behaviors and significantly more likely
dent variables for each person: positive and negative com- to use negative communication strategies when discussing
munication behaviors. Based on previous research, the sexual problems in their relationship.
positive and negative communication behaviors were
allowed to correlate in the model (Crawford & Henry,
Partner Effects for Avoidant Attachment
2004). The correlation matrix for the variables used in our
model is shown in Table 3. Our results revealed a significant partner effect for avoidant
To test our model with a sufficient sample size (Bentler attachment and negative communication behavior, with avoid-
& Chou, 1987), we did not examine gender differences in ance predicting more negative communication behavior by the
our model. To examine gender differences in our model, we partner when discussing sexual problems (β = 0.27, SE = 0.08,
would be required to divide our sample in half. p < 0.001). There was no significant association between an
Furthermore, this divide would increase the complexity of individual’s level of avoidant attachment and his or her part-
our model, which would result in insufficient power. ner’s use of positive communication behaviors during discus-
Research has suggested that underpowered studies can sion of sexual problems (β = −0.13, SE = 0.08, p = 0.11).
lead to erroneous findings (Francis, 2012); therefore, actor
and partner effects were constrained to be equal across
gender. Furthermore, previous studies that have examined Actor Effects for Anxious Attachment
the association between sexual communication and attach- In contrast to the findings seen with the avoidant dimen-
ment, and have reported on gender differences, either did sion, the anxious attachment dimension was not a significant
not find any differences (Timm & Keiley, 2011) or found predictor of an individual’s own positive (β = 0.12, SE = 0.08,
only minor sex differences which did not meaningfully p = 0.14) or negative (β = −0.04, SE = 0.08, p = 0.62) com-
improve model fit (Davis et al., 2006). munication behaviors. That is, there was no association
between anxious attachment and the use of positive or negative
communication behaviors during sexual problem discussions.
Model Results
The proposed theoretical model, shown in Figure 1, was
Partner Effects for Anxious Attachment
tested to examine how well this model fit the data. This
initial model was shown to have good model fit based on There were no significant partner effects for anxious attach-
the χ2 results (χ2 = 12.55, df = 12, p = 0.4) and excellent fit ment and positive communication (β = −0.11, SE = 0.08,
on other fit indices (CFI = .997, RMSEA = 0.024, 90% CI p = 0.19) or negative communication (β = 0.13, SE = 0.08,
[0.00, 0.117]). Because this model was shown to have good p = 0.11). Thus, on average, partners of individuals with higher
fit, it was used to examine the relationship between attach- levels of anxious attachment were no more likely to use
ment and sexual communication. The parameter estimates negative and positive communication when discussing sexual
for this model are shown in Figure 2. problems in their relationships.

Actor Effects for Avoidant Attachment

Discussion
Higher levels of avoidant attachment significantly pre-
dicted greater levels of negative communication (β = 0.30, The present study examined how attachment styles relate
SE = 0.08, p < 0.001) and lower levels of positive commu- to how couples communicate about sexual problems in their
nication (β = −0.25, SE = 0.09, p = 0.005). Thus, indivi- relationships. Previous research has shown that, compared
duals who scored higher on the avoidant attachment to secure attachment, insecure attachment styles are related

197
 MCNEIL, REHMAN, AND FALLIS

Figure 2. Standardized path coefficients for the model. Actor and partner effects are constrained to be equal across gender; *p < .05; **p < .01; ***p < .001;
n.s. = not significant.

