Personality and Individual Differences 70 (2014) 239245

Contents lists available at ScienceDirect

Personality and Individual Differences

journal homepage: www.elsevier.com/locate/paid

Emotional avoidance and rumination as mediators of the relation

between adult attachment and emotional disclosure q
Angela M. Garrison a,, Jeffrey H. Kahn b, Steven A. Miller c, Eric M. Sauer a
a
Western Michigan University, Department of Counselor Education and Counseling Psychology, United States
b
Illinois State University, Department of Psychology, United States
c
Rosalind Franklin University of Medicine and Science, Department of Psychology, United States

a r t i c l e i n f o a b s t r a c t

Article history: The authors evaluated emotional avoidance as a mediator of the relation between attachment avoidance
Received 19 May 2014 (i.e., fear of dependency) and emotional disclosure and rumination as a mediator of the relation between
Received in revised form 3 July 2014 attachment anxiety (i.e., fear of rejection) and emotional disclosure. Two operational denitions were
Accepted 4 July 2014
used for each of three variables emotional avoidance, rumination, and emotional disclosure such that
Available online 1 August 2014
hypotheses were tested on generalized self-appraisals and responses to specic emotional events. College
students (N = 116) rst completed generalized self-report measures of attachment, expressive suppres-
Keywords:
sion (i.e., emotional avoidance), rumination, and emotional-disclosure tendencies. Then, during a 7-day
Attachment avoidance
Attachment anxiety
diary study, they provided daily reports of emotional avoidance, rumination, and disclosure concerning
Emotional avoidance the days most unpleasant event. Attachment avoidance was negatively related to disclosure tendencies
Rumination and daily-event disclosure; emotional avoidance was supported as a mediator in the generalized self-
Emotional disclosure report analyses. Attachment anxiety was positively related to both measures of rumination, and daily-
event rumination was positively related to daily-event disclosure, but mediation was not supported in
either analysis. The ndings suggest implications for theories of attachment and emotion regulation.
2014 Elsevier Ltd. All rights reserved.

1. Introduction experience. That is, individuals differ in their tendency to

In Western cultures, when people are upset, they often look for
someone with whom to talk. Such emotional disclosure (i.e., talking
about ones distress) often increases in concordance with the
amount of distress experienced, such that emotionally intense
experiences are disclosed more than emotionally trivial experi-
ences (Cano, Leong, Williams, May, & Lutz, 2012; Garrison, Kahn,
Sauer, & Florczak, 2012). Although disclosure of ones distress is
not always adaptive, in general, emotional disclosure tends to be
benecial (Frattaroli, 2006; Kennedy-Moore & Watson, 2001). For
example, emotional disclosure to others can reduce intrusive
thoughts (Lepore, Ragan, & Jones, 2000) and reduce the intensity
of the emotion (Zech & Rim, 2005).
Despite the advantages of disclosing distress, many individuals
do not disclose distress even in the face of an emotionally intense
q
Portions of this research were presented at the 120th Annual Convention of the
American Psychological Association, Orlando, FL.
Corresponding author. Address: Western Michigan University, Department of
Counselor Education and Counseling Psychology, 1903 W. Michigan Ave., Kalamazoo,
MI 49008-5226, United States.
E-mail address: angela.m.garrison@wmich.edu (A.M. Garrison).
disclose personally distressing information across situations.
Low-disclosing individuals experience poorer well-being than indi-
viduals who tend to disclose distress, including lower levels of
social support and self-esteem and higher levels of depression
and loneliness (see Kahn, Hucke, Bradley, Glinski, & Malak, 2012).
Because of these negative outcomes, there is value in examining
the processes that are associated with problematic levels of
emotional disclosure.
1.1. Attachment orientation and emotional disclosure
Attachment theory may provide valuable insight into under-
standing individual differences in emotional disclosure. Among
adults, attachment is typically conceptualized along two dimen-
sions (Brennan, Clark, & Shaver, 1998): attachment avoidance
(i.e., fear of dependency, need for self-reliance) and attachment
anxiety (i.e., fear of rejection, need for approval). Individuals high
in attachment anxiety or avoidance (i.e., those with insecure
attachment) have distinct patterns of emotion regulation
(Mikulincer & Shaver, 2007), and these patterns of emotion regula-
tion have theoretical implications for emotional disclosure.

http://dx.doi.org/10.1016/j.paid.2014.07.006
0191-8869/ 2014 Elsevier Ltd. All rights reserved.
