CYBERPSYCHOLOGY, BEHAVIOR, AND SOCIAL NETWORKING

Volume 19, Number 11, 2016

ª Mary Ann Liebert, Inc.
DOI: 10.1089/cyber.2016.0355

Following Your ‘‘Friend’’:

Social Media and the Strength of Adolescents’
Parasocial Relationships with Media Personae

Bradley J. Bond, PhD

Abstract

Social media have permeated the lives of adolescents and may be altering the way that teens engage with their
favorite media celebrities and characters. This study surveyed 316 adolescents to examine the relationship
between social media surveillance and the strength of adolescents’ parasocial relationships (PSRs) with media
personae they follow on social media. Results indicated a significant positive relationship between exposure to
media personae on Twitter and strength of PSRs. Adolescents who had experienced social interactions with
their favorite media personae on Twitter in the form of retweets or responses to tweets had stronger PSRs than
adolescents who had no such interactions. The realism of the media personae did not moderate findings as
expected. The findings suggest that Twitter provides intimate glimpses into the personal lives of media personae
that increase audiences’ feelings of connectedness to those celebrities and characters, an important finding given
that teens are more likely to learn from media personae with whom they have PSRs.

Keywords: adolescence, parasocial relationship, social media, Twitter

Introduction personae. Individual one-sided interactions between audi-

S ocial media platforms have become nearly ubiquitous
among adolescent users. The average United States teen
spends at least 1 hour a day engaged with social media.1
Almost 90 percent of United States teens use social media
and 24 percent report being engaged almost constantly.2
Research suggests that social media can influence adoles-
cents’ identity development, access to information, and
relationships.3 Social media may not only impact real-life
relationships, but may also alter teens’ parasocial rela-
tionships (PSRs)—unilateral, intimate relationships that
audience members develop with media characters and celeb-
rities. The purpose of this study is to explore how social media
platforms like Facebook, Instagram, Twitter, and SnapChat
contribute to the strength of adolescents’ PSRs with media
personae (i.e., characters and celebrities).
Parasocial relationships
PSRs are mediated interpersonal relationships between
audiences and media personae.4 PSRs may arise from our
instinctive need for social relationships; audiences may
gratify their need for attachment to others not only through
their real-life interactions, but also by interacting with media
ences and media personae are referred to as parasocial in-
teractions.5 Parasocial interactions work in similar ways to
real-life interactions; more frequent and intimate interactions
reduce uncertainty and increase feelings of connectedness.4
PSRs are similar to face-to-face friendships in their de-
velopment, maintenance, and function.6,7 Unlike face-to-face
friendships, however, PSRs are one-sided. On television, for
example, characters’ actions are observable to viewers, but
viewers’ actions are not observable to characters. PSRs are,
therefore, characterized by their one-sided sentimentality.
PSRs may be particularly noteworthy for adolescents.
Young people are more likely to learn from media per-
sonae whom they perceive as trusted friends.5 In one study,
adolescents who reported strong PSRs with the cast of Jersey
Shore had sexual attitudes that mirrored those of the cast
more so than adolescents who did not have PSRs with cast
members.8 If teens are more likely to learn from media
personae with whom they have PSRs, scholars must better
understand the underlying factors that influence the creation
and maintenance of PSRs. Social media are likely altering
the way that adolescents experience PSRs because they
provide media personae with novel platforms to reveal tidbits
of personal information to young audiences.

Department of Communication Studies, University of San Diego, San Diego, California.

