Personal Relationships, 12 (2005), 145–163. Printed in the United States of America.
Copyright Ó 2005 IARR. 1350-4126=02
On the interpersonal regulation of emotions:
Emotional reliance across gender,
relationships, and cultures
RICHARD M. RYAN,a JENNIFER G. LA GUARDIA,b JESSICA SOLKY-BUTZEL,c
VALERY CHIRKOV,d AND YOUNGMEE KIMe
a
University of Rochester; bUniversity of Waterloo, Canada; cRoosevelt Hospital;
d
University of Saskatchewan, Canada; eAmerican Cancer Society
Abstract
Three studies examine people’s willingness to rely on others for emotional support. We propose that emotional
reliance (ER) is typically beneficial to well-being. However, due to differing socialization and norms, ER is also
expected to differ across gender and cultures. Further, following a self-determination theory perspective, we hypothesize that ER is facilitated by social partners who support one’s psychological needs for autonomy, competence, and
relatedness. Results from the studies supported the view that ER is generally associated with greater well-being and
that it varies significantly across different relationships, cultural groups, and gender. Within-person variations in
ER were systematically related to levels of need satisfaction within specific relationships, over and above betweenperson differences. The discussion focuses on the adaptive value and dynamics of ER.
When emotionally significant events occur,
some individuals readily turn to others for
support. Yet for others the act of sharing emotional concerns can be threatening or even
viewed as a sign of weakness or inadequacy.
Further, even for those who are willing to
rely on others, there are some social partners
with whom they would be reluctant to share
feelings and others whom they might readily
seek out for emotional support.
In this research, we investigate people’s willingness to rely on others for emotional support,
both as an individual difference and in terms of
the factors leading to selective reliance on particular partners. We term this willingness to
turn to others in emotionally salient situations
emotional reliance (ER). Despite some charac-
Correspondence should be addressed to Richard M.
Ryan, University of Rochester, Department of Clinical
and Social Sciences in Psychology, Rochester, NY
14627, e-mail: ryan@psych.rochester.edu.
terizations of ER as a sign of negative dependency (e.g., Hirschfeld et al., 1977), we predict
that ER will on average be positively associated with well-being. We also expect that tendencies toward ER are influenced by both
gender and culture, resulting in mean differences. Finally, based on self-determination theory (SDT) (Deci & Ryan, 2000; Ryan & Deci,
2000), we predict that ER is selective and that
people are more willing to emotionally rely on
those who they experience as supporting their
basic psychological needs for autonomy, competence, and relatedness.
ER and Social Support
Social support is a broad concept that refers
to the help and care that others can provide
and the effects of that care on coping, health,
and psychological well-being. Social support
can be manifest in many forms, such as
providing information, instrumental help, or
145
146
companionship, among other ways of caring
(Cohen & Wills, 1985; Uchino, Cacioppo, &
Kiekolt-Glaser, 1996). Thus, many distinct
interpersonal processes fall under the rubric
of social support, each of which may have its
own unique dynamics and associations with
outcomes (Goldsmith & Fitch, 1997).
Emotional support is one specific type of
social support, representing the expression
of concern, compassion, and comfort for an
individual during emotional experiences
(Burleson & Kunkel, 1996; Reis & Collins,
2000; Wills & Shiner, 2000). Emotional support, in fact, appears to be one of the most
critical types of social support, facilitating
both coping with specific stressors and
contributing to sustained well-being throughout the life span (Carstensen, Isaacowitz, &
Charles, 1999; Rook, 1987; Ryan & La
Guardia, 2000).
Various literatures suggest that merely having others available who can provide emotional
support is positive for well-being (Cohen,
Sherrod, & Clark, 1986; Pierce, Sarason, &
Sarason, 1991; Ryan, Stiller, & Lynch, 1994;
Windle, 1992) and health (e.g., Emmons &
Colby, 1995). Yet, as Goldsmith and Fitch
(1997) highlight, simply knowing the availability of potential emotional supports overlooks the reactions that recipients have to
these available supports and what features
of interpersonal relationships conduce to the
supports are being utilized.
ER is not a measure of emotional support
per se but rather of a person’s readiness to
enter into interactions where emotional supports may be available. We suggest that ER
represents an individual difference variable
in the sense that people differ in their overall
willingness to turn to others for support. Yet,
ER also varies within individuals from relationship to relationship as a function of the
qualities of these specific relationships.
ER, Psychological Needs, and Well-being
ER is a complex issue that is tied to the
dynamics of psychological needs within specific relationships. In particular, we suggest
that people’s choices about whether to seek
emotional support from specific others
R. M. Ryan et al.
depend on the functional significance of such
support with respect to the recipient’s basic
psychological needs (Butzel & Ryan, 1997;
Ryan & Solky, 1996). The concept of functional significance is derived from SDT and
refers to the meaning of an event with respect
to whether it facilitates or threatens the satisfaction of psychological needs, specifically,
autonomy, competence, and relatedness (Deci
& Ryan, 1985, 2000). According to SDT,
emotionally sensitive partners are those who
provide timely and appropriate responses to
each of these three needs (La Guardia, Ryan,
Couchman, & Deci, 2000), and we predict
that need-supportive partners are those who
most facilitate ER.
Particularly crucial in facilitating ER is the
degree to which a social partner is perceived
as supportive of autonomy. Within SDT,
autonomy concerns volition, the experience
of one’s actions as self-endorsed (Ryan &
Deci, 2000). The opposite of autonomy is heteronomy, which concerns feeling coerced,
compelled, or controlled by forces alien to the
self. A person who is supportive of autonomy
does not attempt to control the partner’s behavior, reactions, or feelings but rather is
attentive to and interested in the partner’s perspective and frame of reference (Ryan, 1993).
Although in some literatures autonomy is
equated with independence, SDT differentiates these constructs by defining dependence
strictly in terms of reliance: One is dependent
when relying on another for resources or supports (Ryan & Lynch, 1989). The opposite of
dependence is thus not autonomy but nonreliance, or independence. Therefore, according
to SDT, one can be autonomously dependent or
reliant, as when one willingly turns to others
for support, or autonomously independent, as
when one reflectively decides not to rely on
others (V. Chirkov, Ryan, Kim, & Kaplan,
2003; Ryan, 1993).
Because the issues of autonomy and independence are often not differentiated, one can
find quite disparate theoretical views on the
meaning and likely effects of ER. Some authors suggest that ER or dependency is problematic, and others laud it as natural and
beneficial. In our view, studies that have shown
negative effects of emotional dependence are
Emotional reliance
often those that conflate such dependence with
heteronomy and/or fears of separateness. For
example, Hirschfeld et al. (1977) introduced
a widely used measure of interpersonal dependency that includes subscales concerning ‘‘emotional reliance’’ and ‘‘assertion of autonomy.’’
However, from an SDT view, Hirschfeld
et al.Õs ER construct concerns more than
simply relying on others for emotional support.
It taps strong fears of aloneness (‘‘I tend to
imagine the worst if a loved one doesn’t arrive
when expected’’), needs for approval (‘‘Disapproval by someone I care about is very painful
to me’’), and vulnerability (‘‘I think that most
people don’t realize how easily they can hurt
me’’). Similarly, the subscale ‘‘assertion of autonomy’’ focuses on interpersonal detachment
(‘‘I don’t need anyone’’), separateness (‘‘I prefer to be by myself’’), and extreme self-focus
(‘‘I am the only person I want to please’’),
rather than volition. Notably, Hirschfeld et al.
have shown that dependence as they assess it
relates to psychopathology, an outcome one
would not expect from either ER or autonomy,
at least as SDT defines them.
Similarly, some of the negative results associated with emotional autonomy (EA) are due
to that construct being construed as a desire for
separateness, detachment, and/or fears about
relying on others, rather than being focused on
volition. Steinberg and Silverberg (1986) introduced a construct called emotional autonomy
that they originally conceptualized as an aspect
of healthy individuation from parents. However, Ryan and Lynch (1989) suggested that
EA as measured by Steinberg and Silverberg
reflects emotional detachment from parents
rather than autonomy. Many EA items describe
an unwillingness to seek, rely on, or trust parental guidance. Ryan and Lynch accordingly
demonstrated that greater EA was associated
with adolescents perceiving parents as less loving and more rejecting and that those high in
EA were less willing to utilize parents for support. In contrast, lower EA was related to lower
perceived parental control and rejection and, on
average, with better overall adjustment. Subsequent studies have shown that EA is higher
when parents are nonnurturing and that teens
high in EA are more likely to conform with
peers, engage in risky behaviors, and achieve
147
less in school (e.g., McBride-Chang & Chang,
1998; McQueen, Getz, & Bray, 2003; Turner,
Irwin, & Millstein, 1991).
