When Social Networking Is Not Working: Individuals With Low Self-Esteem Recognize

but Do Not Reap the Benefits of Self-Disclosure on Facebook

Author(s): Amanda L. Forest and Joanne V. Wood
Source: Psychological Science , MARCH 2012, Vol. 23, No. 3 (MARCH 2012), pp. 295-302
Published by: Sage Publications, Inc. on behalf of the Association for Psychological
Science

Stable URL: https://www.jstor.org/stable/41441787

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/41441787?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Sage Publications, Inc. and Association for Psychological Science are collaborating with JSTOR
to digitize, preserve and extend access to Psychological Science

This content downloaded from

147.46.181.112 on Mon, 31 Oct 2022 11:33:12 UTC
All use subject to https://about.jstor.org/terms
 aps яшшшт I ASSOCIATION FOR
яшшшт I ASSOCIATION FOR
Research Article psychological science

Psychological Science
23(3) 295-302
When Social Networkin
© The Author(s) 20 1 2
Reprints and permission:
Individuals With Low Self-Esteem sagepub.com/joumalsPermissions.nav
DOI: 1 0.1 177/095679761 1429709

Recognize but Do Not Reap the Benefits of

http://pss.sagepub.com

®SAGE
Self-Disclosure on Facebook

Amanda L. Forest and Joanne V. Wood

University of Waterloo

Abstract

The popular media have publicized the idea that social networking Web sites (e.g., Facebook) may enrich the interpe
of people who struggle to make social connections. The opportunity that such sites provide for self-disclosure - a
component in the development of intimacy - could be especially beneficial for people with low self-esteem, who are
hesitant to self-disclose and who have difficulty maintaining satisfying relationships. We suspected that posting on
would reduce the perceived riskiness of self-disclosure, thus encouraging people with low self-esteem to express
more openly. In three studies, we examined whether such individuals see Facebook as a safe and appealing medium
disclosure, and whether their actual Facebook posts enabled them to reap social rewards. We found that although
with low self-esteem considered Facebook an appealing venue for self-disclosure, the low positivity and high negati
disclosures elicited undesirable responses from other people.

Keywords
social networking, Facebook, self-esteem, self-disclosure, interpersonal relationships, social interaction

Received 4/ 1 1 / 1 I ; Revision accepted 9/ 1 6/ 1 1

Since its inception in February of 2004, the social networking

with low self-esteem are more socially anxious, introverted, and
shy than people with high self-esteem (Leary & MacDonald,
Web site Facebook has revolutionized interpersonal communi-
cation and relationship maintenance. With more than 50%
2003). of
Although both types of individuals seem to desire connec-
tion equally (e.g., Anthony, Wood, & Holmes, 2007), people
its 500 million users logging in on any given day (Facebook,
2011b), Facebook appears to be fulfilling its missionwith
"to give
low self-esteem feel lonelier and have less satisfying and
people the power to share and make the world more stable
open and
relationships than do people with high self-esteem (Leary
connected" (Facebook, 2011a). & MacDonald, 2003; Wood, Hogle, & McClellan, 2009). The
Popular media outlets have latched on to the idea that
lowerFace-
quality of the former group's relationships is troubling not
book may serve as a "social lubricant" ("Facebook a Big
onlyHit,"
because feeling connected to other people is considered
2010, para. 18) that helps people who struggle with arelation-
"fundamental human motivation" (Baumeister & Leary, 1995,
ships - the shy, the lonely, and people with low self-esteem -
p. 497), but also because rich social networks promote better
mental
connect with others in a comfortable environment (e.g., health (e.g., Baumeister & Leary, 1995) and even better
Dailey,
2009; Rosenwald, 2011). Some therapists have even specu-health (e.g., Uchino, Cacioppo, & Kiecolt-Glaser,
physical
1996).exam-
lated about the utility of Facebook for such groups. For
ple, Jonathan Dalton noted that "Shy people have difficulty It seems possible that Facebook could help people with low
finding topics to talk about. . . Facebook gives you self-esteem
a starting lead more fulfilling social lives by providing an
point" (as quoted in Rosenwald, 20 1 1 , para. 1 5). Butopportunity
do peoplefor self-disclosure, which is considered crucial
who have difficulty making social connections use these to thesites
development of intimacy (Reis & Shaver, 1988). In
in ways that enrich their interpersonal lives?
To examine this question, we focused on a group of people
Corresponding Author:
who could benefit from the opportunity to enhance their social
Amanda L. Forest, University of Waterloo, Department of Psychology, 200
connections: individuals with low self-esteem - that University
is, peopleAve. West, Waterloo, Ontario N2L 3G I , Canada
whose overall liking for themselves is relatively low. People
E-mail: ahogle@uwaterloo.ca

