HOW DO PEOPLE PURSUE MULTIPLE GOALS WHEN THEY COMMUNICATE
EVERYDAY DISTRESS AND SEEK EMOTIONAL SUPPORT
ON SOCIAL NETWORK SITES?
By
Hyun Jung Oh

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Media and Information Studies – Doctor of Philosophy
2015

PUBLIC ABSTRACT
HOW DO PEOPLE PURSUE MULTIPLE GOALS WHEN THEY COMMUNICATE
EVERYDAY DISTRESS AND SEEK EMOTIONAL SUPPORT
ON SOCIAL NETWORK SITES?
By
Hyun Jung Oh
The present study is designed to understand how people share everyday stress and seek
emotional support on social network sites (SNS) such as Facebook. Although previous studies
have shown that people receive support from their friends through social network sites, it
remains unknown exactly how people share stressful events on social network sites to receive
necessary comfort. To answer this question, this study looks at two different goals people pursue
when they share stressful events on social network sites—receiving necessary support and
maintaining a good impression—and examines how people produce support-seeking messages to
accomplish those goals.
The present study conducted an experiment by assigning 176 college student
participants to one of four experimental conditions: (1) sharing a moderately stressful event
publicly by posting it on a Facebook status update; (2) sharing a mildly stressful event publicly
by posting it on a Facebook status update; (3) sharing a moderately stressful event privately by
sending a private message to a Facebook friend; and (4) sharing a mildly stressful event privately
by sending a private message to a Facebook friend. Participants were then asked how important
it was for them to receive necessary support and to maintain a good impression on others. This
study also measured the amount of time participants spent writing a message and analyzed the
actual message they produced during the experiment.

The result shows that the importance people place on receiving support increases with
the severity of the stressful event. Maintaining a good impression becomes more important when
people share the stressful story using the public status update as opposed to the private message.
The amount of time people spend on writing a message increases when they share a stressful
event that is more serious, and the extent of the increase is greater when people use the public
status update as opposed to the private message to share the stressful story. Finally, the more
time people spend on writing a message, the more likely they are to produce a message that
reflects their need for comfort and impression management.
The findings from this study are meaningful, as they show how people share stressful
events and seek emotional support on social network sites. They suggest that people want to
maintain a good impression on others, and that this desire becomes greater when people share a
stressful story publicly on social network sites. People’s concerns about maintaining a good
impression influence supportive communication and shape the quality of the emotional support
exchanged on social network sites.

ABSTRACT
HOW DO PEOPLE PURSUE MULTIPLE GOALS WHEN THEY COMMUNICATE
EVERYDAY DISTRESS AND SEEK EMOTIONAL SUPPORT
ON SOCIAL NETWORK SITES?
By
Hyun Jung Oh
The present study investigates how people communicate everyday emotional distress
and seek emotional support on social network sites. Although previous studies have shown that
people receive necessary social support when they engage in online social networking, there is a
lack of research on how people produce support-seeking messages to be shared on social
network sites and how consideration of multiple goals shapes the composition of those messages.
Employing the theoretical framework of multiple goals, this study investigates whether people
pursue multiple goals when they seek emotional support on social network sites, and whether
they compose support-seeking messages to address different goals simultaneously. It further
examines whether the severity of the stressful situation and the expected publicity of the supportseeking message influence the amount of effort people put into message composition, and
whether this degree of effort predicts the production of sophisticated messages that address
multiple goals effectively.
The present study employed a two (problem severity: low vs. high) by two (publicity of
message: public vs. private) between-subjects design. A laboratory experiment with 176 college
student participants shows that people’s communication goals change depending on the publicity
of their message, and that people’s levels of effort in message composition in computer-mediated
communication (CMC) increase as the severity of the problem increases. Results also suggest
that the effect of the problem’s severity on CMC message composition time is moderated by the

publicity of the message, such that the effect of the problem severity on CMC message
composition time is greater when people use a public SNS channel as opposed to a private SNS
channel to communicate a stressful situation. Finally, the more effort people put into CMC
message composition, the more likely they are to produce a sophisticated support-seeking
message that serves multiple communication goals.
The finding from this study contributes to the theoretical advancement of social support
and supportive communication in general, and provides new insights to the body of knowledge
on how people utilize social network sites to seek emotional support. It suggests that support
seekers can pursue and address multiple goals when they seek emotional support from others,
and that the affordances of social network sites can amplify the impression management goal,
which further contributes to the dynamics of supportive communication and shapes the quality of
the emotional support exchanged on social network sites.