In addition to the need to feel autonomous,
SDT also suggests that ER is connected with
the basic psychological needs for competence
and relatedness. Competence concerns feeling effective within one’s environment (Deci,
1975; White, 1963). In some contexts, emotionally relying on others may be treated as
incompetence, whereas in others it may be
treated as mature and appropriate, and actually could facilitate perceived competence
(Goldsmith & Fitch, 1997; Ryan & Solky,
1996). Relatedness concerns the feeling of
belonging and connectedness (Baumeister &
Leary, 1995; Deci & Ryan, 1985). We assume that ER can enhance feelings of relatedness and thus well-being (La Guardia et al.,
2000; Reis & Collins, 2000; Reis & Franks,
1994; Russell, Peplau, & Cutrona, 1980).
We expect three findings. First, people will
indicate more willingness to rely on those
they perceive to support their autonomy. People will prefer to turn to others who are able
to listen to and understand their internal frame
of reference without needing to control them.
Second, because we view ER as a resource in
garnering emotional support rather than as
opposed to autonomy or as a problematic
form of dependency, higher levels of ER will
be associated with greater well-being. Finally,
the association between ER and well-being
will be mediated by the fulfillment of psychological needs. That is, one benefits from emotionally relying on another to the extent that
feelings of relatedness, competence, and autonomy are enhanced.
ER and Gender
According to many theorists, boys are socialized to be agentic and independent, whereas
girls are socialized to be communal and interdependent (Helgeson, 1994). As a result of
these different emphases in socialization, the
interpersonal expression of emotions is purported to be more salient (Shields, 1995) and
frequent (Wills, 1998) for women than for
men. Across cultures, women perceive comforting skills, intimate communication, and
148
interdependence to be more important than do
men (Kashima, Yamaguchi, Choi, Gelfand, &
Yuki, 1995; Samter, Whaley, Mortenson, &
Burleson, 1997; Shields, 1995; Ting-Toomey,
1991). Emotional sharing is also considered
to be a more common strategy for women
than men to facilitate intimacy (Caldwell &
Peplau, 1982) and well-being (Jordan, Kaplan,
Miller, Stiver, & Surrey, 1991).
Consistent with these formulations, Ryan
et al. (1994) found that by early adolescence,
girls were less likely than boys to report turning to ‘‘no one’’ when coping with emotional
events. Despite this mean difference, the
effects on well-being were consistent across
gender: ER on ‘‘no one’’ predicted poorer
mental health equally for boys and girls. In
light of such findings, we expect women to
evidence greater overall ER. Yet, we also
expect ER to afford similar benefits to males
and females in most circumstances.
The Present Studies
In Study 1, we assess college studentsÕ ER, or
willingness to turn to a variety of target figures during emotionally salient events, and
relate this assessment of ER to measures of
attachment and unhealthy interpersonal dependence. Additionally, we test for gender effects, predicting women will report higher
levels of ER. Despite mean differences, we
expect ER to be positively associated with
well-being for both males and females. In
line with SDT, we also predict that students
will be more willing to emotionally rely on
the parent who most supports their autonomy
and less on the parent whom they regard as
relatively more controlling. Finally, we predict that the more students report relying on
‘‘no one’’ when emotional events occur, the
poorer their adjustment. In Study 2, we
extend our examination of variability in ER,
hypothesizing that within-person variations in
ER on particular others are related to relationship-specific supports for autonomy,
competence, and relatedness. Study 3 tests
the generalizability of the presumed positive
associations between ER and well-being in
samples from Korea, Russia, Turkey, and the
United States.
R. M. Ryan et al.
Study 1
Method
Participants and procedure
One hundred ninety-five undergraduates (119
female, 76 male) completed assessment packets in small groups (n , 15) and received
extra course credit for their participation.
These participants ranged in age from 17 to
28 years, with a mean age of 20.03 years.
Measures
Emotional reliance.
We assessed ER using
10 items, each rated on a 5-point Likerttype scale for mother, father, best friend,
roommate, and ‘‘no one,’’ resulting in a total
of 50 items. Five of the item stems were
adapted from the utilization subscale of the
Inventory of Adolescent Attachments (Greenberg, Siegel, & Leitch, 1983) by Ryan et al.
(1994). Five new item stems were added to
enhance the range of situations. Table 1 presents all 10 items. Participants were instructed
that if they did not currently have a particular
relationship (e.g., a roommate), they should
leave those items blank. If they had a nontraditional mother or father figure (e.g., a stepfather), they were instructed to refer to that
figure for items pertaining to mother or
father.
A principal components factor analysis of
the 10 ER items was performed for each target figure (mother, father, best friend, roommate, and ‘‘no one’’). Data from each target
supported a one-factor model, accounting for
at least 58% of the variance, with each item
loading above .60. To explore whether positively and negatively valenced items could
form separate factors, a ‘‘forced’’ two-factor
solution was attempted for each target. In no
case did eigenvalues for a second factor
exceed 1.0. As a result, ER scores for each
target were formed by calculating the mean
of all 10 items, and an overall ER score was
calculated by taking the mean of the scores
across targets, excluding ‘‘no one.’’ Alphas
were .93, .95, .91, and .97 for mother, father,
best friend, and roommate, respectively.
Emotional reliance
149
Table 1. Emotional reliance questionnaire
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
When I am alone or depressed, I would turn to (my) ________.
When I am anxious or scared about something, I would turn to (my) ________.
When I am feeling very bad about myself and need a boost, I would turn to
(my) ________.
When I am feeling happy or have good news, I would turn to (my) ____________.
When I have just experienced a tragedy (e.g., the death of a family member or friend),
I would turn to (my) ________.
When I am feeling overwhelmed by responsibilities and commitments I would turn to
(my) ________.
When I am frustrated or angry I would turn to (my) ________.
When I am feeling disappointed I would turn to (my) ________.
When I am proud of my accomplishments I would turn to (my) ________.
When I am confused or indecisive I would turn to (my) ________.
Note. Items 1–5 adapted from Greenberg et al.Õs (1983) Inventory of Adolescent Attachment.
Interpersonal dependency inventory.
The
48-item Interpersonal Dependency Inventory
(IDI; Hirschfeld, et al., 1977) consists of three
subscales labeled (a) emotional reliance on
others, (b) lack of social self-confidence, and
(c) assertion of autonomy. Their ER subscale
represents wishes for contact, approval and
support from others, and dread at loss of the
other. The lack of social self-confidence subscale reflects wishes for help in social decision
making, while assertion of autonomy reflects
preferences for being alone and independent
and not having one’s self-esteem be contingent
on othersÕ approval. Sample items from each
of the subscales are (a) ‘‘Disapproval by someone I care about is very painful to me,’’ (b)
‘‘I am quick to agree with the opinions of
others,’’ and (c) ‘‘I prefer to be by myself.’’
Each subscale score represents the mean of the
items, and an overall IDI score is calculated
by taking the mean of all 48 items. Alphas
were .87, .86, and .83 for each of the subscales, respectively.
Attachment.
We created a 15-item measure consisting of 13 items from the Adult
Attachment Styles inventory (Simpson,
1990), 1 item from Hazan and Shaver’s
(1987) avoidant prototype (‘‘I find it difficult
to allow myself to depend on others’’) and 1
item adapted from their ambivalent prototype
(‘‘I often want to merge completely with
another person’’).1 Scores were calculated by
taking the average of items for each style,
yielding indexes of attachment security,
avoidance, and enmeshment (a ¼ .67, .82,
and .81, respectively).
Perception of parents scale.
The Perception of Parents Scale (Robbins, 1994) assesses participantsÕ perceptions of parents on
dimensions of autonomy support versus control. Participants rate 14 items each for
mother and father on 5-point Likert scales,
yielding a total of 28 items. Sample items
include ‘‘My mother listens to my opinion or
perspective when I’ve got a problem’’ (autonomy support) and ‘‘My mother tries to tell me
how to run my life’’ (control). The mean of
the 14 items for each target constitutes the
1. Simpson’s (1990) measure was originally constructed
by having participants rate 13 of 15 sentences contained in Hazan and Shaver’s (1987) paragraph-form
prototype measure of secure, avoidant, and anxious
prototypes on Likert-type scales as described in Study
1, Methods. Simpson’s technique allows for greater
individual difference variability within each category.
However, whereas Simpson’s adaptation included 13
distinct items, at the suggestion of a prominent adult
attachment researcher, we included 15 because (a)
Simpson did not include one of the avoidant phrases
and (b) one of his anxious items combined two distinct phrases from the anxious prototype that we separated into two independent items. This resulted in our
15 total items. Accordingly, our means cannot be
directly compared with Simpson’s.
—
—
—
.58**, t(193) ¼ 7.83***
.29**, t(194) ¼ 14.37***
Note. For paired t tests, values are Bonferroni corrected at p , .005.
*p , .05. **p , .01. ***p , .001.