This content downloaded from

147.46.181.112 on Mon, 31 Oct 2022 11:33:12 UTC
All use subject to https://about.jstor.org/terms
 296 Forest, Wood

contexts evidence that does exist is mixed. Research in non-Facebook

outside of
associated
contexts has with
found that people who are depressed likab (Coyne,
well as with relations
1976) or high in negative affect (Bell, 1978; Sommers, 1984)
& Hendrick,
are not well liked. However,2004).
it is not clear that expressing neg- H
individualativityshares
is what makes other people dislike themperso
- perhaps it is
ers may disapprove
their reassurance-seeking (Segrin & Abramson, 1994) or their or
People with
failure to show support for low
others (Gotlib & Robinson, sel
1982).
about self-disclosing
Other research suggests that expressing negativity may be
observedendearing:
that People who report a high such
willingness to express i
Whereas people
negative emotions have more intimate with
relationships than peo- h
their good qualities,
ple who are less willing to do so (Graham, Huang, Clark, & p
on avoiding
Helgeson, 2008). revealing
Moreover, it seems likely that the impact of
tion seems to
expressing negativity guide
depends on who receives the disclosure. m
esteem display
People who care about the discloser aretowar
likely to be more inter-
because they are
ested, whereas casual acquaintances espec
and strangers may be put
devaluedoff (cf.
(Leary,
Clark & Taraban, 1991). An interesting aspect of Tam
Face-
faced with
book is that it"risky"
allows people's disclosures to reach a wide audi- op
with ence, including peopleclose
people with whom the discloser has to
an
them emotionally close relationship and people whom
vulnerable to the discloserbe
people with low
barely knows. In Study self-
3, we examined the effects of positive
self-protection
and negative disclosure on strangers and(Fores
on participants' real
low Facebook friends.
self-esteem self-d
esteem do (Gaucher
Facebook 's enormous popularity makes it an important e
Because people
context in which to examine the interpersonal withconsequences of
person disclosure. With over 500
than domillion active Facebook users creat-
people
nity to communicate
ing, on average, 90 pieces of content per month (Facebook,
especially valuable
201 lb), there is no question that Facebook is affecting people's f
feelings online
interpersonal lives. may
Unlike people
In the research reported here, we soughtinto explore whetherfac
who disclose
people with low self-esteem do, their
in fact, see Facebook as a safe f
people's potentially
medium for self-disclosure (i.e., one that offers low risk for
responseshurt orof the
rejection; Study 1) recip
and whether such individuals capi-
perhaps censored.
talize on the opportunity that Facebook provides by express- D
may be a way
ing themselves in
there in ways that enhance their social wh
relate to other individuals without the risks of face-to-face connections (Studies 2 and 3).
disclosure.

At the same time, there is a danger that if people with low

Study
self-esteem feel free to disclose their feelings openly on Face-
I
book, they may express a lot of negativity. Relative to people
Because of their self-protective orientation, people with low
self-esteem may find Facebook especially appealing. They
with high self-esteem, people with low self-esteem experience
higher levels of nearly all negative emotions and are morecan think carefully about what to say and can avoid the "in-
likely to have anxiety and depressive disorders (Learythe-moment"
& awkwardness of face-to-face interactions. We
MacDonald, 2003). People with low self-esteem are alsohypothesized
less that although both people with low self-esteem
likely to savor positive moods (Wood, Heimpel, & Michela,
and people with high self-esteem see Facebook as a place
2003) and to repair negative moods (Heimpel, Wood, Mar-
to connect with other individuals, the former perceive Face-
shall, & Brown, 2002). Although no researchers have reported
book to be a safer context for expressing their emotions and as
associations between self-esteem and negative expressivity,
offering more advantages - such as the opportunity to receive
there is good reason to think that if people with low self-
support - than in-person disclosures do.
esteem do express themselves on Facebook, their disclosures
may be more negative than the disclosures of people with high
Method
self-esteem.