to less effective sexual communication (Davis et al., 2006; First, our study showed that anxious attachment was not
Khoury & Findlay, 2014; Timm & Keiley, 2011). Self- significantly related to sexual communication, either posi-
reports provide insight into how an individual conceptua- tively or negatively. These results are inconsistent with
lizes his or her communication, but with self-report metho- some previous work showing that anxious attachment is
dology it is not possible to determine whether the report related to self-reported sexual communication problems
reflects a perceptual bias or whether there are actual com- (Davis et al., 2006; Khoury & Findlay, 2014). This incon-
munication difficulties, such as increased negativity in the sistency may be the result of individuals high in anxious
tone of the discussion. This issue becomes particularly attachment being more hypervigilant. Hypervigilance may
problematic when examining the association between lead these individuals to be sensitive to noticing normal
attachment style and communication of sexual needs and communication difficulties, which results in them reporting
desires because attachment style could likely influence how more problems when communication is evaluated using
an individual perceives the legitimacy of his or her sexual self-report measures. Alternatively, the communication
needs, how easy it is to express those needs, and how he or codes used in this study may not capture other negative
she reacts to perceived responsiveness or dismissal of those behaviors that could be associated with anxious attachment,
needs by a sexual partner. Thus, even though the actual such as excessive reassurance seeking (Shaver, Schachner,
process of communication may not look different for indi- & Mikulincer, 2005). It is also possible there is a small
viduals with different attachment styles, certain attachment effect of anxious attachment on communication behavior
styles, such as anxious attachment, may make it more likely that our study does not have the power to detect.
for an individual to perceive the communication more In addition to examining the effect of anxious attachment
negatively. on one’s own communication, our study examined how
In this study, we addressed this concern by utilizing an one’s anxious attachment relates to a partner’s communica-
observational method to assess sexual communication. tion behaviors. We found no significant association between
Couples were video-recorded having conversations about anxious attachment in one partner and the other partner’s
sexual difficulties in their relationships. These videos were communication behavior. Our results complement previous
then used to examine how levels of anxious and avoidant research on attachment and sexual satisfaction that sug-
attachment related to observable communication. Our ana- gested anxious attachment is unrelated to partner sexual
lyses revealed several important findings about attachment satisfaction (Butzer & Campbell, 2008).
styles and sexual communication behavior. Similar to findings from self-report studies that show that
avoidant attachment is associated with worse sexual

198
 ATTACHMENT AND SEXUAL COMMUNICATION

communication (Davis et al., 2006; Khoury & Findlay, In addition, our results for anxious attachment did not
2014), the observational data we gathered suggest that indi- replicate findings from past work. Specifically, whereas
viduals high in avoidant attachment express themselves some past studies have found that anxious attachment is
more negatively and less positively when discussing sexual related to worse sexual communication, our data did not
concerns with their partner. Furthermore, our study showed find such an association. We reasoned that this discrepancy
that avoidant attachment is associated with more negative was due to methodological differences between past studies
communication from relationship partners. We did not find and our study. However, this claim would be strengthened
any association between attachment avoidance and positive by demonstrating that the discrepancy between perceived
communication behaviors by the partner. It is possible there sexual communication (as measured by self-report) and
is a small partner effect for positivity, but the effect required observed sexual communication is greater for individuals
more power to detect. high in anxious attachment as compared to individuals high
Overall, our results suggest that only the avoidance in avoidant or secure attachment.
dimension of attachment insecurity is related to less effec- It is also important to highlight that we investigated a
tive sexual communication. One reason avoidant attachment specific type of sexual communication: sexual partners dis-
may be more relevant in these conversations might be that cussing aspects of their sexual relationship with which they
the discussions are centered on difficulties in the sexual were dissatisfied. This is an important domain of commu-
domain. Sex is one of the most intimate acts in relation- nication in intimate relationships, as it is inevitable that
ships, and avoidant attachment involves discomfort with sexual partners will face challenges to their sexual relation-
closeness and intimacy (Bartholomew & Horowitz, 1991); ship that need to be negotiated (e.g., discrepancies in desire,
therefore, these conversations may be especially stressful for or different preferences; Day, Muise, Joel, & Impett, 2015).
individuals high in avoidance. Furthermore, previous find- However, it is unclear if our findings would also generalize
ings suggest that avoidant attachment is associated with an to other types of sexual communication (e.g., a pure dis-
elevated stress response when engaging in conflict discus- closure task in which partners share their sexual likes and
sion, because individuals high in avoidant attachment are dislikes) and to other types of nonsexual communication
prevented from using avoidance strategies (Dozier & between romantic partners. This question also relates to
Kobak, 1992). Another potential reason for the observed the underlying mechanism that explains the association
effects may be that individuals higher in avoidant attach- between attachment avoidance and sexual communication.
ment may predict less benefit from engaging in conversa- We have argued that individuals high on avoidant attach-
tions about sexual difficulties (Gere, MacDonald, Joel, ment may demonstrate less skill during sexual communica-
Spielmann, & Impett, 2013). tion because they fear intimacy, and sexual communication
Our results also suggest anxious attachment might be less tends to be an intimate type of communication. To test this
impactful on sexual disagreements. One potential explana- mechanism, we would need to design observational tasks
tion for why we are not finding significant effects could be that vary in the degree of intimacy and then demonstrate
that sexual disagreements are inherently threatening and that (a) individuals high in attachment avoidance tend to
anxiety provoking for most people. If most people are feel- perform more poorly in tasks that are rated as higher in their
ing highly anxious and threatened during a sexual disagree- potential for evoking intimacy and (b) this finding is specific
ment, people high in anxious attachment may be responding to individuals who are high in attachment avoidance, as
in a more typical way. compared to individuals high in attachment anxiety and
securely attached individuals.
Finally, in the current study, we characterized the inter-
Limitations and Future Directions
action by the mean level of negativity and positivity
This study provides novel insights into how attachment expressed by each partner. While this allows us to view
styles relate to discussions about sexual difficulties. The the conversation as a whole, it is important to note that
observational nature of this study offered many benefits this leaves out a rich source of information related to how
for examining our key questions. For example, it provides partners’ communication behavior changes over the course
a view of what a typical conversation would look like in the of a conversation. By examining how each partner’s beha-
home for both partners (Foster, Caplan, & Howe, 1997; vior changes over time and examining how partners influ-
Gottman & Krokoff, 1989) and allows an independent ence each other over the course of an interaction, we would
rater to judge the effectiveness of the communication. be able to capture the dynamic aspects of the conversation.
There are a number of ways in which the current research For example, are individuals high in avoidance more likely
could be replicated and extended in future research. As we to express negative behavior or to suppress positive beha-
noted, past studies of sexual communication have measured vior in response to a partner’s bid for closeness and inti-
sexual communication using self-report methods, whereas macy? By examining moment-to-moment fluctuations in
we used observational techniques. In future work, it would positive and negative affect, we would be able to investigate
be worthwhile to assess sexual communication using both this question at the macro and micro levels.
methods together. This would allow us to answer questions When interpreting the results of the current study, some
that neither methodology can answer on its own. limitations need to be kept in mind. In our analyses, we