—
—
.14*, t(192) ¼ 3.36**
.13, t(191) ¼ 1.91
.20**, t(192) ¼ 10.32***
—
.22**, t(192) ¼ 9.88***
.21**, t(194) ¼ 6.93***
.20**, t(193) ¼ 12.33***
.35**, t(194) ¼ 21.35***
.53**
.61**
.73**
.75**
.40**
ER mother
ER roommate
ER best friend
Paired sample t tests with Bonferroni corrections for family-wise error (p , .005) were
conducted to compare ER on different relational partners. Sample means revealed that
ER was significantly highest for best friend
(M ¼ 4.36, SD ¼ .74), followed by mother
(M ¼ 3.81, SD ¼ 1.0), roommate (M ¼ 3.45,
SD ¼ 1.23), and father (M ¼ 3.24, SD ¼
1.19), with roommate and father targets not
significantly different from each other. Significantly lower than all other targets was the
endorsement of ‘‘no one’’ (M ¼ 2.15, SD ¼
1.02). Table 2 presents the relations among
the total, target-specific, and ‘‘no one’’ ER
ER total
Results
Table 2. Intercorrelations and paired t tests among emotional reliance (ER) total, target, and ‘‘no one’’ scores (Study 1)
Well-being.
Symptoms of depression were
assessed using the Center for Epidemiological Studies Depression scale (CES-D;
Radloff, 1977), a 20-item, well-validated measure. Anxiety was measured by the widely
used Taylor Manifest Anxiety Scale (TMAS;
Taylor, 1953), which consists of 28 items
rated on 7-point Likert scales. Finally, vitality was measured using the Subjective Vitality Scale (SVS; Ryan & Frederick, 1997), a
7-item measure of personal energy, characterized by feelings of vigor and aliveness.
Alphas for the CES-D, TMAS, and SVS were
.92, .90, and .86, respectively, for this sample. Vitality has particular interest within
SDT as a marker of organismic integrity
(Nix, Ryan, Manly, & Deci, 1999), and our
interest is ultimately studying how congruent
affective expression maintains and enhances
vitality. Thus, we included the SVS among
our central indicators of well-being.
A principal components factor analysis of
the well-being constructs yielded a single factor accounting for 70% of the variance. A
unit-weighted composite score was created
using the mean of the standardized scores.
Both overall and individual scores are used in
analyses.
ER father
scale score. Alphas were .92, .84, .92, and .83
for the separate factors of perceived maternal
autonomy support, maternal control, paternal
autonomy support, and paternal control,
respectively.
—
—
—
—
.26**, t(193) ¼ 8.66***
R. M. Ryan et al.
ER best friend
ER roommate
ER mother
ER father
ER no one
150
Note. Given the number of correlations conducted, more conservative test of significance (p , .01) was selected. ER ¼ emotional reliance; IDI ¼ interpersonal dependency inventory;
AA ¼ assertion of autonomy; SS ¼ social self confidence; CES-D ¼ Center for Epidemiological Studies Depression scale; VIT ¼ subaortic vitality scale; ANX ¼ Taylor Manifest Anxiety Scale; WB ¼ Well-being composite.
**p , .01. ***p , .001.
.19**
.11
.15
.13
.10
.11
.29***
.12
.24**
.26***
.15*
.25**
.25***
.13
.18
.25**
.13
.19**
.29***
.14
.23**
.26***
.15
.22**
.01
.02
.01
.03
.03
.05
.31***
.14
.20**
.31***
.21**
.33***
.31***
.14
.29***
.31***
.12
.18
.18
.12
.13
.22**
.05
.20**
ER total
ER mother
ER father
ER best friend
ER roommate
ER no one
.23**
.16*
.09
.32***
.12
.43***
.03
.07
.04
.01
.03
.14
VIT
CES-D
WB
Enmeshment
Avoidance
Security
IDI ER
IDI SS
IDI AA
Table 3. Intercorrelations of ER, total, and target scores with IDI, attachment, and well-being scales (Study 1)
scores, showing that while total ER relates to
each target, the targets are modestly associated with one another, suggesting considerable within-person variance. To examine
gender effects, a t test for overall ER score
showed that women reported higher overall
ER than men, t(193) ¼ 2.23, p , .05: men ¼
3.80, women ¼ 3.57. Further, women were
higher than men in their ER on best friends,
t(193) ¼ 3.27, p , .01: women ¼ 4.50, men
¼ 4.15, and men reported higher levels of
ER on ‘‘no one’’ than women, t(193) ¼
2.88, p , .01: men ¼ 2.41, women ¼ 1.98.
Correlations of ER with Hirschfeld et al.Õs
(1977) IDI subscales, attachment indexes,
and well-being measures appear in Table 3.
As predicted, overall ER was negatively associated with assertion of autonomy, which
we construed as ‘‘detached independence’’
(r = .23, p , .01). ER on ‘‘no one’’ was
positively correlated with these scales (r ¼
.43, p , .001, assertion of autonomy; r ¼
.20, p , .01, lack of social self-confidence).
These results suggest that overall willingness
to rely on others is negatively associated
with both an insistence on interpersonal
independence and an excessive reliance on
others for esteem. Only one gender difference emerged on the IDI. Assertion of autonomy was higher for men than for women
(Ms ¼ 2.29 vs 1.97), t(193) ¼ 4.46,
p , .001.
Overall, ER was positively correlated with
attachment security (r ¼ .31, p , .001) and
negatively with attachment avoidance (r ¼
.31, p , .001). Conversely, ER on ‘‘no one’’
was correlated positively with avoidance (r ¼
.32, p , .001).
Also as predicted, overall ER was positively related to the well-being composite
(r ¼ .29, p , .001). To determine whether
ER was associated with well-being similarly
for males and females, we examined the
Gender ER interaction predicting the wellbeing composite with a multiple regression
analysis in which gender and ER total
scores were entered in Step 1 and their interaction in Step 2. The interaction term was
nonsignificant.
To test our hypothesis that support for
autonomy facilitates ER, and controlling
151
ANX
Emotional reliance
152
orientations undermine it, we explored the
associations between ER and relative differences in perceived parental autonomy support
and control. Using a statistical method
employed by Kasser and Ryan (1996), we first
created a residual score for each parent separately, in which each parent’s ER score was
regressed onto the overall ER score. Hence,
the remaining residual represents the degree to
which participants were willing to emotionally
rely on each parent relative to their overall
willingness to turn to all targets. Then, we
tested the relation between this relative ER on
a given parent and perceived autonomy support and control by that parent, predicting a pattern of target-specific covariance. Maternal
autonomy support was positively related to
one’s relative ER on mother (r ¼ .57, p , .01)
but not on father (r ¼ .04, ns). Similarly, paternal autonomy support strongly predicted relative ER on father (r ¼ .50, p , .01) but not on
mother (r ¼ .02, ns). Analogous correlations
showed that maternal control was negatively
associated with relative ER on mother (r ¼
.29, p , .01) but was unrelated to ER on
father (r ¼ .07, ns), and paternal control was
associated with less relative ER on father (r ¼
.23, p , .01) but unrelated to ER on mother
(r ¼ .05, ns). The effects were not moderated by gender.
Brief Discussion
Study 1 revealed that ER is positively related
to attachment security and negatively related
to both morbid dependency and extreme
independence or separateness. ER was highest for best friends (often considered the most
intimate relationship of college students; see
La Guardia et al., 2000), followed by mother,
father, and roommate. Gender differences
also emerged as predicted, with ER generally
higher in women. Also as expected, overall
ER was associated with greater well-being.
In contrast, ER on ‘‘no one’’ was associated
with more detached independence from
others and was negatively associated with
well-being. Finally, students reported greater
willingness to rely on the parent who they
perceived to be more autonomy supportive
and less controlling. These findings support
R. M. Ryan et al.
our primary hypotheses concerning the positive effects of ER, gender differences, and
selectivity. Moreover, they suggest the conceptual importance of distinguishing between concepts of dependence, autonomy,
and separateness as our theoretical review
suggested.
Study 2
Because most people have multiple relationship partners on whom they could potentially
rely for emotional support, our aim in Study
2 was to demonstrate variability of ER across
relationships and to account for this variability by demonstrating links to psychological
need fulfillment within relationships. Additionally, we wanted to further explore the
relations of ER and need satisfaction with
well-being. We used multilevel modeling
(hierarchial linear modeling [HLM]; Bryk &
Raudenbush, 1992) to simultaneously attend
to between- and within-person variance. We
expected ER, like other aspects of support
(Davis, Morris, & Krause, 1998), to significantly vary across relationships and that
relationships characterized by higher need
satisfaction would conduce toward greater
ER. Finally, we examined whether need satisfaction mediated the relation between ER and
well-being in general and within each specific target relationship.
Method
Participants and procedure
One hundred sixty undergraduates (105
women, 55 men) were recruited and earned
extra course credit for their participation.
These participants ranged in age from 18 to 29
years, with a mean of 20.28 years. They completed assessments in small group sessions.
Measures
Emotional reliance.
The same 10-item ER
assessment was used. However, in an effort to
remove any ambiguity about ER being a willingness to turn to others, item stems were
altered slightly to read, ‘‘If I were feeling
Emotional reliance
alone or depressed, I would be willing to turn
to..’’ All 10 items were rated for each target
figure, with the mean of items representing
the target score. As in Study 1, factor analyses
similarly supported a one-factor solution for
each target. Alphas were .92, .93, .89, and .88
for mother, father, romantic partner, and best
friend, respectively.