What impact would expressing negativity on FacebookEighty undergraduate Facebook users (17 male, 58 female,
have on a person's interpersonal relationships? The answer is 5 undisclosed; mean age = 21.35 years) completed the Rosen-
not clear. Surprisingly little research has addressed the inter-berg Self-Esteem Scale (Rosenberg, 1965). Using a scale that
personal consequences of negative self-disclosure, and theranged from 1 , not at all true , to 7, very true , participants then

This content downloaded from

147.46.181.112 on Mon, 31 Oct 2022 11:33:12 UTC
All use subject to https://about.jstor.org/terms
 Self-Esteem and Social Networking 297

answered questions from Facebook

four users post status updates - usually
measures about singletheir
sentences
to share
tions of Facebook. First, news, thoughts, or feelings
participants - the posts are the
reported visible to
d
which Facebook enabledallthemof their Facebook
to friends
express - people they have added to their
themselves
list ofother
a = .88) and connect with Facebook contacts.
people (3 items, a = .7
they completed a 9-item measure
Each participant's (awas
set of updates =content
.87)coded.of
Coders the
safety of were undergraduate
self-disclosure on FacebookFacebook users who compared
were blind to the w
person interactions (e.g., hypotheses
"postingand to participanton self-esteem.
Facebook Four coders ratedmake
less self-conscious," "posting the degree of positive self-disclosure ("Overall,
on Facebook how much tha
means
have to see a friend's reaction"). positivity was expressed in the Finally, participa
status updates?"; interrater a =
pleted a 10-item measure .83) and
(a negative
= .93)self-disclosure ("Overall, how much nega-
concerning the ad
of disclosing on Facebook tivityrather
was expressed in than
the status updates?";
in-person interrater a = .85)(e.g.,
on Facebook enables meacross toeach set of updates
get more (1 = none at all , 9 = a great deal).
attention from
Table SI in the Supplemental An example ofMaterial
an update from a set available
rated as highly positive onlin
items for each measure. was, "[The poster] is lucky to have such terrific friends and is
looking forward to a great day tomorrow!" An example from a
highly negative set was, "[The poster] is upset b/c her phone
Results and discussion
got stolen :@." Coders were instructed to treat positivity and
We analyzed participants' perceptions of Facebook in negativity
a series independently. Using the same scales, three new
of regression analyses with mean-centered self-esteem codersas rated
the the degree to which each participant expressed
specificpar-
predictor. Self-esteem did not predict the degree to which types of content, namely sadness, anger, frustration,
ticipants saw Facebook as offering opportunities toanxiety,
expressfear, tiredness, boredom, illness, embarrassment, irri-
themselves (M= 4.20, SD= 1 .69), t < 1 .36. As expected, how-
tability, happiness, excitement, and gratitude.
ever, participants with lower self-esteem saw Facebook Onasthea basis of each participant's status updates, 10 new
coders
safer place to express themselves than did participants rated their liking of the participant using the items
with
higher self-esteem, ß = -0.31, t(78) = -2.89,p = .005, "How much do you like this person?" (1 = not at all , 9 = a
d= 0.65.
The former were also more likely than the latter to great
saydeal),
that "How interested would you be in spending time
Facebook offered opportunities to connect with other withpeople,
him/her?" (1 = not at all interested , 9 = extremely inter-
ested),
ß = -0.22, ¿(76) = -1.99, p = .05, d = 0.46, and to see and "Would this person be someone you'd want as a
advan-
tages - such as getting support and attention withoutfriend?"
burden- (1 = definitely not , 9 = definitely yes). The mean of
ing other people - to disclosing their thoughts and these
feelings
three items (a = .97) was our index of likability (inter-
rater
on Facebook over in person, ß = -0.30, ¿(71) = -2.96,/? a = .69).
= .011,
d = 0.70.

Results and discussion

Study 2 We conducted regression analyses with mean-centered self-
Given that people with low self-esteem see Facebook as a safe esteem as the predictor. Fifty-three participants provided two
and advantageous place to disclose their thoughts and feelings, or fewer status updates, so they were excluded from analyses
do they use Facebook to better their social lives? We hypoth- involving updates. Overall, participants reported spending an
esized that Facebook posts made by participants with low self- average of 26.43 hr per month (SD = 47.12) on Facebook and
esteem would be characterized by more negativity and less having an average of 321.05 Facebook friends (SD = 186.78).
positivity than Facebook posts made by participants with high Participants with lower self-esteem did not differ from partici-
self-esteem. We also examined whether posts made by partici- pants with higher self-esteem on these variables, ts < 1.20.
pants with low self-esteem affected other people's liking for Participants averaged 0.26 updates per day (SD = 0.41), and
them. this did not depend on their level of self-esteem, t < 1 . Thus,
although people with low self-esteem see more advantages in
using Facebook, they do not appear to use Facebook more than
Method
do people with high self-esteem.
One hundred seventy-seven undergraduate Facebook We users
next investigated whether people with high self-esteem
(60 male, 117 female; mean age = 19.95 years) completed the with low self-esteem posted different types of con-
and people
Rosenberg Self-Esteem Scale (Rosenberg, 1965). Participants
tent. Individuals with lower self-esteem expressed less posi-
were then asked, "How many hours per month do tivity,
you spendß = 0.26, ¿(119) = 2.92, p = .004, d = 0.54, and more
on Facebook?" and "How many friends do you havenegativity,
on Face- ß = - 0.3 1, ¿(119) = -3.57,/? =.001, d= 0.65, in their
book?" Next, participants were asked to log in to their
statusFace-
updates than participants with higher self-esteem did.
As Table
book accounts and to provide the 10 most recent status updates1 shows, participants with lower self-esteem also
they had posted along with the date each was posted. When more sadness, anger, frustration, anxiety, fear, and
expressed