240 A.M. Garrison et al. / Personality and Individual Differences 70 (2014) 239245
1.1.1. Attachment avoidance
Attachment avoidance develops when an attachment gure is
not available. Because it does no good to engage in attachment
behaviors, individuals high in attachment avoidance deactivate
their distress (Mikulincer & Shaver, 2007). For example, when dis-
tressed they may use expressive suppression (Gross & John, 2003)
whereby they inhibit the behavioral expression of distress.
Expressive suppression and emotional disclosure, while strongly,
negatively correlated, are not merely opposites (Kahn et al.,
2012). Behavioral expression typically occurs as part of the
momentary experience of emotion (Averill, 1997), whereas talking
about an emotion often occurs well after the emotion episode.
Thus, expressive suppression may be viewed as the inhibition of
nonverbal expressions of emotion in the moment, whereas
emotional disclosure refers to the verbal sharing of emotion at
some later point in time. Because individuals high in attachment
avoidance are attempting to avoid distress, they would have the
opportunity to engage in expressive suppression before they had
the opportunity to engage in emotional disclosure. Thus, emotional
avoidance/suppression is a potential mediating variable in the
relation between attachment avoidance and emotional disclosure.
Empirical research supports the bivariate associations in this
theory-based mediation model. First, there is evidence that
avoidantly attached individuals use suppression as an emotion
regulation strategy (Caldwell & Shaver, 2012; Fraley & Shaver,
1997). For example, when Roisman, Tsai, and Chiang (2004) asked
participants to discuss potentially negative childhood experiences,
only the avoidantly attached individuals showed increases in
physiological activity that signied effortful behavioral inhibition.
Second, expressive suppression is strongly, negatively associated
with emotional-disclosure tendencies (Kahn et al., 2012). Finally,
a negative association between attachment avoidance and emo-
tional-disclosure tendencies has been supported empirically
(Garrison et al., 2012; Wei, Russell, & Zakalik, 2005). Unfortunately,
none of these studies has examined the possible mediating relation
between attachment avoidance and emotional disclosure, as med-
iated by expressive suppression. Doing so was one purpose of this
research.
1.1.2. Attachment anxiety
Unlike attachment avoidance, the emotion-regulation strategy
involved with attachment anxiety is hyperactivation, that is, the
exaggeration of attachment strategies such as monitoring of rela-
tionship partners and proximity seeking (Mikulincer & Shaver,
2007). This develops because an attachment gure is nearby but
is not consistently responsive to the attachment behaviors;
thus, the behaviors are amplied. However, because anxiously
attached individuals fear rejection, they may also engage in
suppression which, if ineffective, will result in rumination
(Liverant, Kamholz, Sloan, & Brown, 2011). Indeed, rumination is
a common form of hyperactivation found among anxiously
attached individuals (along with other forms of emotion-focused
coping; see Mikulincer & Shaver, 2007). Rumination involves
repetitively thinking about distressing emotional experiences
(Nolen-Hoeksema, 2000). For example, Saffrey and Ehrenberg
(2007) found that individuals high in attachment anxiety experi-
enced increased levels of both general rumination and relation-
ship-specic rumination following a recent relationship breakup.
Consistent with the theory of social sharing, this rumination, in
turn, prompts individuals to share their distress with others
(Rim, 2009). Thus, attachment theory and the theory of social
sharing combine to suggest that, when faced with a distressing
event, individuals high in attachment anxiety would engage in
heightened rumination which would, in turn, increase their
likelihood of engaging in emotional disclosure.
Despite the plausibility of this theoretical statement, much of
the empirical research on attachment anxiety and emotional
disclosure has found a small, negative association (e.g., Pistole,
1993; Wei et al., 2005), not a positive one as theory would predict.
We believe that this divergence between theory and research is a
function of the conceptualization of emotional disclosure as a
stable individual-difference tendency versus as an act in response
to a specic event. For example, Wei et al. measured emotional-
disclosure tendencies using generalized self-reports and found
attachment anxiety to be negatively related to emotional disclo-
sure. Generalized self-reports likely tap into a lifetime of memories
of being unable to disclose distress because of inconsistent attach-
ment gures as well as cognitive expectations that disclosure is not
always possible. By contrast, Tan, Overall, and Taylor (2012)
measured the disclosure of specic events that had occurred over
the past week, and they found a positive association between
attachment avoidance and disclosure. Similarly, in a 7-day diary
study, Garrison et al. (2012) found a positive association between
attachment avoidance and the disclosure of intense emotional
events that occurred earlier in the day. Thus, understanding the
association between attachment anxiety and emotional disclosure
including the potential mediating role of rumination requires
two methodological perspectives. The analysis of generalized
self-reports (which partly assess expectations about disclosure)
and the analysis of reports of recent emotional events (which
assess actual behavior) may show differing associations between
attachment anxiety and emotional disclosure as well as divergent
mediating effects. Thus, whereas we posit rumination to serve as
a mediator variable in the relation between attachment anxiety
and emotional disclosure, we believe that the methodology (gener-
alized self-reports versus reports of recent events) will, in essence,
function as a moderator variable.