ACKNOWLEDGEMENTS

Writing this dissertation has been one of the most significant academic challenges I
have ever had to face. Without the support of numerous people, I would not have been able to
complete this study. To them I owe my deepest gratitude.
I would like to express my sincere gratitude to my honorific supervisor, Dr. Robert
LaRose, for his mentorship, scholarly inputs, constant support, and patience. He agreed to act as
my supervisor despite his many other commitments. With his great wisdom and knowledge, Dr.
LaRose helped me broaden my academic perspectives and better understand the true nature of
human behavior.
Many thanks also go to my wonderful committee members, Dr. Stephen Lacy, Dr.
Joseph Walther, and Dr. Saleem Alhabash. I felt truly privileged to be guided by so many of the
greatest scholars in this field. Dr. Lacy proved to be one of the most thoughtful mentors in this
program; he genuinely cares for every single student with his heart. His encouragement enabled
me to remain confident and overcome the difficulties I encountered. Dr. Walther provided
insightful advice and enlightened my way of thinking. I am especially thankful to him for staying
on my committee even after he moved to Singapore. Finally, I am indebted to Dr. Alhabash for
his consistent work in sharing his expertise in experimental research to help me enhance my
work.
I am deeply thankful to Dr. Byoungkwan Lee, my first mentor, who sparked my
academic interest in the field of communication and shaped my direction in academia. During
the last ten years, he has cheered me on in my efforts to achieve my dream; I would never have
come this far without his guidance.
iv

In addition, I would like to thank my former advisor, Dr. Hye-Jin Paek, who originally
accepted me as her student and provided me with meaningful guidance during my early Ph.D.
work at Michigan State University. Her generous support and valuable advice will never be
forgotten.
To my colleagues and the faculty at Michigan State University, who all extended their
help in very special ways, I want to thank you for the lessons I was able to glean through both
personal and scholarly interactions. Special thanks go to Jenni Lee, who administered the
laboratory sessions in Michigan while I was in Korea. My dissertation would not have come to a
successful completion without her help. Thank you Dr. Amanda Holmstrom, for your
illuminating advice and for kindly sharing your experiment stimuli for my study. I would also
like to thank Mikyeung Bae—not only for her help with data collection, but also for her
thoughtful consideration and comfort over the years in Michigan.
I would also like to show my deep appreciation to the great mentors at my home
university, Hanyang University, for their continuous support and encouragement. Dr. Hyunou
Lee and Dr. Miejeong Han helped me build the foundation of my academic knowledge and stood
by me at every stage of my academic progress.
So many friends and great people in both Korea and the United States offered me their
help and support throughout the Ph.D. process. Although they are too numerous to name, every
single one of them have my sincerest thanks. My special thanks go to Chanwoo Kim for filling
my non-academic life during the Ph.D. with joy and laughter. His humor and optimistic view of
life helped me stay physically and mentally healthy throughout this process.
Finally, I could not have succeeded in my academic work without the unconditional
support and encouragement of my parents, Mahn Seug Oh and Ogsoon Shin. They raised me,
v

trusted me, and have always reminded me how precious I am to them. My parents are the origin
of my happiness, and this dissertation is dedicated to them, for their endless love.