656
SOCIAL MEDIA AND TEENS’ PARASOCIAL RELATIONSHIPS
Social media
When media personae post to social media, they provide
audiences with intimate, behind-the-scene peeks into their
everyday lives and taste preferences. The ability of media
personae to carefully craft an image and create a sense of
closeness and familiarity between themselves and their fol-
lowers on social media is referred to as performative inti-
macy.9 For celebrities, performative intimacy could increase
branding and marketing opportunities.10 For audiences,
performative intimacy likely strengthens PSRs.
Let us assume that a teenage girl in 1986 experienced a PSR
with pop musician Cyndi Lauper. In line with the conceptu-
alization of PSRs, the teenage girl would have built her emo-
tional connection to Lauper through parasocial interactions.
However, in 1986 the media landscape would have limited the
teen’s parasocial interactions to watching Lauper on television,
reading about her in print, and listening to her music. Though
the teen may have learned intimate details about Lauper’s
personal life from television appearances or magazine inter-
views, such appearances would have been relatively infre-
quent. Fast-forward 30 years. Today’s teenage girl has a PSR
with Katy Perry. The teen watches Perry on television, reads
about her in print, and listens to her music. She, however, also
has access to Perry’s posts on social media. These posts can
provide quick glimpses into the everyday occurrences of
Perry and, in turn, create or reinforce a sense of privilege
to Perry’s personal backstage.9,11 In just 1 week, Perry posted
to Twitter twelve times, disclosing her support for a presi-
dential candidate, love of tacos, wardrobe difficulties at a
music festival, and excitement for cherry trees to blossom.
Perry’s taste preferences in politics, food, clothing, and botany
were self-disclosed in a matter of days on social media. When
adolescents follow their favorite media personae on social
media, they increase the frequency and intimacy of their
parasocial interactions, reduce uncertainty about their PSRs,
and likely strengthen PSRs.
H1: A positive relationship will exist between PSR strength
and exposure to favorite media personae on social media.
A key element to PSRs is realism. For individuals to de-
velop PSRs, they need to be able to make person judgments
much the same way individuals do when developing real-life
relationships.12,13 Users must perceive social media posts by
media personae as authentic for the information garnered
from social media to strengthen feelings of connectedness.13
Giles12 argued that individuals are more likely to develop
strong PSRs with nonfictional celebrities than with fictional
characters given the importance of realism. Realism may
then moderate the predicted relationship between PSR
strength and social media.
H2: The predicted relationship between PSR strength and
exposure to favorite media personae on social media will be
stronger for participants who reported PSRs with nonfic-
tional personae than for participants who reported PSRs
with fictional personae.
Though PSRs are characterized by their one-sidedness,
social media permit two-way interactions between media
personae and their followers. Media personae can repost and
respond to their followers in unique and novel ways not af-
forded by more traditional media.14 Though such interactions
657
are rare,11 media personae can mention their followers to in-
crease feelings of connectedness, give back to loyal fans, and
manage their image.9 If parasocial interactions lead to PSRs, a
repost or response that briefly transforms the parasocial into
social will likely have an even stronger influence on adoles-
cents’ perceptions of their PSRs with media personae.
H3: Participants who have engaged in computer-mediated
interactions with their favorite media personae will have
stronger PSRs than participants who have not.
Method
Sample
Vice principals from six randomly selected southern Cali-
fornia middle and high schools were invited to recruit their
students to participate in the study. Vice principals reviewed
the questionnaire before providing written consent for partic-
ipation. A university institutional review board approved re-
placing parents’ consent with school administrators’ consent
because the study presented minimal risk to participants. School
administrators who provided consent were instructed to send
recruitment email messages to their students. The recruitment
email included a hyperlink to an assent form. Once participants
provided assent, they were directed to an online survey. The
adolescents in the sample (N = 316) ranged in age from 13 to 17
years (M age = 15.64, SD = 1.19). More females (74 percent)
participated in the study than males (26 percent). The sample
was composed primarily of participants identifying as White (73
percent), mixed race (9 percent), Latino/Hispanic (8 percent),
and Asian (7 percent). Few participants identified as Black (1
percent) or another racial identity (2 percent).
Measures
The first item on the online questionnaire asked participants
to report their favorite media persona, a common practice in
the data collection procedures of studies examining PSRs
based on the assumption that individuals will most likely
have PSRs with media personae that they consider to be their
favorite.15 Participants were instructed to name anyone they
would consider their favorite regardless of the reality status,