1.2. The current study
The purpose of the current study was to test (a) emotional
avoidance (e.g., expressive suppression) as a mediator between
attachment avoidance and emotional disclosure and (b) rumination
as a mediator between attachment anxiety and emotional
disclosure. Given the conceptual differences between emotional-
disclosure tendencies and emotional disclosure as a response to a
specic event, we collected both generalized self-reports (which
characterize how respondents appraise their typical experiences)
and daily reports over a 7-day period (which characterize what
respondents report actually doing on a daily basis). Our hypotheses
utilizing generalized self-reports were that expressive suppression
would mediate the negative relation between attachment avoid-
ance and disclosure tendencies and that rumination would mediate
the negative relation between attachment anxiety and disclosure
tendencies. Our hypotheses utilizing daily reports were that, after
accounting for the intensity of the event, emotional avoidance
would mediate the negative relation between attachment avoid-
ance and emotional disclosure, and rumination would mediate
the positive relation between attachment anxiety and emotional
disclosure. We note that these hypotheses may essentially be
construed as a test of moderated mediation, whereby we expected
the mediation results to differ between the two methodological
perspectives (i.e., methodology is the moderator variable).
2. Method
2.1. Participants
Participants were 116 undergraduate students (73 women, 43
men) from the midwestern United States. Most (78%) of the
 A.M. Garrison et al. / Personality and Individual Differences 70 (2014) 239245
participants identied themselves as European American, 10%
identied as AfricanAmerican, 3% identied as Latino/-a, 3% iden-
tied as biracial or multiracial, 2% identied as Asian or Asian
American, and 4% identied with another racial/ethnic group.
The mean age was 19.60 (SD = 3.40) years.
2.2. Measures
2.2.1. Adult attachment
The 36-item Experiences in Close Relationships Scale (ECRS;
Brennan et al., 1998) is a self-report measure of adult attachment
that ts the two-dimensional conceptualization of adult attach-
ment as comprising attachment avoidance and attachment anxiety
(Mikulincer & Shaver, 2007); thus, this instrument was appropriate
for our conceptualization of attachment. Participants read each
item and rate their level of agreement with the statement using
a scale from 1 (strongly disagree) to 7 (strongly agree). The 18-item
Attachment Anxiety subscale represents participants wish for
extreme closeness with others and worry about the potential loss
of relationship partners. The 18-item Attachment Avoidance
subscale reects the level to which participants emotionally dis-
connect from others and feel discomfort with close relationships.
Coefcients alpha for scores in the current study were .91 for
anxiety and .95 for avoidance.
2.2.2. Distress disclosure
The 12-item Distress Disclosure Index (DDI) is a self-report
measure of individuals general tendencies to disclose negative
emotions and thoughts (Kahn & Hessling, 2001). Participants rate
their agreement with each item using a scale ranging from 1
(strongly disagree) to 5 (strongly agree). Coefcient alpha in the
present study was .94.
2.2.3. Expressive suppression
The 10-item Emotion Regulation Questionnaire (ERQ; Gross &
John, 2003) is a self-report measure of the emotion-regulation
strategies of suppression and cognitive reappraisal. Cognitive
reappraisal was not part of any of our hypotheses; thus, only the
4-item Suppression subscale was analyzed. Participants rate their
agreement with each item using a scale ranging from 1 (strongly
disagree) to 7 (strongly agree). The Suppression scale measures
the extent to which individuals actively constrain their expression
of emotions. In the current study, coefcient alpha for suppression
scores was .71.
2.2.4. Rumination
The 10-item Ruminative Responses Scale (RRS) is a self-report
measure of rumination (Treynor, Gonzalez, & Nolen-Hoeksema,
2003). Participants are instructed to rate each item based on
whether they almost never, sometimes, often, or almost always think
or act in the way described by the item. We only analyzed the
5-item Brooding subscale in this study; this subscale represents a
tendency to reect on situations repeatedly without accompanying
action. Coefcient alpha for scores on this subscale in the current
study was .73.
2.2.5. Daily-event measures
Each night for 7 nights participants were asked to think about
the most signicant unpleasant event that you personally
experienced since the time you woke up, and provide a very brief
description of it. Participants then completed four brief measures
using a scale ranging from 1 (very slightly or not at all) to 5
(extremely). These four measures are described below.