Need satisfaction scale.
Need satisfaction
scale (La Guardia et al., 2000) assesses the
presence of supports for the basic psychological needs of autonomy, competence, and relatedness. Participants rated how well their basic
needs were met within each target relationship
on 5-point Likert scales. The constructs of autonomy, competence, and relatedness are measured in parcels of 3 items each, yielding a
total of 9 items per target, or 36 items in total.
Alphas for the nine-item target scales were .91,
.90, .90, and .84 for mother, father, romantic
partner, and best friend, respectively.
Well-being.
Symptoms of depression were
assessed using the CES-D (a ¼ .91), and
vitality was assessed using the SVS (a ¼ .86)
as in Study 1. Life satisfaction was measured
with the widely used five-item Satisfaction
with Life Scale (SWLS; Diener, Emmons,
Larsen, & Griffin, 1985) (a ¼ .88). Items
were rated on a 9-point scale, with the mean
constituting the total score. Symptoms of
anxiety were measured using the 20-item
State-Trait Anxiety Scale (Spielberger,
Gorsuch, & Lushene, 1971) (a ¼ .93). Selfesteem was measured using the 10-item
general self-esteem scale of the Multidimensional Self-Esteem Inventory (O’Brien &
Epstein, 1988) (a ¼ .91). Items were rated on
7-point scales, with the mean of all items
indexing self-esteem [e.g., ‘‘I sometimes have
a poor opinion of myself’’ (reversed)].
A principal components factor analysis was
performed on measures of well-being. One factor was extracted, accounting for 71% of the
variance, with each factor loading greater than
.72. A unit-weighted composite well-being
score was computed by taking the mean of standardized scores on the variables loading on the
factor, and this was used to index overall wellbeing.
153
Results
Variability of ER
To establish that significant variability in ER
exists across relationships, multilevel modeling (HLM; Bryk & Raudenbush, 1992) was
utilized to simultaneously analyze the
between- and within-person levels. Results
suggested that 29% of the variance was contained at the between-person level, while 71%
of the variance was within-subjects variance
(variability in ER across partners) or error.
Paired samples t tests with Bonferroni corrections (p , .01) were conducted to test the
ER and need satisfaction mean differences by
target (Table 4). ER was highest on best friend
(M ¼ 6.16, SD ¼ .93), followed by romantic
partner (M ¼ 6.00, SD ¼ 1.03), mother (M ¼
5.79, SD ¼ 1.36), and father (M ¼ 5.27, SD ¼
1.48). For need satisfaction, participants
reported being most satisfied in relationships
with best friend (M ¼ 6.23, SD ¼ .73), followed by romantic partner (M ¼ 5.92, SD ¼
1.10) and mother (M ¼ 5.90, SD ¼ 1.10), and
least with father (M ¼ 5.56, SD ¼ 1.11). Further, gender differences were examined using
t tests (Bonferroni corrected at p , .008) for
ER and need satisfaction scales. Women were
significantly higher than men on overall ER,
t(156) ¼ 3.19, p , .01: women ¼ 5.95, men ¼
5.49, and best friends, t(155) ¼ 2.71, p , .01:
women ¼ 6.30, men ¼ 5.89. No significant gender differences emerged for need satisfaction.
To understand whether the within-person
variance was systematic, we constructed a
hierarchical linear model to predict ER from
need satisfaction at the level of the relationship, with need satisfaction centered on the
person’s own mean need satisfaction across
relationships. Further, in each equation, we
controlled for the effects of relationship
type with dummy codes ½e:g:; ER ¼ b0j þ
b1j ðneed satisfactionij Þ þ b2j ðD1Þ þ b3j ðD2Þ þ
b4j ðD3Þ þ rij . Each participant had a maximum of four relationships, with some having
fewer, yielding a total of 604 relationships
nested within 160 people. At the person level,
we assessed whether gender moderated the
intercept and the need satisfaction to ER
slope. Thus, two person-level equations were
constructed to test these effects ½b0j ¼ !00 þ
154
R. M. Ryan et al.
Table 4. Correlations and paired t tests comparing targets for emotional reliance (ER) and
need satisfaction (NS) (Study 2)
ER mother
ER father
ER best friend
ER romantic
partner
ER mother
ER father
ER best friend
ER
romantic
partner
—
.67**, t(149)
¼ 5.96**
.20, t(154)
¼ 2.82**
.21, t(122)
¼ 1.19
—
—
—
—
—
—
.13, t(148)
¼ 6.54***
.27*, t(118)
¼ 4.85***
—
—
.40**, t(121)
¼ 1.49
—
NS mother
NS mother
NS father
NS best friend
NS romantic
partner
—
.59**, t(150)
¼ 4.37**
.32**, t(156)
¼ 3.54**
.25*, t(133)
¼ .45
NS best friend
NS
romantic
partner
—
—
—
—
—
—
.26*, t(149)
¼ 6.97**
.40**, t(128)
¼ 3.63**
—
—
.50**, t(132)
¼ 4.58**
—
NS father
Note. For paired t tests, values are Bonferroni corrected at p , .01.
*p , .01. **p , .001.
!01 ðgenderÞþu0j ; b1j ¼!10 þ!11 ðgenderÞ þ u1j .
The intercept and need satisfaction effects in
the person-level equations were treated as
random (u nj), and dummy coded effects were
fixed. Thus, the random difference between
persons (u nj) was included for the estimates
of both the intercept and the need satisfaction slope.
Results indicated that, on average, women
reported greater overall ER, replicating findings
from Study 1. Regardless of gender, the effect
of need satisfaction on ER was significant, such
that on average, greater need satisfaction was
associated with greater willingness to rely in
relationships (coefficients for men and women,
.86 and .71, respectively, p , .001).
We then explored whether level of need
satisfaction was predictive of ER within specific relationships. Using multiple linear regression, residualized scores of ER and need
satisfaction for each target were created using
the Kasser and Ryan (1996) method described
in Study 1. Then, for each relationship and
controlling for gender, the residual ER score
was regressed onto the residual need satisfaction score. This procedure yielded significant
results for relationships to mother, F(1, 153)
¼ 189.31, b ¼ .74, p , .001; father, F(1,
149) ¼ 106.43, b ¼ .64, p , .001; romantic
partner, F(1, 122) ¼ 111.50, b ¼ .69, p ,
.001; and best friend, F(1, 152) ¼ 213.28,
b ¼ .76, p , .001, indicating that greater
need satisfaction is robustly associated with
greater ER within relationships, regardless of
relationship type.
Relations of ER, need satisfaction, and
well-being
Overall, need satisfaction was positively correlated with vitality (r ¼ .28, p , .001),
self-esteem (r ¼ .41, p , .001), and life satisfaction (r ¼ .48, p , .001) and negatively
correlated with symptoms of depression (r ¼
.43, p , .001) and anxiety (r ¼ .35, p ,
.001). ER also related to the same five wellbeing outcomes similarly (rs ¼ .26, .22, .47,
Emotional reliance
.30, .25, all ps , .001). Using the wellbeing composite as the dependent measure, we
tested whether gender moderated the impact of
ER or need satisfaction on well-being. As in
Study 1, regressions testing these moderation
effects were not significant. Controlling for the
main effects of gender, betas for the main
effects of ER for mother, father, romantic partner, and best friend were .39, .18, .19, .45, and
.39, and for need satisfaction were .47, .20, .29,
.53, and .50, respectively (all ps , .05).
Following methods outlined by Kenny,
Kashy, and Bolger (1998), mediation by need
satisfaction of the overall ER to well-being
relationship was then tested. ER significantly
predicted well-being, F(1, 156) ¼ 22.84, b ¼
.36, p , .001, and overall need satisfaction,
F(1, 156) ¼ 134.54, b ¼ .68, p , .001. When
well-being was regressed onto need satisfaction, controlling for ER, the effect was significant, F(1, 155) ¼ 18.21, b ¼ .41, p , .001.
However, the path from ER to well-being was
no longer significant with overall need satisfaction in the equation. The Sobel test revealed that the decrease in the beta coefficient
from .33 to .08 was significant (z ¼ 3.47, n ¼
155, p , .001), and because the coefficient
.08 was itself not significant, F(1, 155) ¼ .62,
ns, these results suggest full mediation.
Analysis of mediation at the level of individual relationships supported a similar model
for relationships to father, romantic partner,
and best friend. However, for mother, mediation by need satisfaction was not significant.