This content downloaded from

147.46.181.112 on Mon, 31 Oct 2022 11:33:12 UTC
All use subject to https://about.jstor.org/terms
 298 Forest, Wood

Table I . would participants' Facebook friends respond

Results of differently
Re than
Predictor of
complete Emotions
strangers, such as our coders, did?
Study 2

Emotion ß F( 1 , 1 1 9) Study 3
Sadness -0.35 16.71** One might expect that friends, who care about the discloser's
Anger -0.19 4.20* well-being, would respond more positively than strangers
Frustration -0.25 8.00** would to negative disclosures. However, we suspected that
Anxiety -0.26 8.63** friends' responses would depend on whom the negative disclo-
Fear -0.24 7.13** sure came from. Friends have access to information that
Tiredness -0. 1 6 3.29* strangers do not have: knowledge of the discloser's typical
Boredom -0.08 0.73
moods and expressive style. Recent research has shown that
Illness -0.10 1.16
when people interpret another person's self-disclosure, they
Embarrassment -0.18 3.77*
take into account that person's typical degree of negativity.
Irritability -0.19 4.46* When individuals habitually express negativity, others inter-
Happiness 0.28 10.39** pret their negative statements as less indicative of real prob-
Excitement 0.29 1 0.59**
lems and respond less supportively (Forest, Kille, Wood, &
Gratitude 0.28 9.88**
Holmes, 2011). Hence, when people with high self-esteem,
< .08. *p < .05. **/j who express negativity relatively infrequently, post a negative
<.01.
update, their friends may worry that something must really be
irritability, and less wronghappiness, excitement
and offer attention and support. In contrast, negative
their posts than did posts from people who
people commonly express
with higher negativity - asself
our
What effect did the data suggest that people with lowhigh
relatively self-esteem do - should be
negativ
tivity expressed byless worrisome. Their friends may even
participants experience the
with lowpost as s
strangers who examined their
tiresome and refrain from Facebook
expressing interest in hopes of dis-
liked participants with couraging further
lower negativity. The opposite should be true forles
self-esteem
with higher self-esteem, highly positive updates: ßPeople=with0.21,
low self-esteem/(119
should
d = 0.43. To examine whether the overall valence of self- receive more interest from friends than should people with
disclosure (positivity expressed minus negativity expressed)
high self-esteem. The friends of people with low self-esteem
mediated the association between self-esteem and liking, we
may lavish attention so as to encourage more pleasant updates
ran a bias-corrected bootstrap analysis (Shrout & in the future.
Bolger,
2002) with 5,000 resamples. As predicted, valence of disclo-
In Study 3, we again collected participants' 10 most recent
sure mediated the association between self-esteem andFacebook
liking, posts and coded them for positivity, negativity, and
95% confidence interval (CI) = [.04, .17],/? = .001. Bootstrap
likability by strangers (coders). We also examined the degree
of social reward each post received from participants' Face-
analyses conducted on each of the specific emotions expressed
by participants revealed that sadness (95% CI = [.03, book.15],
friends.
p = .001), anger (95% CI = [.01, -.12],/? = .025), frustration
(95% CI = [.02, .12], p = .003), irritability (95% CI = [.01,
Method
-.09], p = .019), happiness (95% CI = [.03, .15], p = .001),
Ninety-eight
excitement (95% CI = [.01, .11],/? = .023), and gratitude (95% undergraduate Facebook users (21 male, 77
CI = [.03, .16], p = .001) each mediated the association
female; mean age = 21.18 years) completed the Rosenberg
between self-esteem and liking. Self-Esteem Scale (Rosenberg, 1965). As in Study 2, partici-
Study 2, then, yielded findings that are unfortunate pants also stated how many hours per month they spent and
for par-
how many friends they had on Facebook, and they provided
ticipants with low self-esteem: Although these participants
saw great promise in using Facebook to safely reap rewarding
their 10 most recent status updates. Updates from all partici-
social experiences (Study 1), they failed to capitalize pants
on this
were randomly ordered and given to three coders, who
opportunity. Participants with lower self-esteem made posts
rated each update for positivity and negativity as in Study 2.
The
that were more negative and less positive than those of mean of the coders' scores across each participant's set of
partici-
pants with higher self-esteem and were liked less as aupdates
result -formed indices of positive self-disclosure (interrater
a = .95)
"liked less" by strangers, that is. Our coders did not know theand negative self-disclosure (interrater a = .94). Ten
people posting Facebook updates. Because nearly separate half of coders rated their liking for each update 's writer using
Facebook friends are strangers or acquaintances rather thethan
same three items (a = .99) as in Study 2. These ratings
close or even casual friends (Forest & Wood, 2011a), werethe averaged across each participant's set of updates (inter-
effects of posts on these distant others are important. raterBut
a = .82).