First, the emotional intensity of each days event was assessed
with two items from Garrison et al.s (2012) study (e.g., How
intense was your emotional reaction to the unpleasant event right
241
when it occurred?). Responses to these items were averaged to
create a measure of event intensity that ranged from 1 to 5
(a = .81). Second, disclosure of each days event was assessed with
Garrison et al.s (2012) three-item measure (e.g., To what degree
have you shared information about this unpleasant event with
someone today, i.e., by talking with someone, texting someone,
posting an online message, etc.?). The three disclosure items were
averaged to yield a measure of event disclosure (a = .92). Third,
participants completed a six-item measure of emotional avoidance
(e.g., How hard did you try not to express your emotions related
to this unpleasant event?). Responses to these items were aver-
aged to form a measure of emotional avoidance (a = .76). Finally,
participants completed two items to measure rumination about
the event (e.g., How much did you worry about this unpleasant
event?). We averaged responses to these two items to yield an
overall measure of rumination (a = .76).
2.3. Procedure
Participants met in small groups and provided informed con-
sent. They then completed the ECRS, DDI, ERQ, and RRS. Next,
instructions were given for how to complete the daily diary portion
of the study which began that night. During the diary part of the
study, participants completed the same daily-event measures each
evening for 7 days. Reminder emails were sent to participants each
evening.
Participants responses to the online questionnaire were used as
long as the responses were entered between 5 p.m. on the day they
were reporting about and 5 a.m. the following day. Across the 116
participants, 80 (69%) completed the daily survey on time for 5 or
more days, 26 (22%) completed the daily survey on time for 3 or
4 days, and 10 completed the survey on time for 2 or fewer days.
To provide an acceptable number of daily events per participant,
we only analyzed data for the 106 participants who provided at
least three daily reports for the daily-report analyses, but we
included all 116 participants in the generalized self-report
analyses.
3. Results
3.1. Preliminary analyses
Correlations, means, and standard deviations among the vari-
ables are displayed in Table 1. Correlations among daily reports
were based on all 570 daily reports regardless of whose report it
was. Event disclosure was positively correlated with the degree
to which the participant ruminated about the event but was not
associated with the degree to which emotions concerning the
event were avoided. Correlations among generalized self-reports
indicated that attachment avoidance was associated with lower
emotional-disclosure tendencies and greater expressive suppres-
sion. Attachment anxiety was not signicantly correlated with
emotional-disclosure tendencies but was associated with greater
rumination. We also aggregated daily reports within-person and
correlated those aggregated daily reports with generalized
self-reports (see dashed box on Table 1). Aggregated daily reports
of disclosure, emotional avoidance, and rumination were positively
associated with generalized self-reports of disclosure tendencies,
expressive suppression, and rumination, respectively, thus
supporting the validity of the daily reports. Finally, although not
shown on Table 1, the number of daily reports submitted was
not signicantly correlated with either disclosure tendencies,
r = .11, p = .25, nor with aggregated daily disclosure, r = .04,
p = .65.
242 A.M. Garrison et al. / Personality and Individual Differences 70 (2014) 239245
Table 1
Correlations among daily reports and generalized self-reports.
Note: Daily reports variables are at the level of the daily event uncorrected for nesting (N = 570). Generalized self-reports are at the level of the person (N = 116). Correlations
between daily reports and generalized self-reports (indicated in the dashed box) used aggregated daily reports and are based on N = 106. p < .05, p < .01, p < .001.
3.2. Generalized self-report analyses
We estimated a single path model with attachment avoidance
and attachment anxiety as independent variables, expressive
suppression and brooding rumination as mediator variables, and
emotional disclosure tendencies as the dependent variable. Attach-
ment anxiety and avoidance were specied to be correlated with
each other, and the two mediator variables were specied to be
correlated with each other. Given the potential for gender affecting
the relations among variables, we rst estimated this model with
gender as a covariate of the two mediator variables and the
dependent variable, but the pattern of statistical signicance was
no different from when we did not include gender in the model.
Thus, for simplicity, we report a model without gender included.
This model provided a good t to the data, v2 (2, N = 116) = 4.14,
CFI = .98, RMSEA = .10 (90% C.I. = .00.23), SRMR = .05. As Fig. 1
illustrates, attachment avoidance was associated with higher
expressive suppression, and expressive suppression was, in turn,
associated with lower disclosure tendencies. Although attachment
avoidance and disclosure tendencies had a zero-order correlation
of r = .38, while controlling for expressive suppression (as well
as other variables in the model) the standardized path coefcient
for attachment avoidance was .14 and non-signicant, p = .06.
We tested the signicance of the indirect effect using 10,000
bootstrap samples to generate 95% condence intervals around
the unstandardized indirect effect of 0.11. The condence interval
(C.I.) ranged from .18 to .06; thus, the indirect effect was statis-
tically signicant, thus suggesting mediation. The correlation
between expressive suppression and rumination was r = .19,
p = .04, yet xing the correlation between the mediator variables
to 0 did not change the pattern of the mediation results.