199
 MCNEIL, REHMAN, AND FALLIS

were able to capitalize on the dyadic nature of the study to Personality and Social Psychology, 111(4), 530–546. doi:10.1037/
increase statistical power, but our overall sample size lim- pspi0000069
Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. New York,
ited some of the analyses we were able to conduct. Most NY: Basic Books.
significantly, we did not have sufficient power to test for Bowlby, J. (1980). Attachment and loss. New York, NY: Basic Books.
gender differences, and future observational studies will Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). New
need to examine whether gender moderates any of the find- York, NY: Basic Books.
ings that we obtained in the current study. Butzer, B., & Campbell, L. (2008). Adult attachment, sexual satisfaction,
and relationship satisfaction: A study of married couples. Personal
In addition, this study focused exclusively on couples in Relationships, 15(1), 141–154. doi:10.1111/pere.2008.15.issue-1
long-term relationships and how they navigate sexual diffi- Byers, E. S., & MacNeil, S. (1997). The relationships between sexual
culties. This focus means that it is still unclear how attach- problems, communication, and sexual satisfaction. The Canadian
ment styles might affect sexual negotiations for newer Journal of Human Sexuality, 6(4), 277.
couples, who are still in the process of establishing sexual Byers, E. S., & Rehman, U. (2014). Sexual well-being. In D. Tolman & L.
Diamond (Eds.), APA handbook of sexuality and psychology (Vol. 1,
relationships. It is possible that anxious attachment may be pp. 317–337). Washington, DC: American Psychological Association.
more important in these early relationships because there is a Campbell, L., Simpson, J. A., Kashy, D. A., & Rholes, W. S. (2001).
higher likelihood of rejection when a relationship is still new. Attachment orientations, dependence, and behavior in a stressful situa-
Finally, our sample is fairly homogenous in ethnic back- tion: An application of the actor-partner interdependence model.
ground, and the couples in our sample reported high levels Journal of Social and Personal Relationships, 18(6), 821–843.
doi:10.1177/0265407501186005
of sexual and relationship satisfaction. Thus, our findings Cassidy, J., & Berlin, L. J. (1994). The insecure/ambivalent pattern of
may not generalize to more ethnically diverse samples and attachment: Theory and research. Child Development, 65(4), 971–
to couples lower in relationship and sexual satisfaction. In 991. doi:10.2307/1131298
future work, it will be important to examine these questions Christensen, A., & Heavey, C. L. (1990). Gender and social structure in the
in samples characterized by greater diversity on these demand/withdraw pattern of marital conflict. Journal of Personality
and Social Psychology, 59(1), 73. doi:10.1037/0022-3514.59.1.73
dimensions. Collins, N. L., & Feeney, B. C. (2000). A safe haven: An attachment theory
perspective on support seeking and caregiving in intimate relation-
ships. Journal of Personality and Social Psychology, 78(6), 1053.
Conclusion doi:10.1037/0022-3514.78.6.1053
Crawford, J. R., & Henry, J. D. (2004). The Positive and Negative Affect
In conclusion, this study highlights the important role Schedule (PANAS): Construct validity, measurement properties, and
attachment styles play in relationship processes and how normative data in a large nonclinical sample. British Journal of
they are especially relevant to the sexual domain. Clinical Psychology, 43(3), 245–265. doi:10.1348/0144665031752934
Specifically, it expands on previous literature to show that Creasey, G. (2002). Associations between working models of attach-