vi

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................................... ix
LIST OF FIGURES ........................................................................................................................ x
CHAPTER 1 ................................................................................................................................... 1
INTRODUCTION ...................................................................................................................... 1
CHAPTER 2 ................................................................................................................................... 5
LITERATURE REVIEW AND HYPOTHESES ........................................................................ 5
Stress and Emotional Support ................................................................................................. 5
Daily Stress and Well-Being ............................................................................................... 5
Emotional Support .............................................................................................................. 7
Computer-Mediated Social Support and Self-Disclosure on Social Network Sites ............... 9
Computer-Mediated Social Support on Social Network Sites ............................................ 9
Self-Disclosure on Social Network Sites .......................................................................... 11
Support Providers, Support Seekers, and Supportive Communication ................................ 13
Features of Effective Supportive Communication ............................................................ 13
Support Providers.............................................................................................................. 14
Support Seekers ................................................................................................................ 15
Multiple Goals of Seeking Emotional Support on Social Network Sites ............................. 19
Multiple Goal Framework................................................................................................. 19
Primary Goal of Support-Seeking Communication .......................................................... 22
Secondary Goal of Support-Seeking Communication on Social Network Sites .................. 24
Public vs. Private Communication on Social Network Sites ............................................ 24
Publicity of Support-Seeking Message and Pursuit of Impression Management Goal .... 25
Production of Support-Seeking CMC Message on Social Network Sites ............................ 29
Production of Support-Seeking CMC Message ................................................................ 29
Sophistication of Support-Seeking CMC Message: Message Design Logic Approach ... 31
CHAPTER 3 ................................................................................................................................. 36
METHOD ................................................................................................................................. 36
Participants ............................................................................................................................ 36
Procedure .............................................................................................................................. 38
Stimuli ............................................................................................................................... 40
Manipulation Check .......................................................................................................... 42
Order Effect ...................................................................................................................... 43
Measures ............................................................................................................................... 44
Items for Pretest and Manipulation Check........................................................................ 44
Goal Importance................................................................................................................ 44
Time Spent on Message Composition .............................................................................. 48
Level of Message Sophistication ...................................................................................... 48
vii

Analytic Strategy .................................................................................................................. 53
CHAPTER 4 ................................................................................................................................. 55
RESULTS ................................................................................................................................. 55
Effect of Problem Severity on Importance of Support-Receiving Goal ............................... 55
Effect of Message Publicity on Importance of Impression Management Goal .................... 57
Effect of Problem Severity and Message Publicity on Message Composition Time ........... 58
Effect of Message Composition Time on Sophistication of Support-Seeking CMC Message
............................................................................................................................................... 61
CHAPTER 5 ................................................................................................................................. 63
DISCUSSION ........................................................................................................................... 63
Importance of Support-Receiving and Impression Management Goals ............................... 63
Pursuit of Multiple Goals and Production of Support-Seeking CMC Message ................... 67
Sophistication of Supportive-Seeking CMC Message .......................................................... 72
Opportunities and Constraints of Supportive Communication on Social Network Sites ..... 74
Limitations ............................................................................................................................ 77
Conclusion ............................................................................................................................ 83
APPENDICES .............................................................................................................................. 84
APPENDIX A. Distribution of Original, Log-Transformed, and Square Root-Transformed
Number of Facebook Friends and Time Spent on Message Composition ........................... 85
APPENDIX B. Results of Neutral Stimuli Pretest ............................................................... 86
APPENDIX C. Emotional State after Neutralization Treatment .......................................... 87
APPENDIX D. Results of First Pretest................................................................................. 88
APPENDIX E. Results of Second Pretest............................................................................. 89
APPENDIX F. Stressful Scenarios Used in Main Experiment............................................. 90
APPENDIX G. Results of Manipulation Check ................................................................... 91
APPENDIX H. Codebook for Content-Analysis of Message Sophistication of SupportSeeking CMC Messages ....................................................................................................... 92
APPENDIX I. Consent Form, Survey, and Debriefing Form............................................... 95
BIBLIOGRAPHY ....................................................................................................................... 101