nature, or role of the persona. However, participants were
asked to list only individual names rather than groups (e.g.,
the entire cast of a television show or a musical group). Two
coders blind to the purpose of the study then coded each of
the favorite media personae to determine whether they were
fictional characters or nonfictional celebrities using infor-
mation garnered through independent internet searches of
the media personae (j = 1.00). Following the favorite media
persona item, participants responded to items measuring
PSR strength, exposure to favorite media personae in tradi-
tional and social media, and demographics. Age and race
were presented as open-ended questions; gender was mea-
sured using a single close-ended item. Following the favorite
persona item, all other measures were presented to partici-
pants in random order to control for potential order effects.
PSR strength. The 15-item parasocial interaction scale
quantifies the strength of one-way emotional connections that
television audiences experience with television characters.16
The measure was altered for the purpose of this study to
658
account for media personae not on television and to allow the
questionnaire to electronically pipe the names of the partic-
ipants’ favorite media personae into each item. For example,
the original item, ‘‘I look forward to watching this character
when his/her show is on’’ was altered to read, ‘‘I look forward
to watching, listening to, or reading about [name of media
persona].’’ Other examples include, ‘‘[Name of media per-
sona] would fit in well with my group of friends’’ and ‘‘[Name
of media persona] makes me feel comfortable, like I am with a
friend.’’ Responses to the items were measured on a five-point
scale from 1 (strongly disagree) to 5 (strongly agree). Mean
scores were calculated for the 15 items (a = 0.86) and ranged
from 2.20 to 5.00 (M = 4.10, SD = 0.57).
Exposure to favorite media personae. Participants re-
ported how often they were exposed to their favorite media
persona ‘‘on television, in film, in magazines, or on the radio’’
(i.e., traditional screen media) as one item on a scale from 1
(never) to 7 (at least once/day) (M = 4.05, SD = 1.95). The
same anchor points were then used to measure how often
participants checked the posts of their favorite media persona
on Facebook (M = 2.76, SD = 2.29), Instagram (M = 3.37,
SD = 2.56), SnapChat (M = 1.95, SD = 1.88), and Twitter
(M = 2.68, SD = 2.42). These four social media platforms were
included on the questionnaire because current national survey
data suggest that they are the most popular among adoles-
cents.2 Participants were also asked how many times their
favorite media persona had ever reposted or replied to any of
their posts on Facebook, Instagram, SnapChat, or Twitter.
Results
Descriptive analyses
Favorite media personae. Participants reported 201 dif-
ferent media personae as favorites. The most frequently cited
favorites were Jennifer Lawrence (11 percent), Taylor Swift
(3 percent), Harry Styles (2 percent), Sheldon Cooper (2
percent), and Miley Cyrus (2 percent). Only 28 percent of
participants reported favorite media personae that were among
the top 10 most cited, indicating considerable variance among
adolescents’ favorite media personae. Media personae were
more likely nonfictional celebrities (60 percent) than fictional
characters (40 percent).
Media use. Most adolescents (70 percent) reported fol-
lowing their favorite media persona on at least one social media
platform, and half (50 percent) reported following their favor-
ite on at least two different social media platforms. A
smaller percent (15 percent) noted that they use all four
social media platforms included in the study to follow their
favorite media persona. Instagram was the most popular for
following media personae (55 percent), followed by Face-
book (46 percent), Twitter (36 percent), and SnapChat (26
percent).
Hypotheses and research questions
The first hypothesis predicted a relationship between PSR
strength and frequency of exposure to favorite media personae
via social media. The second hypothesis predicted that this
relationship would be moderated by the realism of the media
personae. To test these hypotheses, a linear regression analysis
BOND
Table 1. Summary of Regression Analysis
Predicting Parasocial Relationship Strength
B SE b
Sex 0.32 0.07 0.26***
Age 0.03 0.03 0.07
Race 0.09 0.07 0.07
Realism -0.02 0.07 -0.01
Traditional Media 0.01 0.02 0.02
Facebook 0.02 0.02 0.06
Instagram -0.01 0.02 -0.05
Twitter 0.05 0.02 0.22**
SnapChat 0.06 0.05 0.07
Note. Sex (male/female), race (White/racial minority), and
realism (fictional/non-fictional) were dichotomously coded. Inter-
action terms were nonsignificant and thus removed from analysis to
preserve power.
F (9, 289) = 5.68, p < 0.001. R2 = 0.15. **p < 0.01, ***p < 0.001.
was conducted using age, sex, race, media personae realism,
traditional media exposure, exposure to the four social media
platforms, and interaction terms as predictor variables and
strength of PSR as the criterion variable. All variables were
mean-centered before analysis to prevent issues related to
multicollinearity. The resulting model was significant, F (9,
289) = 5.68, p < 0.001. Participant sex was related to PSR
strength (b = 0.26, p < 0.001) such that females had stronger
PSRs than males. Twitter (b = 0.22, p < 0.01) significantly
contributed to PSR strength. Facebook (b = 0.06), Instagram