This content downloaded from

147.46.181.112 on Mon, 31 Oct 2022 11:33:12 UTC
All use subject to https://about.jstor.org/terms
 Self-Esteem and Social Networking 299

When a person posts= .08

a (SE
status
= .05), ¿(49.84)update, Facebook
= 1.70 ,p = .096. The predicted Self-
can write a
click comment,
on a interaction
Esteem x Positivity "Like" button
also emerged, parameter toesti-ind
ing the post, or simplymate
not=-.11 (SE=. 05), ¿(57. 13) = -2.37,
respond. p = .021 (Fig. 1). This w
Participants
instructions on Facebook tothat
suggests report the
update positivity number
was associated of "L
with social reward
of their most recent updates received
only for participants with low and the
self-esteem: number
The more positivity
ent people who commented on
participants withthe post.
low self-esteem expressed, the more com-
ments and "Likes" their friends gave, parameter estimate = .20
(SE = .07), ¿(52.55) = 2.88, p = .006 (for participants with high
Results and discussion
self-esteem, ¿ < 1).
We conducted regression analyses with mean-centeredLMM self-analyses involving self-esteem, negativity, and their
interaction
esteem as the predictor. Twenty-five participants provided two as predictors yielded only the predicted Self-
or fewer status updates, so they were excluded fromEsteem x Negativity interaction, parameter estimate = .10
analyses
(SE = .04),
involving updates. Overall, participants reported spending an ¿(557.77) = 2.37, p = .018 (Fig. 2): Participants
average of 27.56 hr per month (SD = 3 1 .62) on Facebook and self-esteem received more social rewards for their
with high
having an average of 352.51 Facebook friends (SD = more
240.77).
negative updates than for their less negative ones, param-
etermore
Participants with lower self-esteem reported spending estimate = .18 (SE = .07), ¿(558.46) = 2.53, p = .012,
time on Facebook each month than did participants with higher
whereas participants with low self-esteem did not, ¿ < 1 .
self-esteem, ß = -0.31, ¿(86) = -3.07, p = .003, d = 0.66, but we conducted parallel LMM analyses of coders'
Finally,
did not differ in number of friends, t < 1 . As in Studyliking for each update 's writer. As predicted, participant self-
2, partici-
pants with lower self-esteem expressed less positivity, ß = not interact with update positivity or negativity to
esteem did
0.23, ¿(71) = 2.01 ,/7 = .048, d = 0.48, and more negativity in
predict liking by strangers, ¿s < 1.61. 2
their status updates, ß = -0.33, ¿(71) = -2.93, p = .005, d = the responses of strangers and of participants' Face-
In sum,
0.70, than did participants with higher self-esteem,book and friends
the suggest that people with low self-esteem are not
former were also deemed less likable by strangers, rewarded ß = 0.22, for their tendency to express negativity. The more
¿(71) = 1.94,/? = .056, d = 0.46. Once again, valence of disclo- an update contained, the less participants were liked
negativity
sure mediated the association between self-esteem and coders' by coders. Although participants with high self-esteem gar-
liking (95% CI = [.03, .19 ],p = .005). nered more attention and validation from friends the more they
Do the negative posts made by participants with low self-expressed negativity, participants with low self-esteem did
esteem elicit social rewards from their Facebook friends? To not. Indeed, friends of participants with low self-esteem
answer this question, we used linear-mixed-modeling (LMM)rewarded the latter 's posts with more validation and attention
analyses to assess friends' responses to individual updates the more positive they were, perhaps trying to encourage this
posted by people with lower and higher self-esteem as a func-
atypical behavior.
tion of each update 's positivity or negativity (we used LMM
analyses because updates were nested within and unbalanced
across participants). We expected that expressing negativity
would garner more interest from friends of participants with
high self-esteem, and expressing positivity would elicit more
social rewards from friends of participants with low self-
esteem. We did not expect such an interaction on coders' liking
because coders lacked information about the discloser's
self-esteem and typical expressivity. In each analysis, the pre-
dictors were mean-centered self-esteem, person-mean-cen-
tered update positivity or negativity, and the Self-Esteem x
Positivity or Self-Esteem x Negativity interaction.
On average, each update received 0.90 (SD = 1 .46) "Likes"
and comments from 1.30 different commenters (SD = 2.04).
We summed the number of "Likes" each update received and
the number of commenters who responded to it. The resulting
"social reward" composite reflected the degree to which par-
ticipants' Facebook friends rewarded each update with valida-
Fig. I. Results from Study 3: social-reward total score as a function of self-
tion and attention.1 esteem and coder-rated positivity of Facebook status updates. Social-reward
scores
We first conducted LMM analyses using self-esteem, posi-were calculated by summing the number of "Likes" each update
received on Facebook and the number of commenters who responded to
tivity, and their interaction as predictors. A marginal main effect
each update. For both self-esteem and positivity, tow refers to the value I
of update positivity indicated that more positive updates
standard deviation below the mean, and high refers to the value I standard
received more social rewards from friends, parameter estimate
deviation above the mean.