viii

LIST OF TABLES

Table 1. Participants’ Demographic Characteristics and Number of Their SNS Friends ............ 37
Table 2. Descriptive, Reliability, and Validity of Key Variables ................................................. 47
Table 3. Content Analysis of Support-Seeking CMC Messages: Presence of Goal-Oriented
Content, Level of Message Sophistication, and Inter-Coder Reliability ........................ 52
Table 4. Means and Standard Deviations of Perceived Importance of Support-Receiving Goal
between Individuals Facing Moderately vs. Mildly Severe Problems ........................... 56
Table 5. Means and Standard Deviations of Perceived Importance of Impression Management
Goal between Individuals Seeking Emotional Support through Public vs. Private SNS
Channels ......................................................................................................................... 57
Table 6. Main and Interaction Effects of Problem Severity and Message Publicity on Time Spent
on CMC Message Composition ...................................................................................... 59
Table 7. Means and Standard Deviations of Time Spent on CMC Message Composition across
Four Experimental Conditions ........................................................................................ 60
Table 8. Correlations among Key Variables ................................................................................. 61
Table 9. Ordinal Regression Analyses Examining Effect of Time Spent on CMC Message
Composition on Sophistication of Support-Seeking CMC Message .............................. 62
Table 10. Pretest of IAPS Stimuli: Means and Standard Deviations of Arousal Scores .............. 86
Table 11. Means and Standard Deviations of SAM Scores after Neutralization Treatment ........ 87
Table 12. First Pretest of Twelve Scenarios: Means and Standard Deviations of Perceived Reality
and Perceived Severity ................................................................................................... 88
Table 13. Second Pretest of Six Scenarios: Means and Standard Deviations of Perceived Reality,
Perceived Severity, and Likelihood of Seeking Support on Facebook .......................... 89
Table 14. Means and Standard Deviations of Perceived Severity and SAM Scores after Reading
Scenario .......................................................................................................................... 91
Table 15. ANOVA Results of Perceived Severity and SAM Scores after Reading the Scenario 91

ix

LIST OF FIGURES

Figure 1. Main and Interaction Effects of Problem Severity and Message Publicity on Time
Spent on CMC Message Composition .................................................................................. 60
Figure 2. Probability-Probability Plots of Number of SNS Friends and Time Spent on CMC
Message Composition ........................................................................................................... 85
Figure 3. SAM Scale – Emotional Valence .................................................................................. 96
Figure 4. SAM Scale – Emotional Arousal .................................................................................. 96

x

CHAPTER 1
INTRODUCTION

Social network sites (SNS) are communication platforms where people can form and
maintain a wide network of social connections (Ellison, Steinfield, & Lampe, 2011). Acquiring
social support is one of the key reasons for visiting social network sites (Park, Kee, &
Valenzuela, 2009). Social support helps people maintain psychological well-being directly or
through alleviating distress caused by negative events (Cobb, 1976; Cohen & Hoberman, 1983;
Cohen & Wills, 1985; House & Kahn, 1985). Studies have indicated that SNS users receive
multiple types of support from their SNS friends (Hampton, Goulet, Rainie, & Purcell, 2011),
with emotional support being the most common type (Oh, Lauckner, Boehmer, Fewins-Bliss, &
Li, 2013).
Although existing SNS studies suggest that people emotionally benefit from having
many SNS friends and engaging in various networking activities on social network sites, most of
these studies focused on the receiver’s perception of available social support or positive
outcomes of having such perception (e.g., Lee, Noh, & Koo, 2013; Nabi, Prestin, & So, 2013;
Wright, 2012; Wright et al., 2013). There has thus far not been a study examining how people
compose the messages they post on social network sites. Moreover, previous studies have
suggested that people manage their impression on social network sites by disclosing positive
aspects of themselves (e.g., Qiu, Lin, Leung, & Tov, 2012; Zhao, Grasmuck, & Martin, 2008).
However, it remains unknown how the general motivation of impression management on social
network sites shapes ways people seek emotional support on social network sites, and how