(b = -0.05), SnapChat (b = 0.07), traditional media (b = 0.02),
and realism (b = -0.01) failed to contribute to PSR strength
(Table 1). Hypothesis 1 was partially supported; exposure to
favorite media personae on Twitter was significantly related
to PSR strength. Interaction terms were not significant con-
tributors to PSR strength. Thus, realism of the media personae
failed to moderate the relationship between social media and
PSR strength. Hypothesis 2 was not supported.
The third hypothesis predicted that participants who had
engaged in computer-mediated interactions with favorite
media personae would have stronger PSRs than participants
who had not engaged in such interactions. Only 3 percent of
the sample (n = 9) reported engaging in computer-mediated
interactions with their favorite media persona. All of these
interactions came in the form of a retweet or response tweet
on Twitter. No participant reported more than one such in-
teraction. Participants who had interacted with their favorite
media personae on Twitter had stronger PSRs (M = 4.50,
SD = 0.40) than participants who had no such interactions
(M = 4.08, SD = 0.56), t (312) = 2.23, p < 0.05. Hypothesis 3
was supported.
Discussion
The purpose of this study was to examine how exposure to
media personae on social media was related to PSR strength
among adolescents. A positive relationship was found be-
tween surveilling favorite media personae on Twitter and
strength of PSRs that was not significant for other social
media platforms. Though more adolescents reported fol-
lowing favorite media personae on Instagram and Facebook
than Twitter, the relationship between Twitter and PSRs may
be explained by the nature and function of the social media
SOCIAL MEDIA AND TEENS’ PARASOCIAL RELATIONSHIPS
platform. Twitter has emerged as a legitimate channel of
celebrity voice. Tweets from celebrities and politicians are
often cited as official statements by news media, a privilege
not often afforded to other social media.17 The improvisatory
quality of tweets or the inability to edit posted tweets may
contribute to their perceived legitimacy. Not only are tweets
deemed authentic, but they are also often intimate. A qual-
itative analysis of 12 media personae Twitter accounts sug-
gested that celebrities disclose personal information on
Twitter and that such disclosure is meaningful and poten-
tially influential for those who consume the messages.11 The
legitimacy and intimacy of tweets may increase the impor-
tance of parasocial interactions on Twitter compared to other
social media and, in turn, influence the strength of PSRs
between teen audiences and their favorite media personae.
Frequency may also explain the contribution of Twitter to
PSR strength. Media personae may post more frequently on
Twitter than other social media. The casual observation of
Katy Perry’s Twitter account noted earlier found that Perry
had tweeted 12 times in 1 week. In the same week, she had
posted to Instagram just once. Perry’s decision to tweet
considerably more than to post on Instagram may be because
Twitter followers are more than twice as likely than other
social media users to repost content on Twitter and other
social media.18 Reposts from everyday people are the life-
blood of building and maintaining fan loyalty on social
media.10 If media personae are posting more frequently on
Twitter, teen audiences may engage in more parasocial in-
teractions on Twitter and, thus, strengthen PSRs.
The relationship between Twitter surveillance and PSR
strength was not moderated by the realism of the media
personae as hypothesized. Twitter may give life to fictional
characters in ways that traditional media never could. Vol-
demort, the Harry Potter villain, has never been interviewed
on The Tonight Show or featured in GQ. Voldemort, how-
ever, is active on Twitter and is followed by over 2 million
people. The account has no affiliation with those who own
rights to Voldemort’s likeness. Though followers are aware
that a fictional character cannot tweet, Voldemort’s tweets
serve as parasocial interaction between him and his followers,
allowing fans to suspend reality in novel ways beyond the
books and films and strengthen their PSRs with the villain.
Adolescents who had experienced computer-mediated
communication with media personae on Twitter via a retweet
or response to a tweet had stronger PSRs than those who had
not experienced such communication. Such a finding rein-
forces the argument that a one-time social interaction can
intensify an otherwise PSR.9,11 Retweets or response tweets
strengthen emotional connections, and emotional connec-
tions characterize PSRs. Computer-mediated communica-
tion with media personae may even have a greater impact on
perceived intimacy than more traditional public appearances
where audiences might engage in face-to-face interactions
with their favorite celebrities like autograph signings. Au-
tograph signings are scheduled and highly structured; social
media posts are perceived as spontaneous and genuine.9 Such
a finding could have implications for media personae man-
aging performative intimacy as a means of branding and
marketing. Social media platforms like Twitter may be safe
spaces from which to connect with fans and increase PSR
strength, all while avoiding the risks associated with real-life
interactions with fans.11
659
This study is not without limitations. The cross-sectional
nature of the data does not allow causal claims. Teens with