On the attachment-anxiety side of the model, attachment anx-
iety was positively predictive of brooding rumination, yet brooding
rumination was not related to disclosure tendencies, p = .33. The
bootstrapped unstandardized indirect effect was .02 (95%
C.I. = .07, .02), indicating a non-signicant indirect effect.
3.3. Daily-event analyses
For daily reports, we had multilevel data. The independent vari-
ables (attachment avoidance and attachment anxiety) were at level
2, and the mediator variables (event avoidance and event rumina-
tion) and the dependent variable (event disclosure) were at level 1.
We therefore estimated a multilevel mediation model using Mplus
6.11 (Muthn & Muthn, 1998-2010). We used Preacher, Zyphur,
and Zhangs (2010) multilevel structural equation modeling
(MSEM) approach which partitions paths into between-person
and within-person components. Because attachment was mea-
sured at level 2, the test of mediation can only occur at the
between-person level. Thus, the mediation analyses only incorpo-
rated the between-person effects. Gender was not correlated with
any of the daily reports, so it was not included in the analyses.
In this model (see Fig. 2) the path coefcients between the
mediators and the dependent variable were specied as random,
meaning that these paths were allowed to vary across participants.
We also accounted for the intensity of the days emotional event in
the prediction of disclosure of these specic emotional events.
Because the predictive effects of emotional avoidance and rumina-
tion on emotional disclosure might vary as a function of the
emotional intensity of the event, we examined whether event
intensity interacted with either emotional avoidance or rumina-
tion when predicting disclosure. Neither the interaction with event
avoidance, p = .89, nor the interaction with event rumination,
p = .45, was signicant. Thus, we specied intensity as a covariate
but not as a moderator variable.
Attachment avoidance predicted higher emotional avoidance of
the event, meaning that individuals with relatively higher attach-
ment avoidance tended to avoid their emotions in response to
unpleasant events over the course of the week. At the within-
person level, higher avoidance of the event predicted lower
disclosure of the event, meaning that when an individual avoided
emotions concerning an event he or she was less likely to talk
about that event. In Preacher et al.s (2010) MSEM approach, there
is also a between-person component to the relation between the
mediator (emotional avoidance, in this case) and dependent
variable (see Table 2); this coefcient was also signicant but
positive, suggesting that participants who tend to avoid emotions
surrounding events also tend to talk about these events. The
indirect effect between attachment avoidance and event disclosure
at the between-person level, however, was just below the thresh-
old of statistical signicance, p = .06. Thus, despite a between-
person association between attachment anxiety and the emotional
 A.M. Garrison et al. / Personality and Individual Differences 70 (2014) 239245
avoidance of daily events, and within-person and between-person
associations between emotional avoidance and disclosure of daily
events, the full mediation hypothesis for attachment avoidance
and emotional disclosure was not supported.
The other half of the model addressed whether rumination
mediates the relation between attachment anxiety and event
disclosure. Attachment anxiety was positively predictive of
rumination, and rumination was associated with greater event
disclosure; this was true at both the between-person and
within-person level (see Table 2). As was the case for attachment
avoidance, the indirect effect between attachment anxiety and
event disclosure was not signicant at the .05 level, p = .06. Thus,
despite bivariate relations being in the predicted directions, rumi-
nation did not mediate the relation between attachment anxiety
and the daily disclosure of emotional events.
4. Discussion
The purpose of this study was to explore the relation between
attachment and emotional disclosure by investigating emotional
avoidance and rumination as potential mediating variables. In
order to obtain a methodologically comprehensive test of these
relationships, we measured mediator variables and emotional dis-
closure both as generalized self-reports and as reports of daily
events. This appears to be the rst time these two approaches have
been used together to understand potential mediating variables in
the relation between attachment orientations and emotional
disclosure.
Based on generalized self-reports, individuals with higher levels
of attachment avoidance exhibited less of a tendency to share their
distress with others than individuals lower in attachment avoid-
ance. This nding is not novel (e.g., Garrison et al., 2012; Wei
et al., 2005), but we extended this literature by nding that expres-
sive suppression mediated the relation between attachment avoid-
ance and emotional disclosure. This is consistent with attachment
theory, as avoidantly attached individuals seek ways to diminish
their experiences of distress (Mikulincer & Shaver, 2007). Thus, if
avoidantly attached individuals are suppressing their emotions,
they are unlikely to share their distress with others.