Because of this deviation from the general
pattern for mothers, we explored the interaction of maternal need satisfaction and ER
on mother in predicting well-being. This analysis revealed a significant result, F(1, 154) ¼
5.31, b ¼ .23, p , .05, indicating that those
who were highly need fulfilled and highly
emotionally reliant and those with low need
fulfillment and low ER had relatively higher
levels of well-being than all others. Those who
were low on need fulfillment yet highly emotionally reliant on their mothers showed the
lowest level of well-being. This pattern indicates that congruency—matching one’s level
of reliance on mother to level of need satisfaction—is better for mental health, while having
little of one’s psychological needs met by one’s
155
mother, yet still being willing to rely highly on
her, bodes badly for well-being. Having examined this interaction with mothers as the target,
we further examined the same interaction for
each of the remaining targets (three additional
tests), but no significant effects emerged.
Brief Discussion
Results confirmed that there is substantial
within-person variability in ER across relationships. They also suggest that ER is positively related to need satisfaction, such that the
more people experience fulfillment of needs
for autonomy, competence, and relatedness in
a relationship, the more they are willing to rely
on that person for emotional support.
This study also demonstrated that need satisfaction mediated the relationship between
ER and well-being. It appears that ER yields
benefits to well-being as a function of engendering greater feelings of relatedness, competence, and autonomy at the general level and
in specific relationships to father, romantic
partner, and best friend.
Interestingly, ER on mothers did not fit this
mediational pattern. In post hoc analyses, we
found instead a significant interaction between
ER and need satisfaction on well-being for
mothers. The best psychological adjustment
was found for individuals who are willing to
rely on mothers who are responsive to their
psychological needs and for those who are not
willing to rely on mothers who are not. Why
does this moderation pattern only emerge for
mothers? We suspect that this finding is
related to the critical role mothers play in the
lives of many students. Whereas students may
have greater discretion about whether to rely
on other targets who are not need supportive,
some students may feel compelled to turn to
their mother for support regardless of her
actual responsiveness. These data suggest that
such reliance can be problematic. Because it
was not predicted, this finding warrants replication in future research.
Study 3
Many researchers have begun to investigate
interpersonal relationships in individualistic
156
and collectivistic cultures. This interest has
been spurred by some characterizations of
western cultures, such as those of the United
States and Western Europe, as placing less
emphasis on relational ties and more on independence and self-focus (Markus & Kitayama,
1991). Conversely, collectivistic cultures,
particularly those from East Asia, are often
described as emphasizing group cohesion
and interpersonal harmony (Hofstede, 1980;
U. Kim, 1994; Triandis, 1989).
Although one might assume that the emphasis on interpersonal connection and cohesion in collectivistic cultures would beget
more ER on others, we suggest that crosscultural patterns and forms of reliance are not
monolithic. In fact, many western cultures
are characterized by an openness of emotional expression between intimates. By contrast, the norm in some East Asian cultures
(themselves hardly homogenous) is to restrict
the expression of personal emotions (Suh,
Diener, Shigehiro, & Triandis, 1998). In fact,
too much attention to one’s own thoughts and
feelings, irrespective of their impact on
others, may be regarded as selfish or immature (Y. Kim, Deci, & Zuckerman, 2002; Suh
et al., 1998). Thus, it could be argued that
a strong concern for the welfare of the group
and subordination of personal needs (Kwan,
Bond, & Singelis, 1997; Matsumoto, Kudoh,
Scherer, & Wallbott, 1988) may actually lead
to lower ER. Some researchers further suggest that collectivists are more likely than
individualists to assume that disclosure is not
necessary for one to cope with emotional
events. Markus and Kitayama (1991) state
that in individualistic cultures ‘‘it is the individual’s responsibility to Ôsay what’s on one’s
mindÕ if one expects to be attended to or
understood’’ while in collectivistic cultures
providers of support have ‘‘the willingness
and ability to feel and think what others are
feeling and thinking . without being told’’
(p. 229). These assumptions suggest that ER
may not be as salient or helpful in collectivistic cultures, but this is a largely untested
hypothesis.
A recent review by Burleson (2003) suggested, in fact, that some of these characterizations of cultural differences in the value
R. M. Ryan et al.
and form of emotional support may need to
be revisited. He reported data from two studies comparing Chinese and U.S. participants,
finding that members of both samples placed
high value on emotional support. He concluded that these findings call into question
whether cultural values like individualism
and collectivism truly foster differences in
what people desire from close relationships.
At the same time, he suggested that the way
in which emotional support may be conveyed
may differ across cultures.
As these viewpoints underscore, we expect
that ER may be valued differently across cultures, as well as expressed differently. However, because emotional sharing is a form of
intimacy and promotes awareness and integration, we suggest that ER will be associated with greater well-being across cultures.
Thus, we predict that although mean levels of
ER may differ, ER will yield positive outcomes across cultures and, as in Studies 1
and 2, across gender.
We examine samples from four cultures
selected to vary in their relative emphasis on
individualistic versus collectivistic practices.
We recruited student samples from Russia,
South Korea, Turkey, and the United States.
Existing literatures (e.g., V. Chirkov et al.,
2003; Triandis, 1989) suggest that U.S. participants are the most individualistic of the
four; South Koreans are the most collectivistic, with Turkish and Russian participants
between the other two.
Method
Participants and procedure
Participants2 were 557 students drawn from
four nations: 195 from a northeastern U.S.
university (52 men, 143 women; mean age ¼
19.6 years); 159 from two universities in
north central Russia (42 men, 117 women;
mean age ¼ 20.7 years); 94 from a midsize
university in southwestern Turkey (54 men,
2. The data collection for this Study 3 was also the basis
of an article by V. Chirkov et al. (2003) concerning cross-cultural aspects of autonomy for specific behaviors.
Emotional reliance
40 women; mean age ¼ 21.7 years), and 109
from a South Korean university (79 men, 30
women; mean age ¼ 20.2 years). All of the
universities were in urban areas and enrolled
economically diverse students. Eligible participants were those who were born in and
whose native language was that of the country
being studied. Demographic variables such as
marital status and family income (adjusted to
the national currency and level of wealth)
were comparable across samples, with the
exception of a somewhat lower income level
among Turkish participants. A trained assistant administered surveys in small groups,
and participants all received either course
credit or a small monetary compensation.
157
tion. Alphas for Korean, Russian, Turkish, and
U.S. samples were .40, .40, .50, and .64, respectively. Low alphas are typical for this measure, which aggregates across facets from a
longer multidimensional measure. It has nonetheless stood up in prior studies to crosscultural comparability criteria (e.g., Ryan et al.,
1999). A third measure was the 10-item SelfEsteem Scale (Rosenberg, 1965) assessing
global self-worth. Alphas for Korean, Russian,
Turkish, and U.S. samples were .85, .77, .83,
and .85, respectively. Finally, we used six
items from the Center for Epidemiological
Studies Depression Inventory (Radloff, 1977)
to assess depressive symptoms. Alphas for
Korea, Russia, Turkey, and the United States
were .86, .79, .77, and .89, respectively.
Measures
Persons highly fluent in both English and the
language of the country translated instruments that appeared originally in English,
and back translations were performed to test
the fidelity of the measures.
Emotional reliance.
We used a brief fouritem measure of ER (items 1–4 in Table 1).
On a 5-point Likert-type scale, participants
rated each of the 4 items for their relationships to family and friends. Alpha coefficients for Korea, Russia, Turkey, and the
United States were .76, .79, .76, and .85 for
family, and .75, .79, .85, and .80 for friends,
respectively.
Well-being.
Four indicators were chosen
to assess both hedonic and eudaimonic wellbeing (Ryan & Deci, 2001). Each has previously shown cross-cultural comparability
using Means and Covariance Structure
(MACS) analysis in U.S./Russian and U.S./
Korean comparisons (e.g., Y. Kim et al.,
2002; Ryan et al., 1999), criteria to be
retested, below. The first indicator was the
SWLS (Diener et al., 1985). Alpha coefficients for Korea, Russia, Turkey, and the
United States were .83, .75, .76, and .85,
respectively. The second measure was the
Short Index of Self-Actualization (Jones &
Crandal, 1986), a 15-item scale consisting of
varied indicators of growth and self-realiza-
Results
Gender and cultural differences
Omnibus ANOVAs tested effects of gender,
country, and their interaction on ER for family and friends. Means and standard deviations are shown in Table 5 for ER on family
and friends. ER family results revealed main
effects for gender, F(1, 556) ¼ 13.76, p ,
.001; country, F(3, 556) ¼ 10. 57, p , .001;
and a marginally significant interaction, F(3,
556) ¼ 2.18, p , .09. Overall, ER on family
was significantly higher for women (M ¼
3.73) than for men (M ¼ 3.40). Further, ER
on family was significantly higher for the
U.S. (M ¼ 3.79) and Russian (M ¼ 4.03)
samples than for Turkish (M ¼ 3.16) and
Korean (M ¼ 3.52) samples. Koreans were
significantly lower than all other samples on
ER on family. The t tests comparing gender
differences within country (Bonferroni correction p , .0025) showed that ER on family
was higher for Russian women than for Russian men, t(157) ¼ 3.34, p , .01: women
¼ 4.18, men ¼ 3.60, and for U.S. women
than for U.S. men, t(193) ¼ 3.40, p , .01:
women ¼ 3.94, men ¼ 3.40. Korean men did
not differ significantly from Korean women
and Turkish men did not differ from Turkish
women in ER on family.