This content downloaded from

147.46.181.112 on Mon, 31 Oct 2022 11:33:12 UTC
All use subject to https://about.jstor.org/terms
 300 Forest, Wood

a manner that may push other people away. It is ironic that

feeling safe enough to disclose their feelings on Facebook may
encourage people with low self-esteem to reveal things that
could lead to the very rejection they fear. The usual guarded,
self-protective orientation of people with low self-esteem is
generally thought to hamper their ability to cultivate satisfying
relationships, but in the studies reported here, more self-
protectiveness might have led them to minimize their negativ-
ity and brought about better interpersonal outcomes.
Do the negative disclosures of people with low self-esteem
always make them dislikable, or are such disclosures particu-
larly objectionable on Facebook? We know of no research that
has found a direct association between the degree to which
typical (i.e., nondepressed) individuals express negativity and
those individuals' likability. Moreover, negative disclosure
may, at times, benefit relationships. Disclosing negative per-
sonal information can signal that the discloser trusts the person
he or she is communicating with and desires connection
Fig. 2. Results from Study
(Graham et al., 2008). However, expressing negativity on
esteem and coder-rated n
scores Facebook maycalculated
were lack a key relationship-promoting quality: fos- b
received tering intimacy
on Facebook (Altman & Taylor, 1973). If a person makes aan
each update. For
negative disclosure both
to hundreds of other people over Face- sel
standard deviation below
deviation above the mean. book, any one of the discloser's friends is unlikely to see the
disclosure as a sign of trust or intimacy seeking with him or her
in particular. Moreover, expressing negativity may lose its rela-
General Discussion tionship-boosting benefits when it is "constant or indiscrimi-
nate" (Graham et al., 2008, p. 395). Although further research
The rapid growth of social networking Web sitesisin needed on the effects of expressing negativity outside of
recent
years has created a fantastic opportunity for relationship
Facebook, we suspect that frequently expressing negativity is
development by enabling people to connect with detrimental
others byto people's relationships in many contexts.
What, then,
sharing their thoughts and experiences. As popular media out- should people with low self-esteem do to use
social networking
lets have speculated, this opportunity could be especially valu- more constructively? Discouraging people
able for people such as those with low self-esteem,
with lowwhose
self-esteem from ever expressing negativity on Face-
self-protectiveness normally makes them hesitant booktoseems
self-unwise. Their negative Facebook disclosures may
disclose, and whose relationships could use a boost. provide them with nonsocial benefits (e.g., improved physical
The results of the three studies reported here, however, and mental sug-health; Pennebaker & Chung, 2007). And we do
gest that the way in which people with low self-esteem not advocate use being inauthentic. However, given that more
Facebook may prevent them from reaping its potential positive social
and less negative updates are better liked by strangers
benefits. We found that people with low self-esteem thanperceive
are less positive, more negative updates (Studies 2 and 3)
Facebook as a safe, appealing place for self-disclosure and that people and with low self-esteem receive better responses
that they spend as much (or more) time using Facebook from Facebook
as do friends for more positive updates than for less
people with high self-esteem. The comfort that people with (Study 3), people with low self-esteem might
positive updates
low self-esteem feel in this low-risk environment could make benefit from making more positive and less negative updates.
Rather than posting phony positive updates, however, people
it a great place for them to enrich their relationships by sharing
things they otherwise would not. However, people with low with low self-esteem might try to share more of the positive
self-esteem tend to make updates that are higher in negativitythings that do happen to them and to be selective about what
and lower in positivity than those of people with high self- negative things they post. Perhaps, then, Facebook really
esteem, and they are liked less than people with high self- could be a tool that not only makes the world "open and con-
esteem as a result. nected" (Facebook, 201 la), but also one that helps people with
The finding that people with low self-esteem express them- low self-esteem create rewarding social relationships.
selves in ways that are not especially likable is unfortunate
because people spend time with only 24% of their FacebookAcknowledgments
friends in face-to-face interactions (Forest & Wood, 2011a).We thank Richard Eibach, John Holmes, and Danu Stinson for their
Facebook contact may be the only way that many friends com- valuable comments on this manuscript, and Erik Woody for his sta-
municate, then, yet people with low self-esteem use the site tistical
in advice.