Analyses with daily reports supported some aspects of
attachment and emotion-regulation theories. We found a positive
between-person relation between attachment avoidance and emo-
tional avoidance of unpleasant events, and we found the expected
negative within-person relation between emotional avoidance and
disclosure. These bivariate relations replicate the generalized
self-report analyses as well as previous research linking attach-
ment avoidance with higher levels of emotional suppression (e.g.,
Fig. 1. Generalized self-report model depicting hypothesized indirect relations
between attachment and disclosure tendencies as mediated by expressive sup-
pression and brooding rumination. p < .05; p < .01; p < .001.
243
Caldwell & Shaver, 2012) and emotional suppression with lower
levels of disclosure (e.g., Kahn et al., 2012). Of interest was a signif-
icant positive between-person effect between emotional avoidance
and event disclosure. Thus, people who tended to avoid the emo-
tions surrounding the weeks events also tended to disclose their
emotions regarding the weeks events. This highlights the idea that
even though individuals may rely on particular coping strategies
more often than others, over the course of a week they might make
use of multiple strategies in an effort to effectively manage their
distress. This pattern of results also underscores the importance
of using both between- and within-person research approaches.
Despite the signicant bivariate relations, the analysis of daily
reports failed to support the mediation hypothesis for attachment
avoidance. This was not the case in the analysis of generalized
self-reports, suggesting that peoples generalized appraisals of
their emotional-disclosure tendencies (yet not their daily disclo-
sure behaviors) might partly be a function of attachment
avoidance. Specically, in addition to measuring a history of
disclosure behaviors, generalized self-reports of emotional-
disclosure tendencies may tap into avoidant individuals desire
for interpersonal distance.
The other half of our analyses focused on understanding the
relation between attachment anxiety and emotional disclosure.
Although the hyperactivation of emotions inherent in attachment
anxiety (Mikulincer & Shaver, 2007) should lead anxiously
attached individuals to disclose distress, our analyses did not
support this hypothesis. Instead, we found a negative (but non-
signicant) zero-order correlation between attachment anxiety
and disclosure tendencies, a nding that has now appeared multi-
ple times in the literature (see Kahn et al., 2012). The internal
working models of others held by anxiously attached individuals
are that other people will not be reliable attachment gures; thus,
it does not pay to disclose distress. Thus, generalized self-reports of
emotional-disclosure tendencies partly reect these relational
schemas (Pierce, Baldwin, & Lydon, 1997). It may also be that
anxiously attached individuals fear that others will perceive them
as being unable to cope with lifes demands; others might choose
to leave the relationship if that were the case. If so, they may
inhibit the disclosure of distress as a means to appear emotionally
stable. This is speculative, however; integrating variables such as
need for approval and social desirability into this model might be
valuable in future work.
Fig. 2. Daily-reports model depicting unstandardized paths between attachment
and emotional disclosure as mediated by emotional avoidance and rumination. The
circles on the mediator-to-dependent-variable paths indicate that these paths were
specied as random effects; values of the coefcients in Fig. 2 reect the
unstandardized within-person effect for these paths. p < .05; p < .01; p < .001.
244 A.M. Garrison et al. / Personality and Individual Differences 70 (2014) 239245
Table 2
Unstandardized between- and within-person paths between mediator and event disclosure and indirect effects from the daily reports analyses.
Mediator variable Level of effect
Emotional avoidancea Between
Within
b
Rumination Between
Within
Note: Dependent variable was event disclosure. Within estimates are xed effects (i.e., averaged across all participants). Dashes indicate the effect was not estimated.
*
p < .05.
**
p < .01.
***
p < .001.
a
Mediator of relation between attachment avoidance and event disclosure.
b
Mediator of relation between attachment anxiety and event disclosure.
At the level of daily events, a somewhat different pattern
emerged. Attachment anxiety was positively related to rumination,
as theory would suggest (e.g., Mikulincer & Shaver, 2007).
Rumination about daily emotional events was also positively
predictive of event disclosure, such that thinking about events over
and over was associated with talking about those events. Never-
theless, mediation was not strong enough to be statistically
signicant at the .05 level. Thus, it is best to be conservative and
consider these relations among attachment anxiety, rumination,
and event disclosure only at the bivariate (versus multivariate)
level.
4.1. Limitations and ideas for future research
Our sample primarily consisted of European American individ-
uals; therefore, readers should not generalize our ndings to other
groups of individuals. Future research could focus on understand-
ing more about how, when, and why individuals from other
cultural groups use emotional disclosure as a potential means for
managing their distress. Similarly, it is important to note that our
participants were all college students. In future studies, research-
ers may wish to explore whether the supported relations between
attachment and emotional disclosure are found when non-student
samples are used.