Similarly, ER on friends revealed main
effects for gender, F(1, 556) ¼ 16.98, p ,
158
R. M. Ryan et al.
Table 5. Means and standard deviations of emotional reliance (ER) on family and friends
ER family
United States (total)
Men
Women
Russia (total)
Men
Women
Korea (total)
Men
Women
Turkey (total)
Men
Women
ER friends
n
M
SD
n
M
SD
194
52
142
159
42
117
109
79
30
94
54
40
3.79a
3.40
3.94
4.03bc
3.60
4.18
3.16abd
3.16
3.18
3.52cd
3.44
3.63
0.98
1.01
0.93
0.97
0.98
0.92
0.87
0.85
0.92
0.96
1.02
0.89
194
52
142
159
42
117
109
79
30
94
54
40
3.98ef
3.63
4.10
3.90gh
3.56
4.02
3.61egi
3.55
3.79
3.30fhi
3.31
3.27
0.80
0.99
0.68
0.94
1.05
0.92
0.77
0.73
0.86
0.84
0.68
1.02
Note. Subscripts within columns indicate means that differ from each other, at least p , .05.
.001; country, F(3, 556) ¼ 10.28, p , .001;
and a marginally significant interaction, F(3,
556) ¼ 2.51, p , .06. Overall, ER on friends
was significantly higher for women (M ¼
3.84) than for men (M ¼ 3.52). Further, ER
on friends was similar in the United States
(M ¼ 3.98) and Russia (M ¼ 3.90), while
both groups were higher than Koreans (M ¼
3.61) and Turks (M ¼ 3.30), with Turks significantly lower than all groups. The t tests
comparing the gender differences within
country of ER on friends (Bonferroni correction p , .0025) showed that Russian women
reported greater ER on friends than Russian
men, t(157) ¼ 2.54, p , .01: women ¼
4.02, men ¼ 3.56, and U.S. women reported
greater ER on friends than U.S. men, t(193)
¼ 3.16, p , .01: women ¼ 4.10, men ¼
3.63. No gender differences were found
within the Korean and Turkish samples.
MACS analyses and model testing
To ensure measurement equivalence and
comparability of constructs, we employed
MACS analyses (Little, 1997, 2000). MACS
procedures test the factorial invariance of
measurement models across samples by placing equality constraints on both intercepts
and factor loadings but not on the unique
parameters and correlations between con-
structs. The goodness-of-fit index (GFI), the
root mean square error of approximation
(RMSEA), the incremental fit index (IFI),
and the comparative fit index (CFI) were
used to assess the model fit.3
For ER, two latent constructs were created
from two observed indicators pertaining to
family and friends, with each indicator consisting of two items. These two scales were
invariant across samples, v2(22, N ¼ 553) ¼
56.5, p , .001, RMSEA ¼ .10 (90% CI ¼
.07 to .14, CFI ¼ .94; IFI ¼ .94), indicating
that ER measures were comparable across
cultures.
ER and well-being.
For well-being, three
observed indicators for each of the depression
and self-actualization variables and two indicators for each of the self-esteem and life satisfaction variables were used to test models
3. For the indexes GFI, IFI, and CFI, values of .90 or
higher were considered acceptable, and for the invariance test, we used a difference-in-fit criterion of
, .05 (Little, 1997). For RMSEA, we used the conventional decision rule: , .05 represents a small error
of approximation and a very good fit of the model,
.10 . RMSEA . .05 suggests a reasonable error of
approximation, and . .10 indicates poor fit. Because
of its high sensitivity to the trivial discrepancies
between covariance matrices, the likelihood ratio chisquare statistic was not used in the decision making
about the model acceptability.
Emotional reliance
of cross-cultural comparability. The factorial
invariance fit for the measurement models
across four countries was acceptable
(RMSEA ¼ .063; CFI ¼ .96, IFI ¼ .96).
However, the model based on strong factorial
invariance fell just below the optimal levels,
RMSEA ¼ .11 (90% CI ¼ .095 to .012), CFI
¼ .89, IFI ¼ .89. Despite this, we decided to
treat the well-being constructs as comparable
because the measurement model yielded relatively high factorial invariance (Ghorpade,
Hattrup, & Lackritz, 1999), and prior research using MACS with these scales
revealed comparability across three of these
cultures (e.g., V. I. Chirkov & Ryan, 2002;
Y. Kim et al., 2002; Ryan et al., 1999).
To test the relations of ER to well-being,
two multiple regressions were performed. In
each, we used the composite well-being index
as the dependent variable and friend or family
ER, country contrast codes (used to test comparisons of constructs by country membership), and interactions of ER by country code
as the independent variables. Contrast 1 compared U.S. to Russian, Korean, and Turkish
participants taken together; Contrast 2 compared Russians to Koreans and Turks, and
Contrast 3 evaluated Koreans in relation to
Turks. Thus, collectively, these contrasts capture all the possible variance due to betweencountry differences.
Regression results revealed main effects
for Contrast 1 (comparing the United States
vs. all others) and Contrast 2 (comparing
Russians to Koreans and Turks) on wellbeing. Well-being was greater, F(3, 540) ¼
64.84, b ¼ .32, p , .01, among U.S. participants (latent composite mean set at 0.00) compared with Russian (latent mean ¼ .24),
Turkish (latent mean ¼ .57), or Korean (latent mean ¼ .50) participants. Also, Russians had higher composite well-being than
either Turks or Koreans, F(3, 540) ¼ 20.90, b
¼ .19, p , .01.
For ER on family, a main effect emerged
such that greater ER with family was associated with greater well-being, F(1, 543) ¼
21.69, b ¼ .19, p , .01. A main effect also
emerged for ER on friends, with greater ER
associated with greater well-being, F(1, 543)
¼ 15.25, b ¼ .16, p , .01. Of the six poten-
159
tial contrast codes by ER interactions only
one emerged as significant, namely ER on
Friends Contrast 1 (b ¼ .38, p , .01),
indicating that ER on friends was more associated with well-being in the U.S. than in the
other three samples.
Subsequently, within-country regression
analyses were conducted separately for each
indicator of ER with gender and its interaction
with ER variables as predictors of well-being.
Consistent with the descriptive analyses previously reported, a significant main effect for
gender emerged only in the Russian sample,
F(1, 158) ¼ 4.81, b ¼ .17, p , .05, showing
that men (M ¼ 8.42) reported greater wellbeing than women (M ¼ 7.84). The effect of
ER on well-being was moderated by gender
in only one case. In the Russian sample, a significant interaction of ER with family by gender on well-being indicated that the effect
was more pronounced for women than for
men, F(1, 159) ¼ 4.43, b ¼ .63, p , .05.
Brief Discussion
Using MACS analyses, we found that the
construct of ER as measured was psychometrically comparable across cultures. As expected, means on ER differed by country and
gender. ER on friends was highest in the
United States and lowest in Turkey, whereas
ER on family was highest among Russians
and lowest among Koreans. Where gender
differences emerged, women were higher than
men. However, the relations of ER on family
and friends to well-being were similar across
cultures: the more emotionally reliant, the
greater the benefits for well-being. Gender
also did not generally moderate the effects of
ER on well-being. An exception was that reliance on family was associated with greater
well-being outcomes for Russian women than
for men, whereas the reverse was true in the
United States. The observed mean-level variations suggest that cultural context and gender
may impact upon both the level of ER and to
whom ER is directed. Yet, the findings also
suggest that ER on family and friends is associated similarly with well-being across these
cultural settings despite variations in collectivism versus individualism.
160
General Discussion
Emotional support has been identified as one
of the most critical types of support for adaptation to stress and sustained well-being (Burleson, 2003; Reis & Collins, 2000; Wills &
Shinar, 2000). However, the receipt of emotional support is undoubtedly influenced by
one’s willingness to rely on others during
emotionally laden events. Although it would
seem that a willingness to turn to others
might be beneficial, theories and measures of
ER have often infused the construct of reliance
with other attributes such as neediness, insecurity, or detachment, thus making it a negative
construct. In this research, we investigated
people’s self-reported willingness to turn to
others during emotionally salient times, which
we termed emotional reliance. ER is viewed
as both an individual difference and an attribute that varies across a person’s different
relationships as a function of their quality. Our
interest was to explore the association of ER
with psychological well-being, and the qualities of interpersonal relationships within which
ER is more likely to occur.
In accord with our major hypothesis, ER
was positively associated with well-being. This
relation was replicated in all three studies. Having distinguished ER from earlier constructs
entailing more problematic forms of emotional
dependence, it appears that being willing to
seek out interpersonal support during emotional
events may be neither harmful nor indicative of
psychological problems. The findings also underscore the importance of keeping constructs
such as autonomy, detachment, and insecurity
distinct in measurement and theory concerning
support and dependence. Autonomy and dependence are not inherently antithetical and,
indeed, people often solicit care from others by
choice and with positive effects.