ment and conflict management behavior in romantic couples.
avoidant attachment is a particular barrier to effective com- Journal of Counseling Psychology, 49(3), 365. doi:10.1037/0022-
munication, for both people high in avoidant attachment and 0167.49.3.365
their partners. Understanding these barriers is an essential Creasey, G., & Hesson-McInnis, M. (2001). Affective responses, cognitive
part of helping couples improve their sexual communication. appraisals, and conflict tactics in late adolescent romantic relation-
ships: Associations with attachment orientations. Journal of
Counseling Psychology, 48(1), 85. doi:10.1037/0022-0167.48.1.85
Davis, D., Shaver, P. R., & Vernon, M. L. (2004). Attachment style and
References subjective motivations for sex. Personality and Social Psychology
Bulletin, 30(8), 1076–1090. doi:10.1177/0146167204264794
Davis, D., Shaver, P. R., Widaman, K. F., Vernon, M. L., Follette, W. C., &
Alexander, R., Feeney, J., Hohaus, L., & Noller, P. (2001). Attachment style
and coping resources as predictors of coping strategies in the transition Beitz, K. (2006). “I can’t get no satisfaction”: Insecure attachment,
to parenthood. Personal Relationships, 8(2), 137–152. doi:10.1111/ inhibited sexual communication, and sexual dissatisfaction. Personal
Relationships, 13(4), 465–483. doi:10.1111/pere.2006.13.issue-4
pere.2001.8.issue-2
Day, L. C., Muise, A., Joel, S., & Impett, E. A. (2015). To do it or not to do
Anderson, M., Kunkel, A., & Dennis, M. R. (2011). “Let’s (not) talk about
that”: Bridging the past sexual experiences taboo to build healthy it? How communally motivated people navigate sexual interdepen-
dence dilemmas. Personality and Social Psychology Bulletin, 41(6),
romantic relationships. Journal of Sex Research, 48(4), 381–391.
791–804. doi:10.1177/0146167215580129
doi:10.1080/00224499.2010.482215
Dozier, M., & Kobak, R. R. (1992). Psychophysiology in attachment inter-
Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among
young adults: A test of a four-category model. Journal of Personality views: Converging evidence for deactivating strategies. Child
Development, 63(6), 1473–1480. doi:10.2307/1131569
and Social Psychology, 61(2), 226–244. doi:10.1037/0022-
Ein-Dor, T., Mikulincer, M., & Shaver, P. R. (2011). Attachment insecu-
3514.61.2.226
Bentler, P. M., & Chou, C. P. (1987). Practical issues in structural modeling. rities and the processing of threat-related information: Studying the
Sociological Methods and Research, 16(1), 78–117. doi:10.1177/ schemas involved in insecure people’s coping strategies. Journal of
Personality and Social Psychology, 101(1), 78. doi:10.1037/a0022503
0049124187016001004
Feeney, J. A., & Noller, P. (2004). Attachment and sexuality in close
Birnbaum, G. E., Reis, H. T., Mikulincer, M., Gillath, O., & Orpaz, A.
(2006). When sex is more than just sex: Attachment orientations, relationships. In J. H. Harvey, A. Wenzel, & S. Sprecher (Eds.), The
handbook of sexuality in close relationships (pp. 183–201). Mahwah,
sexual experience, and relationship quality. Journal of Personality
NJ: Erlbaum.
and Social Psychology, 91(5), 929. doi:10.1037/0022-3514.91.5.929
Feeney, J. A., Peterson, C., Gallois, C., & Terry, D. J. (2000). Attachment style as
Birnbaum, G. E., Reis, H. T., Mizrahi, M., Kanat-Maymon, Y., Sass, O., &
Granovski-Milner, C. (2016). Intimately connected: The importance of a piedictor of sexual attitudes and behavior in late adolescence. Psychology
and Health, 14(6), 1105–1122. doi:10.1080/08870440008407370
partner responsiveness for experiencing sexual desire. Journal of