Additionally, all measures in this study were based on partici-
pant self-reports which may have been a source of bias. Whereas
the daily measures in our study may have decreased retrospective
bias, participants were still providing reports at the end of each
day. Future researchers may wish to gather responses from
participants more than once per day to capture events as they
happen. Furthermore, although we based portions of our
daily-event measures on Garrison et al.s (2012) measures, we also
created several new items specically for this study. Thus, the
validity of our daily questionnaire could not be assessed fully.
Finally, other variables are worthy of including into models of
emotional disclosure. For example, we only focused on two poten-
tial mediators. Although the chosen mediators were based on
theory, additional mediators may exist. For instance, self-efcacy
for coping, perceived level of social support, and outcome expecta-
tions regarding disclosure may impact disclosure behavior.
Researchers may also wish to deepen our understanding of the dis-
closure behavior of insecurely attached individuals by asking about
the reasoning behind their choices to disclose or not. Such informa-
tion could be obtained in a mixed-method or qualitative study. In
addition, other personality-related variables could be examined as
potential predictors of emotional disclosure; one candidate would
be emotional intelligence. On the outcome side, relatively little
work has been done on the disclosure of positive emotions. Exper-
imental research has shown that suppressing positive emotions
leads to a decreased subjective experience of those emotions
Mediator to event-disclosure path Indirect effect
Estimate SE Estimate SE
1.94*** 0.53
0.38*** 0.09 0.24 0.13
0.58** 0.20
0.20*** 0.05 0.10 0.06
(Gross & Levenson, 1997), so expanding the domain of emotions
in disclosure research would be of interest.
4.2. Implications for research and theory
Our ndings suggest the importance of incorporating attach-
ment into our understanding of disclosure behaviors. Specically,
attachment avoidance clearly seems to be related to lower levels
of emotional-disclosure tendencies because these individuals tend
to suppress their emotions, yet more theoretical and empirical
work is needed on understanding when and why attachment anx-
iety is related to disclosure. For example, how does disclosure to
different relationship targets vary as a function of attachment anx-
iety? How does the content of disclosure affect how anxiously
attached individuals engage in disclosure? To what degree do
methodological issues affect conclusions about attachment anxiety
and disclosure? Addressing issues such as these would represent
important points of departure from this research.
Furthermore, these ndings emphasize the importance of build-
ing on theory related to individual differences in the normative use
of daily and generalized disclosure. Our results indicate there are
meaningful individual differences that exist with regards to disclo-
sure behavior. Moreover, these disclosure patterns differ based on
interpersonal and affective variables that are of interest to person-
ality researchers, such as individuals attachment orientations and
their emotion-regulation patterns. Thus, this research may contrib-
ute to a more complete understanding of the complex relationships
among interpersonal factors, emotion regulation, and disclosure.
References
Averill, J. R. (1997). The emotions: An integrative approach. In R. Hogan, J. Johnson,
& S. Briggs (Eds.), Handbook of personality psychology (pp. 513541). San Diego,
CA: Academic Press. http://dx.doi.org/10.1016/B978-012134645-4/50022-6.
Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult
romantic attachment: An integrative overview. In J. A. Simpson & W. S. Rholes
(Eds.), Attachment theory and close relationships (pp. 4676). New York: Guilford.
Caldwell, J. G., & Shaver, P. R. (2012). Exploring the cognitive-emotional pathways
between adult attachment and ego-resiliency. Individual Differences Research,
10, 141152.
Cano, A., Leong, L. E. M., Williams, A. M., May, D. K. K., & Lutz, J. R. (2012). Correlates
and consequences of the disclosure of pain-related distress to ones spouse.
Pain, 153, 24412447. http://dx.doi.org/10.1016/j.pain.2012.08.015.
Fraley, R. C., & Shaver, P. R. (1997). Adult attachment and the suppression of
unwanted thoughts. Journal of Personality and Social Psychology, 73, 10801091.
http://dx.doi.org/10.1037/0022-3514.73.5.1080.
Frattaroli, J. (2006). Experimental disclosure and its moderators: A meta-analysis.
Psychological Bulletin, 132, 823865. http://dx.doi.org/10.1037/0033-
2909.132.6.823.
Garrison, A. M., Kahn, J. H., Sauer, E. M., & Florczak, M. A. (2012). Disentangling the
effects of depression symptoms and adult attachment on emotional disclosure.
Journal of Counseling Psychology, 59, 230239. http://dx.doi.org/10.1037/
a0026132.
Gross, J. J., & John, O. P. (2003). Individual differences in two emotion regulation
processes: Implications for affect, relationships, and well-being. Journal of
Personality and Social Psychology, 85, 348362. http://dx.doi.org/10.1037/0022-
3514.85.2.348.