Results across studies also demonstrated
that gender and culture can affect levels of
ER. Gender theories have suggested differences between men and women in the function
of emotional support and the resultant effects
of emotional support on well-being, with
women being more apt to seek emotional support than men, and perhaps benefit more from
it. Although we did not invariantly find
R. M. Ryan et al.
gender differences in level of ER, when we
did, they showed greater ER for women, as past
theories have suggested. However, the beneficial effects of ER on well-being were similar
across gender, despite mean differences.
Cross-cultural theories have also suggested
that the mode and functional impact of ER
may differ based on cultural orientation. In
much of this work, it is suggested that collectivist cultures promote more concern with
othersÕ emotions but less willingness to turn
to others explicitly for emotional support.
Four samples drawn from countries that varied along the collectivism versus individualism dimension showed some differences in
ER for family and friends. South Koreans, as
these theories would predict, indicated they
would be less likely than Russians, Turks, or
Americans to emotionally rely on family members. Also as these theories would suggest,
Americans reported more ER on friends than
did other samples. However, there were no
monolithic trends; gender differences tended
to outstrip cultural ones, and ER appeared to
be both comparably measurable and meaningful within each culture. Indeed, our findings
fit with those of Burleson (2003), who suggested that emotional support is desirable in
close relationships across the spectrum of individualism and collectivism.
These results highlight the potential importance of socialization for ER. Women may
typically be encouraged to be emotionally
engaging, oriented to relationships, and disclose more than men and thus, as demonstrated, report greater willingness to turn to
others when emotional events occur. Norms
about ER have also been claimed to vary across
cultures, and some have argued that there is
less emphasis on such personal forms of reliance among collectivists (Suh et al., 1998).
Nonetheless, despite mean differences in the
endorsement of ER, greater levels of ER were
reliably associated with greater mental health.
Within-person variations in ER were evident across the studies. We demonstrated
important links between ER and need satisfaction, such that greater need support within
a given relationship is associated with greater
ER with that person. Need satisfaction was
also shown to mediate, or in the case of rela-
Emotional reliance
tionships to mothers moderate, the relation of
ER to well-being. Future research is needed
to further test these results in cross-age and
cross-cultural samples. However, these mediation and moderation effects are intriguing.
In general, the mediation models suggested
that ER may work by fostering feelings of
relatedness, competence, and autonomy and
thus diminishing feelings of depression, loneliness, and insecurity and raising life satisfaction and self-esteem. These results fit with
the theoretical benefits outlined by Reis and
Franks (1994) and Wills and Shinar (2000),
among others. However, our one moderation
result, found only with mothers, suggests that
if one must rely emotionally on someone
who is not supportive of basic needs, reliance
may indeed not yield benefits. This intriguing
but unexpected result warrants replication.
Future directions for research include the
study of ER in different age-groups and developmental epochs. At this point, we have prior
related work with adolescents that suggests
that ER on parents, teachers, and friends may
be important for different outcomes (Ryan
et al., 1994). We also have pilot work with
adults suggesting the benefits of ER on wellbeing and the importance of spouses or partners in adultsÕ patterns of ER. Studies with
varied ages will be important to understanding how both developmental changes and
changes in social support constellations affect
both ER and emotional support more generally. Other methods, such as longitudinal or
diary methods, may also further illuminate the
nature of variability in ER and need satisfaction (as evidenced in Study 2) by predicting
well-being outcomes across time in both ER
and need satisfaction. The present findings
also speak to the need for future research on
emotional support to focus on within-person
variation, which in the case of ER, outstripped
the variance between persons.
In sum, our results suggest that although
levels of ER may differ across gender and
culture, individuals who report more willingness to turn to others when emotional events
occur also report greater well-being. ER
appears, however, to be used selectively with
others who are perceived as supporting needs
for autonomy, competence, and relatedness.
161
Because emotions play such an important
role in health and self-regulation, understanding the willingness to share and express them
has implications for a variety of relational
and clinical contexts and supplies an interesting agenda for future research.
References
Baumeister, R. F., & Leary, M. R. (1995). The need to
belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin,
117, 497–529.
Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical
linear models. Newbury Park, CA: Sage.
Burleson, B. R. (2003). The experience and effects of
emotional support: What the study of cultural and
gender differences can tell us about close relationships, emotion and interpersonal communication.
Personal Relationships, 10, 1–24.
Burleson, B. R., & Kunkel, A. W. (1996). The socialization of emotional support skills in childhood. In
G. R. Pierce, B. R. Sarason, & I. G. Sarason (Eds.),
Handbook of social support and the family (pp. 105–
140). New York: Plenum Press.
Butzel, J. S., & Ryan, R. M. (1997). The dynamics of volitional reliance: A motivational perspective on dependence, independence, and social support. In G. R.
Pierce, B. Lakey, I. G. Sarason, & B. R. Sarason
(Eds.), Sourcebook of social support and personality
(pp. 49–67). New York: Plenum Press.
Caldwell, M., & Peplau, L. (1982). Sex differences in
same-sex friendships. Sex Roles, 8, 721–732.
Carstensen, L. L., Isaacowitz, D. M., & Charles, S. T.
(1999). Taking time seriously: A theory of socioemotional selectivity. American Psychologist, 54, 165–181.
Chirkov, V. I., & Ryan, R. M. (2002). Parent and teacher
autonomy support in Russian and U.S. adolescents:
Common effects on well-being and academic motivation. Journal of Cross Cultural Psychology, 32,
618–635.
Chirkov, V., Ryan, R. M., Kim, Y., & Kaplan, U. (2003).
Differentiating autonomy from individualism and
independence: A self-determination theory perspective on internalization of cultural orientations and
well-being. Journal of Personality and Social Psychology, 84, 97–110.
Cohen, S., Sherrod, D. R., & Clark, M. S. (1986). Social
skills and the stress-protective role of social support.
Journal of Personality and Social Psychology, 50,
963–973.
Cohen, S., & Wills, T. A. (1985). Stress, social support,
and the buffering hypothesis. Psychological Bulletin,
98, 310–357.
Davis, M. H., Morris, M. M., & Krause, L. A. (1998).
Relationship-specific and global perceptions of social
support: Associations with well-being and attachment. Journal of Personality and Social Psychology,
74, 468–481.
Deci, E. L. (1975). Intrinsic motivation. New York: Plenum
Press.
Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation
and self-determination in human behavior. New York:
Plenum Press.
162
Deci, E. L., & Ryan, R. M. (2000). The ‘‘what’’ and
‘‘why’’ of goal pursuits: Human needs and the selfdetermination of behavior. Psychological Inquiry,
11, 227–268.
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S.
(1985). The satisfaction with life scale. Journal of
Personality Assessment, 49, 71–75.
Emmons, R. A., & Colby, P. M. (1995). Emotional conflict and well-being: Relation to perceived availability, daily utilization, and observer reports of social
support. Journal of Personality and Social Psychology, 68, 947–959.
Ghorpade, J., Hattrup, K., & Lackritz, J. R. (1999). The
use of personality measures in cross-cultural research:
A test of three personality scale across two countries.
Journal of Applied Psychology, 84, 670–679.
Goldsmith, D. J., & Fitch, K. (1997). The normative context of advice as social support. Human Communication Research, 23, 454–476.
Greenberg, M. T., Siegel, J. M., & Leitch, C. L. (1983).
The nature and importance of attachment relationships to parents and peers during adolescence. Journal of Youth and Adolescence, 12, 373–386.
Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52, 511–524.
Helgeson, V. S. (1994). Relation of agency and communion to well-being: Evidence and potential explanations. Psychological Bulletin, 116, 412–428.
Hirschfeld, R. M., Klerman, G. L., Gough, H. G., Barrett,
J., Korchin, S. J., & Chodoff, P. (1977). A measure
of interpersonal dependency. Journal of Personality
Assessment, 41, 610–621.
Hofstede, G. (1980). Culture’s consequences. Beverly
Hills, CA: Sage.
Jones, A., & Crandal, R. (1986). Validation of a short
index of self-actualization. Personality and Social
Psychology Bulletin, 12, 63–73.
Jordan, J. V., Kaplan, A. G., Miller, J. B., Stiver, I. P., &
Surrey, J. L. (1991). Women’s growth in connection:
Writings from the Stone Center. New York: Guilford.
Kashima, Y., Yamaguchi, S. U., Choi, S. C., Gelfand,
M. J., & Yuki, M. (1995). Culture, gender, and self:
A perspective from individualism collectivism research. Journal of Personality and Social Psychology, 69, 925–937.
Kasser, T., & Ryan, R. M. (1996). Further examining the
American dream: Differential correlates of intrinsic
and extrinsic goals. Personality and Social Psychology Bulletin, 22, 80–87.
Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data
analysis in social psychology. In D. T. Gilbert,
S. T. Fiske, & G. Lindzey (Eds.), The handbook of
social psychology: Vol. 1. (4th ed., pp. 233–265).