200
 ATTACHMENT AND SEXUAL COMMUNICATION

Foster, D. A., Caplan, R. D., & Howe, G. W. (1997). Representativeness of Loving (Eds.), Interdisciplinary research on close relationships: The
observed couple interaction: Couples can tell, and it does make a case for integration (pp. 27–52). Washington, DC: American
difference. Psychological Assessment, 9(3), 285. doi:10.1037/1040- Psychological Association.
3590.9.3.285 Roberts, N., & Noller, P. (1998). The associations between adult attachment
Francis, G. (2012). Publication bias and the failure of replication in experi- and couple violence: The role of communication patterns and relation-
mental psychology. Psychonomic Bulletin and Review, 19(6), 975– ship satisfaction. In J. Simpson & W. Rholes (Eds.), Attachment theory
991. doi:10.3758/s13423-012-0322-y and close relationships (pp. 317–350). New York, NY: Guilford Press.
Frank, E., Anderson, C., & Rubinstein, D. (1978). Frequency of sexual Schachner, D. A., & Shaver, P. R. (2004). Attachment dimensions and
dysfunction in normal couples. New England Journal of Medicine, sexual motives. Personal Relationships, 11(2), 179–195. doi:10.1111/
299(3), 111–115. doi:10.1056/NEJM197807202990302 pere.2004.11.issue-2
Fricker, J., & Moore, S. (2002). Relationship satisfaction: The role of love Schermelleh-Engel, K., Moosbrugger, H., & Müller, H. (2003). Evaluating
styles and attachment styles. Current Research in Social Psychology, 7 the fit of structural equation models: Tests of significance and descrip-
(11), 182–204. tive goodness-of-fit measures. Methods of Psychological Research
Gere, J., MacDonald, G., Joel, S., Spielmann, S. S., & Impett, E. A. (2013). Online, 8(2), 23–74. doi:10.1016/S0065-2601(03)01002-5
The independent contributions of social reward and threat perceptions Shaver, P. R., Hazan, C., & Bradshaw, D. (1988). Love as attachment: The
to romantic commitment. Journal of Personality and Social integration of three behavioral systems. In R. J. Sternberg & M.
Psychology, 105(6), 961. doi:10.1037/a0033874 Barnes (Eds.), The anatomy of love (pp. 68–98). New Haven, CT:
Gottman, J. M., & Krokoff, L. J. (1989). Marital interaction and satisfac- Yale University Press.
tion: A longitudinal view. Journal of Consulting and Clinical Shaver, P. R., & Mikulincer, M. (2002). Attachment-related psychody-
Psychology, 57(1), 47–52. doi:10.1037/0022-006X.57.1.47 namics. Attachment and Human Development, 4(2), 133–161.
Kashy, D. A., & Kenny, D. A. (1999). The analysis of data from dyads and doi:10.1080/14616730210154171
groups. In H. T. Reis & C. M. Judd (Eds.), Handbook of research Shaver, P. R., & Mikulincer, M. (2006). Attachment theory, individual
methods in social psychology (pp. 451–477). New York, NY: psychodynamics, and relationship functioning. In D. Perlman & A.
Cambridge University Press. Vangelisti (Eds.), The Cambridge handbook of personal relationships
Kenny, D. A., Kaniskan, B., & McCoach, D. B. (2015). The performance of (pp. 251–271). New York, NY: Cambridge University Press.
RMSEA in models with small degrees of freedom. Sociological Methods Shaver, P. R., Schachner, D. A., & Mikulincer, M. (2005). Attachment
and Research, 44(3), 486–507. doi:10.1177/0049124114543236 style, excessive reassurance seeking, relationship processes, and
Khoury, C. B., & Findlay, B. M. (2014). What makes for good sex? The depression. Personality and Social Psychology Bulletin, 31(3), 343–