1

features of computer-mediated communication (CMC) facilitate deliberate composition of the
support-seeking messages.
A considerable body of literature on supportive communication has suggested
characteristics of providers, seekers, and situations associated with accomplishing desirable
support outcomes (e.g., Barbee, Gulley, & Cunningham, 1990; Burleson, 2003; Goldsmith, 2000,
2004; High & Dillard, 2012; Jones, 2004; MacGeorge, Gillihan, Samter, & Clark, 2003;
MacGeorge, Lichtman, & Pressey, 2002; Mortenson, 2009). Most of these studies, however,
viewed support seekers as passive recipients of supportive messages or ineffective
communicators whose support-seeking attempts are often hindered by other situational factors.
They seldom examined whether and how support seekers strategically address the need for
support in order to accomplish their support-seeking goals effectively.
Since stressful situations vary in severity or other contextual facets (e.g., cause, duration
etc.), support seekers are likely to have different goals and concerns, when they share stressful
stories with others. Since humans constantly manage the impression they make on others in order
to maintain an ideal self-image (Goffman, 1967; Leary & Kowalski, 1990; Rosenberg & Egbert,
2011; Turnley & Bolino, 2001), it is natural that people are concerned about the impression they
are making, especially when they reveal their problems to others (Albrecht & Goldsmith, 2003;
Brasher, Neidig, & Goldsmith, 2004). Moreover, because people desire to maintain a good
impression when their behavior can be seen by a large audience (Leary & Kowalski, 1990),
impression-oriented concerns would be more prominent when people seek support on social
network sites, especially when they use public channels (e.g., wall posts or status updates) as
opposed to private channels (e.g., private messages).

2

The multiple goal framework suggests that people pursue more than one goal during
interpersonal communication and that each goal differs in its importance relative to other goals
(Caughlin, 2010; Dillard, Segrin, & Harden, 1989; Meyer, 2002; Samp & Solomon, 1998; 1999;
Wilson & Feng, 2007). Employing the theoretical framework of multiple goals, the present study
explores whether people pursue multiple goals when they seek emotional support on social
network sites and whether the consideration of multiple goals shapes their support-seeking
messages. More specifically, this study tests whether the severity of the stressful situation
motivates people to strongly pursue the support-receiving goal, and whether the expected
publicity of the support-seeking message further motivates people to pursue the impression
management goal in addition to the support-receiving goal. By using an asynchronous CMC
channel that allows more time for message production and edits than do face-to-face or
synchronous channels (Walther, 2007), this study also examines whether individuals put more
time and effort into CMC message composition when they are motivated to pursue multiple
goals as opposed to when they are not motivated to do so, and whether such effort leads to the
production of sophisticated CMC messages that effectively address those goals.
The present study will contribute to the theoretical advancement of social support and
supportive communication by providing a theory-grounded explanation of support-activating
behavior on social network sites. Although existing studies on supportive communication have
examined the quality of supportive messages and the characteristics of providers and situations
associated with the quality of those messages, little is known about how people solicit support
from others (Lim, Teo, & Zhao, 2013; Taylor et al., 2004). It is often considered natural that
people will automatically seek support when they need it. An overlooked fact is that support
seekers can also frame and communicate their problem differently to receive optimal support. By
3

examining how multiple goals shape the way people present stressful situations and seek support
on social network sites, this study will contribute to a better understanding of how support
seekers can influence the quality of supportive interaction.
Findings from the present study will also contribute to the body of knowledge on how
people utilize social network sites to maintain psychological well-being. Although SNS
researchers have advocated the benefits of online social networking in promoting psychological
well-being, there have been few theoretical explanations of the unique nature of social network
sites pertaining to sharing everyday stressors and exchanging social support. A majority of SNS
studies on social support have associated the characteristics of social ties or relationships with
perception of social support and psychological well-being. This simply reflects an untested
assumption of sociological studies that “merely having these ties may have benefits” (Taylor et
al., 2004, p. 355). Exploring how people represent stressful events on social network sites and
examining how affordances of those sites highlight certain communication goals may help us
better understand the unique opportunities and constraints of social network sites in delivering
needed emotional support.