New York: McGraw-Hill.
Kim, U. (1994). Individualism and collectivism: Conceptual clarification and elaboration. In K. Uichol,
H. C. Triandis, C. Kagitcibasi, S. C. Choi, & G. Yoon
(Eds.), Individualism and collectivism: Theory,
method and applications (pp. 19–40). Thousand
Oaks, CA: Sage.
Kim, Y., Deci, E. L., & Zuckerman, M. (2002). The
development of the Self-Regulation of Withholding
Negative Emotions Questionnaire. Educational and
Psychological Measurement, 62, 316–336.
Kwan, V. S. Y., Bond, M. H., & Singelis, T. M. (1997).
Pancultural explanations for life satisfaction: Adding
R. M. Ryan et al.
relationship harmony to self-esteem. Journal of Personality and Social Psychology, 73, 1038–1051.
La Guardia, J. G., Ryan, R. M., Couchman, C., & Deci,
E. L. (2000). Within-person variation in security of
attachment: A self-determination theory perspective on
attachment, need fulfillment, and well-being. Journal
of Personality and Social Psychology, 79, 367–384.
Little, T. D. (1997). Mean and covariance structure
(MACS) analysis of cross-cultural data: Practical and
theoretical issues. Multivariate Behavioral Research,
32, 53–76.
Little, T. D. (2000). On the comparability of constructs
in cross-cultural research: A critique of Cheung and
Rensvold. Journal of Cross Cultural Psychology, 31,
213–219.
Markus, H. R., & Kitayama, S. (1991). Culture and the
self: Implications for cognition, emotion, and motivation. Psychological Review, 98, 224–253.
Matsumoto, D., Kudoh, T., Scherer, K., & Wallbott, H.
(1988). Antecedents of and reactions to emotions in
the United States and Japan. Journal of Cross Cultural Psychology, 19, 267–286.
McBride-Chang, C., & Chang, I. (1998). Adolescentparent relations in Hong Kong: Parenting styles, emotional autonomy, and school achievement. Journal of
Genetic Psychology, 159, 421–436.
McQueen, A., Getz, J. G., & Bray, J. H. (2003). Acculturation, substance use and deviant behavior: Examining separation and family conflict as mediators.
Child Development, 74, 1737–1750.
Nix, G., Ryan, R. M., Manly, J. B., & Deci, E. L. (1999).
Revitalization through self-regulation: The effects of
autonomous versus controlled motivation on happiness and vitality. Journal of Experimental Social
Psychology, 35, 266–284.
O’Brien, E., & Epstein, S. (1988). The Multidimensional
Self-Esteem Inventory: Professional Manual. Odessa,
FL: Psychological Assessment Resources.
Pierce, G. R., Sarason, I. G., & Sarason, B. R. (1991).
General and relationship-based perceptions of social
support: Are two constructs better than one? Journal
of Personality and Social Psychology, 61, 1028–
1039.
Radloff, L. S. (1977). The CES-D scale: A self-report
depression scale for research in the general population. Applied Psychological Measurement, 1, 385–
401.
Reis, H. T., & Collins, N. (2000). Measuring relationship
properties and interactions relevant to social support.
In S. Cohen, L. G. Underwood, & B. H. Gottlieb
(Eds.), Social support measurement and intervention
(pp. 136–192). New York: Oxford University Press.
Reis, H. T., & Franks, P. (1994). The role of intimacy
and social support in health outcomes: Two processes
or one? Parental Relationships, 2, 185–197.
Robbins, R. J. (1994). An assessment of perceived parental autonomy—support and control. Unpublished
doctoral dissertation, University of Rochester,
Rochester, NY.
Rook, K. S. (1987). Social support versus companionship: Effects on life stress, loneliness, and evaluations
by others. Journal of Personality and Social Psychology, 52, 1132–1147.
Rosenberg, M. (1965). Society and the adolescent selfimage. Princeton, NJ: Princeton University Press.
Russell, D., Peplau, L. A., & Cutrona, C. E. (1980). The
revised UCLA Loneliness scale: Concurrent and dis-
Emotional reliance
criminant validity evidence. Journal of Personality
and Social Psychology, 39, 472–480.
Ryan, R. M. (1993). Agency and organization: Intrinsic
motivation, autonomy, and the self in psychological
development. In J. Jacobs (Ed.), Nebraska symposium
on motivation: Developmental perspectives on motivation, Vol. 40 (pp. 1–56). Lincoln: University of
Nebraska Press.
Ryan, R. M., Chirkov, V. I., Little, T. D., Sheldon, K. M.,
Timoshina, E., & Deci, E. L. (1999). The American
dream in Russia: Extrinsic aspirations and wellbeing in two cultures. Personality and Social Psychology Bulletin, 25, 1509–1524.
Ryan, R. M., & Deci, E. L. (2000). Self-determination
theory and the facilitation of intrinsic motivation,
social development, and well-being. American Psychologist, 55, 68–78.
Ryan, R. M., & Deci, E. L. (2001). On happiness and
human potentials: A review of research on hedonic
and eudaimonic well-being. Annual Review of Psychology, 52, 141–166.
Ryan, R. M., & Frederick, C. M. (1997). On energy, personality, and health: Subjective vitality as a dynamic
reflection of well-being. Journal of Personality, 65,
529–565.
Ryan, R. M., & La Guardia, J. G. (2000). What is being
optimized?: Self-determination theory and basic psychological needs. In S. Qualls & R. Abeles (Eds.),
Psychology and the aging revolution: How we adapt
to longer life (pp. 145–172). Washington, DC: APA
Books.
Ryan, R. M., & Lynch, J. (1989). Emotional autonomy
versus detachment: Revisiting the vicissitudes of adolescence and young adulthood. Child Development,
60, 340–356.
Ryan, R. M., & Solky, J. A. (1996). What is supportive about social support? On the psychological needs
for autonomy and relatedness. In G. R. Pierce, B. R.
Sarason, & I. G. Sarason (Eds.), Handbook of social
support and the family (pp. 249–267). New York:
Plenum Press.
Ryan, R. M., Stiller, J. D., & Lynch, J. (1994). Representations of relationships to teachers, parents, and
friends as predictors of academic motivation and selfesteem. Journal of Early Adolescence, 14, 226–249.
Samter, W., Whaley, B. B., Mortenson, S. T., & Burleson,
B. R. (1997). Ethnicity and emotional support in samesex friendship: A comparison of Asian-Americans,
African-Americans, and Euro-Americans. Personal
Relationships, 4, 413–430.
163
Shields, S. A. (1995). The role of emotion beliefs and
values in gender development. In N. Eisenberg (Ed.),
Review of personality and social psychology: Vol.
15. Social development (pp. 212–232). Thousand
Oaks, CA: Sage.
Simpson, J. A. (1990). The influence of attachment styles
on romantic relationships. Journal of Personality and
Social Psychology, 62, 434–446.
Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E.
(1971). The State-Trait Anxiety Inventory. Palo Alto,
CA: Consulting Psychologists Press.
Steinberg, L., & Silverberg, S. (1986). The vicissitudes of
autonomy in adolescence. Child Development, 57,
841–851.
Suh, I., Diener, E., Shigehiro, O., & Triandis, H. C.
(1998). The shifting basis of life satisfaction judgments
across cultures: Emotions versus norms. Journal of
Personality and Social Psychology, 74, 482–493.
Taylor, J. A. (1953). A personality scale of manifest anxiety. Journal of Abnormal and Social Psychology,
48, 285–290.
Ting-Toomey, S. (1991). Intimacy expressions in three
cultures: France, Japan, and the United States. International Journal of Intercultural Relations, 15, 29–46.
Triandis, H. C. (1989). The self and social behavior in
differing cultural contexts. Psychological Review,
96, 506–520.
Turner, R. A., Irwin, C. E., & Millstein, S. G. (1991).
Family structure, family processes and experimenting
with substances during adolescence. Journal of
Research on Adolescence, 1, 83–106.
Uchino, B. N., Cacioppo, J. T., & Kiekolt-Glaser, J. K.
(1996). The relationship between social support and
physiological processes: A review with emphasis on
underlying mechanisms and implications for health.
Psychological Bulletin, 119, 488–531.
White, R. W. (1963). Ego and reality in psychoanalytic
theory. New York: International Universities Press.
Wills, T. A. (1998). Social support and health in women.
In E. Blechman & K. Brownell (Eds.), Behavioral medicine and women (pp. 118–123). New York: Guilford.
Wills, T. A., & Shinar, O. (2000). Measuring perceived
and received social support. In S. Cohen, L. G.
Underwood, & B. H. Gottlieb (Eds.), Social support
measurement and intervention (pp. 66–135). New
York: Oxford University Press.
Windle, M. (1992). Temperament and social support in
adolescence: Relations with depressive symptoms
and delinquent behaviors. Journal of Youth and Adolescence, 21, 1–21.