associations among attachment style, inhibited communication, and 359. doi:10.1177/0146167204271709
sexual satisfaction. Journal of Relationships Research, 5(7), 1–11. Sillars, A. L., Coletti, S. F., Parry, D., & Rogers, M. A. (1982). Coding verbal
doi:10.1017/jrr.2014.7 conflict tactics: Nonverbal and perceptual correlates of the “avoidance–
MacNeil, S., & Byers, E. S. (2009). Role of sexual self-disclosure in the distributive–integrative” distinction. Human Communication Research,
sexual satisfaction of long-term heterosexual couples. Journal of Sex 9(1), 83–95. doi:10.1111/hcre.1982.9.issue-1
Research, 46(1), 3–14. doi:10.1080/00224490802398399 Simon, W., & Gagnon, J. H. (1986). Sexual scripts: Permanence and
Metts, S., & Cupach, W. R. (1989). Situational influence on the use of change. Archives of Sexual Behavior, 15(2), 97–120. doi:10.1007/
remedial strategies in embarrassing predicaments. Communications BF01542219
Monographs, 56(2), 151–162. doi:10.1080/03637758909390256 Simpson, J. A., Ickes, W., & Grich, J. (1999). When accuracy hurts:
Mikulincer, M., & Shaver, P. R. (2003). The attachment behavioral system Reactions of anxious–ambivalent dating partners to a relationship-
in adulthood: Activation, psychodynamics, and interpersonal pro- threatening situation. Journal of Personality and Social Psychology,
cesses. Advances in Experimental Social Psychology, 35, 53–152. 76(5), 754. doi:10.1037/0022-3514.76.5.754
doi:10.1016/S0065-2601(03)01002-5 Simpson, J. A., Rholes, W. S., & Nelligan, J. S. (1992). Support seeking
Mikulincer, M., & Shaver, P. R. (2007). Attachment in adulthood: and support giving within couples in an anxiety-provoking situation:
Structure, dynamics, and change. New York, NY: Guilford Press. The role of attachment styles. Journal of Personality and Social
Morrison, T. L., Goodlin-Jones, B. L., & Urquiza, A. J. (1997). Attachment Psychology, 62(3), 434–446. doi:10.1037/0022-3514.62.3.434
and the representation of intimate relationships in adulthood. Journal Simpson, J. A., Rholes, W. S., & Phillips, D. (1996). Conflict in close
of Psychology, 131(1), 57–71. doi:10.1080/00223989709603504 relationships: An attachment perspective. Journal of Personality and
Ramsey, M. A., & Gentzler, A. L. (2015). An upward spiral: Bidirectional Social Psychology, 71(5), 899–914. doi:10.1037/0022-3514.71.5.899
associations between positive affect and positive aspects of close Timm, T. M., & Keiley, M. K. (2011). The effects of differentiation of self,
relationships across the life span. Developmental Review, 36, 58– adult attachment, and sexual communication on sexual and marital
104. doi:10.1016/j.dr.2015.01.003 satisfaction: A path analysis. Journal of Sex and Marital Therapy, 37
Rehman, U. S., Fallis, E., & Byers, E. S. (2013). Sexual satisfaction in (3), 206–223. doi:10.1080/0092623X.2011.564513
heterosexual women. In D. Castaneda (Ed.), An essential handbook of Wei, M., Russell, D. W., Mallinckrodt, B., & Vogel, D. L. (2007). The
women’s sexuality (Vol. 1, pp. 25–45). Westport, CT: Praeger. Experiences in Close Relationship Scale (ECR)—Short Form:
Reis, H. (2012). Perceived partner responsiveness as an organizing theme Reliability, validity, and factor structure. Journal of Personality
for the study of relationships and well-being. In L. Compbell & T. J. Assessment, 88(2), 187–204. doi:10.1080/00223890701268041

201
Copyright of Journal of Sex Research is the property of Routledge and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.