4

CHAPTER 2
LITERATURE REVIEW AND HYPOTHESES

Stress and Emotional Support
Daily Stress and Well-Being
Psychological stress is defined as “a reaction to the environment in which there is (a)
the threat of a net loss of resources, (b) the net loss of resources, or (c) a lack of resource gain
following the investment” (Hobfòll, 1989, p. 516). Reaction to stress involves “an alarming
response, eliciting a general increase in wakefulness and brain arousal, and specific response to
deal with the reasons for the alarm” (Ursin & Eriksen, 2004). Resources include objects,
personal characteristics, conditions, or energies that have instrumental or symbolic value to the
person or serve as means of obtaining those things (Hobfòll, 1989). People experience a mild
level of stress regularly due to daily hassles, difficulties, or emotionally upsetting events
(Brantley, Waggoner, Jones, & Rappaport, 1987; DeLongis, Folkman, & Lazarus, 1988; Nixon,
Mazzola, Bauer, Krueger, & Spector, 2011). People sometimes experience severe stress because
of major life events such as traumatic events or losses (Brewin, Andrew, & Valentine, 2000;
Tracy, Norris, & Galea, 2011). The accumulation of stress can negatively affect physical health
(Herbert & Cohen, 1993; Nixon et al., 2011; Segerstrom & Miller, 2004) as well as
psychological well-being (Kraaij, Arensman, & Spinhoven, 2002).
The present study is conducted in the context of mild-to-moderate hassles,
disappointments, or difficulties that people experience and become upset about on a regular
basis. Daily stressful events are negative events that have a higher probability of occurring on a
daily basis but have less significant impacts upon the individuals than negative major life events
5

do (Brantley et al., 1987; Flett, Blankstein, Hicken, & Watson, 1995). Daily stressors are “closer
in temporal proximity to the onset, exacerbation, or recurrence of physical or psychological
symptoms than major life events” (Brantley et al., 1987, p. 62). Previous research has shown that
daily upsetting events are major determinants of mood and psychological well-being (Blankstein
& Flett, 1992; DeLongis et al., 1988). Some studies further suggested that daily stressors are
even stronger predictors of psychological symptoms than major life events (DeLongis, Coyne,
Dakof, Folkman, & Lazarus, 1982; Kanner, Coyne, Schaefer, & Lazarus, 1981). This is partly
because daily hassles are less visible compared to major life events, and elicit less support from
others unless people express their feelings and actively seek support (Flett et al., 1995). This
further creates a paradox in support availability. For example, a study found that individuals
suffering from mild mental problems tended to have fewer supporters and received less
sympathy than those experiencing severer and more visible mental illnesses (Perry, 2011).
Therefore, the support-seeking behavior and ways people represent the problematic situation
would be critical for receiving necessary support, especially when people face daily upsetting
events as they are not readily observable or clearly perceived by others as stressful.
Furthermore, previous studies have shown that supportive communication or the content
of supportive messages matters only when support seekers can effectively process the message
(Bodie, Burleson, & Holmstrom et al., 2011). Since having an extreme level of stress impairs the
ability to process information and regulate thoughts (Arnsten, 2009; Shansky & Lipps, 2013), the
communicative element (i.e., sensitivity, quality) would be more important when people
experience mild-to-moderate levels of stress as opposed to an extreme level of stress (Bodie,
Burleson, & Holmstrom et al., 2011). More discussion on the relationship between stress and