 A.M. Garrison et al. / Personality and Individual Differences 70 (2014) 239245
Gross, J. J., & Levenson, R. W. (1997). Hiding feelings: The acute effects of inhibiting
negative and positive emotion. Journal of Abnormal Psychology, 106, 95103.
http://dx.doi.org/10.1037/0021-843X.106.1.95.
Kahn, J. H., & Hessling, R. M. (2001). Measuring the tendency to conceal versus
disclose psychological distress. Journal of Social and Clinical Psychology, 20,
4165. http://dx.doi.org/10.1521/jscp.20.1.41.22254.
Kahn, J. H., Hucke, B. E., Bradley, A. M., Glinski, A. J., & Malak, B. L. (2012). The
Distress Disclosure Index: A research review and multitrait-multimethod
examination. Journal of Counseling Psychology, 59, 139149. http://dx.doi.org/
10.1037/a0025716.
Kennedy-Moore, E., & Watson, J. C. (2001). How and when does emotional
expression help? Review of General Psychology, 5, 187212. http://dx.doi.org/
10.1037/1089-2680.5.3.187.
Lepore, S. J., Ragan, J. D., & Jones, S. (2000). Talking facilitates cognitive-emotional
processes of adaptation to an acute stressor. Journal of Personality and Social
Psychology, 78, 499508. http://dx.doi.org/10.1037/0022-3514.78.3.499.
Liverant, G. I., Kamholz, B. W., Sloan, D. M., & Brown, T. A. (2011). Rumination in
clinical depression: A type of emotional suppression? Cognitive Therapy and
Research, 35, 253265. http://dx.doi.org/10.1007/s10608-010-9304-4.
Mikulincer, M., & Shaver, P. R. (2007). Attachment in adulthood: Structure, dynamics,
and change. New York: Guilford.
Muthn, L. K., & Muthn, B. O. (19982010). Mplus users guide (6th ed.). Los Angeles,
CA: Muthn & Muthn.
Nolen-Hoeksema, S. (2000). The role of rumination in depressive disorders and
mixed anxiety/depressive symptoms. Journal of Abnormal Psychology, 109,
504511. http://dx.doi.org/10.1037/10021-843X.109.3.504.
Pierce, T., Baldwin, M. W., & Lydon, J. E. (1997). A relational schema approach to
social support. In G. R. Pierce, B. Lakey, I. G. Sarason, & B. R. Sarason (Eds.),
Sourcebook of social support and personality (pp. 1947). New York: Plenum
Press.
245
Pistole, M. C. (1993). Attachment relationships: Self-disclosure and trust. Journal of
Mental Health Counseling, 15, 94106.
Preacher, K. J., Zyphur, M. J., & Zhang, Z. (2010). A general multilevel SEM framework
for assessing multilevel mediation. Psychological Methods, 15, 209233. http://
dx.doi.org/10.1037/a0020141.
Rim, B. (2009). Emotion elicits the social sharing of emotion: Theory and
empirical review. Emotion Review, 1, 6085. http://dx.doi.org/10.1177/
1754073908097189.
Roisman, G. I., Tsai, J. L., & Chiang, K. S. (2004). The emotional integration of
childhood experience: Physiological, facial expressive, and self-reported
emotional response during the Adult Attachment Interview. Developmental
Psychology, 40, 776789. http://dx.doi.org/10.1037/0012-1649.40.5.776.
Saffrey, C., & Ehrenberg, M. (2007). When thinking hurts: Attachment, rumination,
and post-relationship adjustment. Personal Relationships, 14, 351368. http://
dx.doi.org/10.1111/j.1475-6811.2007.00160.x.
Tan, R., Overall, N. C., & Taylor, J. K. (2012). Lets talk about us: Attachment,
relationship-focused disclosure, and relationship quality. Personal Relationships,
19, 521534. http://dx.doi.org/10.1111/j.1475-6811.2011.01383.x.
Treynor, W., Gonzalez, R., & Nolen-Hoeksema, S. (2003). Rumination reconsidered:
A psychometric analysis. Cognitive Therapy and Research, 27, 247259. http://
dx.doi.org/10.1023/A:1023910315561.
Wei, M., Russell, D. W., & Zakalik, R. A. (2005). Adult attachment, social self-efcacy,
self-disclosure, loneliness, and subsequent depression for freshman college
students: A longitudinal study. Journal of Counseling Psychology, 52, 602614.
http://dx.doi.org/10.1037/0022-0167.52.4.602.
Zech, E., & Rim, B. (2005). Is talking about an emotional experience helpful? Effects
on emotional recovery and perceived benets. Clinical Psychology and
Psychotherapy, 12, 270287. http://dx.doi.org/10.1002/cpp.460.