This content downloaded from

147.46.181.112 on Mon, 31 Oct 2022 11:33:12 UTC
All use subject to https://about.jstor.org/terms
 Self-Esteem and Social Networking 30 1

Declaration of Conflicting
Collins, N. L.,Interests
& Miller, L. C. (1994). Self-disclosure and liking: A
meta-analytic review. Psychological Bulletin, 116, 451-415.
The authors declared
they had no conflicts of intere that
Coyne, J. C. (1976). Depression and the responses of others. Journal
respect to their authorship or the publication of this article.
of Abnormal Psychology, 85, 186-193.
Dailey, K. (2009, June). Friends with benefits: Do Facebook friends
Funding
provide the same support as those in real life? Newsweek.
This research was funded by a grant from the Social Sciences and Retrieved from http://www.thedailybeast.com/newsweek/blogs/
Humanities Research Council of Canada (SSHRC) to Joanne V. Wood. the-human-condition/2009/06/ 1 5/friends-with-benefits-do-

facebook-friends-provide-the-same-support-as-those-in-real-life
Supplemental Material .html

Additional supporting information may be found at http://pss.sagepub Facebook. (2011a). [Facebook Info page]. Retrieved from http://
.com/content/by/supplemental-data www.facebook.com/facebook?v=info

Facebook. (2011b). Statistics. Retrieved from http ://www. facebook

Notes
.com/press/info.php?statistics
Facebook
1. We consider comments as rewards for two reasons. First, a big hit with narcissists: Study. (2010, September). The
because
Globe and Mail. Retrieved from http://www.theglobeandmail
responding to Facebook updates is voluntary, commenting suggests
that the commenter cares enough to post a response. Second,.com/news/technology/personal-tech/facebook-a-big-hit-with-
partici-
pants' ratings of how much they liked the first comment they narcissists-study/article
received 1 698694/
Forest,
for each update (1 = not at all , 9 = very much) fell above the A. L., Kille, D. R., Wood, J. V., & Holmes, J. G. (201 1). Does
midpoint
in 91.2% of cases, and only 3.4% of first comments were likedthe squeaky
"not at wheel get the grease? Ones emotional baseline
affects others ' responses to one s emotional expressions. Manu-
all" (M = 6.97, SD = 2.04); this finding suggests that these comments
were generally well received. script in preparation.
2. Although we were interested primarily in self-esteem's role
Forest, A. in
L., & Wood, J. V. (201 la). [Closeness of Facebook friends].
moderating the effects of positivity and negativity on socialUnpublished
reward, raw data.
Forest,
we also tested mediation (the possibility that self-esteem A. L., & Wood, J. V. (2011b). Self protection versus self-
yielded
optimization:
positivity or negativity, which in turn yielded social reward) using Managing conflicting self-esteem-related motives
LMM analyses with grand-mean-centered positivity or negativity
in interpersonal relationships. Manuscript in preparation.
and the joint significance approach (MacKinnon, Lockwood,
Gaucher, D., Wood, J. V., Stinson, D. A., Holmes, J. G., Logel, C.,
& Forest,
Hoffman, West, & Sheets, 2002). There was no direct effect of self- A. L. (2012). Self-esteem differences in expressivity:
esteem on social reward, but there was an indirect effect How
of self-
feeling secure in a partner's regard promotes expression.
Manuscript
esteem on social reward via positivity (but not via negativity): Lower in preparation.
self-esteem predicted expressing less positivity, parameterGotlib, I. H.,
estimate = & Robinson, L. A. (1982). Responses to depressed indi-
viduals:
.29 (SE = .14), ¿(69.65) = 2.11, /7 = .039; in turn, less positivity wasDiscrepancies between self-report and observer-rated