strong PSRs could subsequently follow their favorite media
personae on Twitter. The relationship is likely mutually causal,
but longitudinal and experimental research would provide
additional insight into causation. A larger, more diversified
sample would be beneficial given the lack of males and the
small number of participants who had experienced computer-
mediated communication with their favorite media personae.
Additionally, the content of tweets from media personae was
not examined here. A content analysis of celebrity tweets that
teens are exposed to would also add to our understanding of
how performative intimacy takes shape and its relationship to
PSRs. Another area for future research lies in the social media
literacy of adolescents. For example, assistants often author
celebrity Twitter accounts rather than the celebrities them-
selves, a practice known as ghost-tweeting.13 Ghost-tweeting
could alter the perceived authenticity and, in turn, influence
PSRs.13 Research has also suggested that reliance on social
media to follow media personae may be related to lower
psychological well-being.19 Future research should continue
to examine how adolescents interpret and understand para-
social interactions on social media and the effects of PSRs.
In sum, the findings presented here suggest that Twitter
provides adolescents intimate glimpses into the backstage,
personal lives of media personae. Exposure to the backstage
increases teen audiences’ feelings of connectedness to the
characters and celebrities they follow and strengthens their
PSRs. Given that research suggests teens are more likely to
learn from media personae with whom they have PSRs, so-
cial media may be altering the function of media personae in
creating, reinforcing, or changing adolescents’ beliefs, atti-
tudes, and behaviors.
Acknowledgments
The author wishes to thank Kat Pfost and Hayley West for
their assistance with data collection and analysis.
Author Disclosure Statement
No competing financial interests exist.
References
1. Rideout V. (2015) The common sense census: media use by
tweens and teens. San Francisco, CA: Common Sense
Media.
2. Lenhart A. (2015) Teens, social media, and technology
overview 2015. Washington, DC: Pew Research Center.
3. Moreno M, Kota R. (2014) Social media. In Strasburger
VC, Wilson BJ, Jordan AB, eds. Children, adolescents, and
the media, 3rd ed. Thousand Oaks, CA: Sage, pp. 435–456.
4. Horton D, Wohl RR. Mass communication and para-social
interaction: observations on intimacy at a distance. Psy-
chiatry 1956; 19:215–229.
5. Calvert SL, Richards MN. (2014) Children’s parasocial
relationships. In Jordan AB, Romer D, eds. Media and the
well-being of children and adolescents. New York: Oxford,
pp. 187–200.
6. Rubin AM, Perse EM. Audience activity and soap opera in-
volvement: a uses and effects investigation. Human Com-
munication Research 1987; 14:246–268.
660
7. Rubin RB, McHugh MP. Development of parasocial in-
teraction relationships. Journal of Broadcasting and Elec-
tronic Media 1987; 31:279–292.
8. Bond BJ, Drogos KL. Sex on the shore: wishful identifi-
cation and parasocial relationships as mediators in the re-
lationship between Jersey Shore exposure and emerging
adults’ sexual attitudes and behaviors. Media Psychology
2014; 17:102–126.
9. Marwick A, Boyd D. To see and be seen: celebrity practice
on Twitter. Convergence 2011; 17:139–158.
10. Lueck JA. Friend-zone with benefits: the parasocial ad-
vertising of Kim Kardashian. Journal of Marketing Com-
munications 2015; 21:91–109.
11. Stever GS, Lawson K. Twitter as a way for celebrities to
communicate with fans: implications for the study of para-
social interaction. North American Journal of Psychology
2013; 15:339–354.
12. Giles DC. Parasocial interaction: a review of literature and a
model for future research. Media Psychology 2002; 4:279–305.
13. Cohen EL, Tyler WJ. Examining perceived distance and
personal authenticity as mediators of the effects of ghost-
tweeting on parasocial interaction. Cyberpsychology, Be-
havior, and Social Networking 2016; 19:342–346.
14. Tsiotsou RH. The role of social and parasocial relationships
on social networking sites loyalty. Computers in Human
Behavior 2015; 48:401–414.
15. Tsay M, Bodine BM. Exploring parasocial interaction in
college students as a multidimensional construct: do per-
BOND
sonality, interpersonal need, and television motive predict
their relationships with media characters? Psychology of
Popular Media Culture 2012; 1:185–200.
16. Rosaen SF, Dibble JF. Investigating the relationships
among child’s age, parasocial interactions, and the social
realism of favorite television characters. Communication
Research Reports 2008; 25:145–154.
17. Muntean N, Petersen AH. Celebrity Twitter: strategies of
intrusion and disclosure in the age of technoculture. M/C
Journal 2009;12: http://journal.media-culture.org.au/index
.php/mcjournal/article/view/194 Last accessed August 25,
2016.
18. Zarrella D. (2010) The social media marketing book. Se-
bastopol, CA: O’Reilly.
19. Baek YM, Bae Y, Jang H. Social and parasocial relationships
on social network sites and their differential relationships with
users’ psychological well-being. Cyberpsychology, Behavior,
and Social Networking 2013; 16:512–517.
Address correspondence to:
Dr. Bradley J. Bond
Department of Communication Studies
University of San Diego
5998 Alcala Park
San Diego, CA 92110
E-mail: bond@sandiego.edu