behavior.
associated with receiving less social reward, parameter estimate = .11 Journal of Abnormal Psychology, 91, 231-240.
(SE = .05), ¿(49.29) = 2.06, p = .045. Graham, S. M., Huang, J. Y., Clark, M. S., & Helgeson, V. S. (2008).
The positives of negative emotions: Willingness to express nega-
References
tive emotions promotes relationships. Personality and Social
Psychology Bulletin , 34, 394-406.
Altman, I., & Taylor, D. A. (1973). Social penetration : The develop-
Heimpel,
ment of interpersonal relationships. New York, NY: Holt, S. A., Wood, J. V., Marshall, M. A., & Brown, J. D. (2002).
Rine-
hart, and Winston. Do people with low self-esteem really want to feel better? Self-
Anthony, D. В., Wood, J. V., & Holmes, J. G. (2007). Testing soci-differences in motivation to repair negative moods. Jour-
esteem
nal of
ometer theory: Self-esteem and the importance of acceptance forPersonality and Social Psychology, 82, 128-147.
social decision-making. Journal of Experimental SocialLeary, M. R., & MacDonald, G. (2003). Individual differences in
Psychol-
ogy, 43 , 425^132. self-esteem: A review and theoretical integration. In M. R. Leary
Baumeister, R. F., & Leary, M. R. (1995). The need to belong:& Desire
J. R Tangney (Eds.), Handbook of self and identity (pp. 401-
418). New York, NY: Guilford Press.
for interpersonal attachments as a fundamental human motiva-
tion. Psychological Bulletin , 117, 497-529. Leary, M. R., Tambor, E. S., Terdal, S. К., & Downs, D. L. (1995).
Self-esteem as an interpersonal monitor: The sociometer
Baumeister, R. F., Tice, D. M., & Hutton, D. G. (1989). Self-presen-
hypothesis. Journal of Personality and Social Psychology, 68,
tational motivations and personality differences in self-esteem.
Journal of Personality, 57, 547-579. 518-530.

MacKinnon,
Bell, R A. (1978). Affective state, attraction and affiliation: Misery D. R, Lockwood, С. M., Hoffman, J. M., West, S. G.,
loves happy company too. Personality and Social Psychology
& Sheets, V. (2002). A comparison of methods to test mediation
Bulletin, 4, 616-619. and other intervening variable effects. Psychological Methods, 7,
83-104.
Clark, M. S., & Taraban, С. (1991). Reactions to and willingness to
Pennebaker,
express emotions in communal and exchange relationships. Jour- J. W., & Chung, С. K. (2007). Expressive writing, emo-
nal of Experimental Social Psychology, 27, 324-336. tional upheavals, and health. In H. S. Friedman & R. C. Silver

This content downloaded from

147.46.181.112 on Mon, 31 Oct 2022 11:33:12 UTC
All use subject to https://about.jstor.org/terms
302 Forest, Wood

(Eds.), Foundations
Sprecher, o
S.
York, NY: relationship
Oxford Uni
Reis, H. T., & Shaver,
acteristics o
cess. In S. 23,
Duck857-877
(Ed.),
367-389). Uchino, B.
Chichester, N
Rosenberg, M. (1965).
relationship
ton, NJ: A review
Princeton wi
Un
Rosenwald, M.
tions S.
for(201
he
shyness? Wood,
The J. V
Washin
ing pain in
.washingtonpost.com/l
20 1 1 /02/ Sedikides
1 (E
2/ABxotpQ
Segrin, C., protection
& Abramso
sive Wood,
behaviors: A J.
comV
Abnormal versus dam
Psychology
Shrout, P. E.,
tive& Bolge
affect.
566-580.
nonexperimental studi
Psychological
Wood, J. V., Hogle, A. L., &Methods
McClellan, J. C. D. (2009). Self-esteem,
Sommers, effects
S. on relationships.
(1984). In H. Reis & S. Sprecher (Eds.),
Rep Ency-
ality clopedia of humancollege
among relationships (pp. 1422-1425). Thousand
st
Oaks, CA: Sage.
Psychology , 46 , 207-

This content downloaded from

147.46.181.112 on Mon, 31 Oct 2022 11:33:12 UTC
All use subject to https://about.jstor.org/terms