6

cognitive functioning, and its implication in the context of supportive communication is provided
in discussion.
Emotional Support
The concept of social support has been widely studied among social psychologists due
to its key role in alleviating the negative impact of stressful events on physical as well as
psychological health (e.g., Cobb, 1976; Cohen & Hoberman, 1983; Cohen & Wills, 1985;
Siedlecki, Salthouse, Oishi, & Jeswani, 2014; Thoits, 2011; See Holt-Lunstad, Smith, & Layton,
2010 for a recent meta-analysis). In general, social supports are resources exchanged between
individuals with the purpose of helping each other (House & Kahn, 1985). Functionally, they can
be classified based on the type of resources conveyed through supportive interaction (Cohen &
Hoberman, 1983; Cutrona & Russell, 1990; House, 1981). For example, House (1981) identified
four categories of support: informational support, emotional support, appraisal support, and
instrumental support. Cutrona and Russell (1990) also proposed similar classification involving
emotional support, social integration support, esteem support, tangible support, and information
support. This taxonomy has been continuously applied in social support studies, including those
conducted in online and SNS contexts (e.g., Donovan, LeFebvre, Tardif, Brown, & Love, 2014;
Hirsch & Barton, 2011; Oh & Lee, 2012; Oh et al., 2013). Recently, Thoits (2011) suggested two
overarching functional categories of social support: coping assistance and emotional sustenance.
Coping assistance includes advice or instrumental aids that are problem-focused, whereas
emotional sustenance provides caring, sympathy, and comfort for the purpose of reducing stressinduced negative affect.
The present study focuses on a specific stressful situation that requires emotional
support. From the functional perspective, emotional support conveys, “comfort and security
7

during times of stress that lead a person to feel he or she is cared for by others” (Cutrona &
Russell, 1990, p. 322). It is often treated as an intentional and goal-directed behavior “enacted by
one party with the intent of helping another cope effectively with emotional distress” (Burleson,
2003, p. 552). Although different types of support can be more or less effective depending on the
situation, emotional support has been generally perceived as helpful and appropriate across
various situations (Burleson & Goldsmith, 1998; Cutrona & Suhr, 1992).
Acknowledging that the effectiveness of emotional support depends largely on whether
the provider’s supportive attempt is thoughtfully conveyed through supportive communication,
many communication scholars have dedicated their research to understanding the mechanisms of
supportive communication that conveys emotional support (e.g., Burleson, 2003; High & Dillard,
2012; Holmstrom, Burleson, & Jones, 2005; Jones, 2004). Following this line of literature, the
current study also focuses on emotion-oriented communication from the perspective of support
seekers and reveals how people communicate about stressful events to activate emotional support
on social network sites. The next section summarizes previous studies on social support in the
context of social network sites and discusses how SNS scholars have examined supportive
communication occurring on these sites.

8

Computer-Mediated Social Support and Self-Disclosure on Social Network Sites
Computer-Mediated Social Support on Social Network Sites
Online communication technologies provide people with new opportunities for building
relationships and receiving social support. Although earlier studies focused on how individuals’
face-to-face support networks become weakened with their increased Internet use (e.g., Kraut et
al., 1998), later studies have revealed positive aspects of online communication by showing how
CMC users build online social networks and exchange support with each other (e.g., AmichaiHamburger, Kingsbury, & Schneider, 2013; Eastin & LaRose, 2005; LaRose, Eastin, & Gregg,
2001; Turner, Grube, & Meyers, 2001; Walther & Boyd, 2002).
With the emergence of online platforms built for social networking, researchers have
revealed how people utilize social network sites to exchange social support (e.g., Kim & Lee,
2011; Oh et al., 2013; Wright et al., 2013). Social network sites are communication platforms
that offer various communication technologies for forming and maintaining “a wide network of
social connections” (Ellison et al., 2011, pp. 875-876). A recent report showed that SNS users
perceived a greater level of social support (including emotional, companionship, and tangible
support) than did general Internet users (Hampton et al., 2011). The perception of social support
on social network sites has been associated with positive psychological outcomes such as
enhanced self-esteem, self-efficacy, and subjective well-being (Burke, Marlow, & Lento, 2010;
Oh et al., 2013; Oh, Ozkaya, & LaRose, 2014; Phua, 2013; Steinfield, Ellison, & Lampe, 2008).
Another line of SNS research took a sociological approach to investigate social support.
Social capital, which refers to the resources people acquire through social interactions (Putnam,
2000), has been a popular concept among many SNS researchers (Ellison, Gray, Vitak, Lampe,
& Fiore, 2013; Ellison, Steinfield, & Lampe, 2007, Ellison et al., 2011; Jung, Gray, Lampe, &
9

Ellison, 2013; Park et al., 2009; Valenzuela, Park, & Kee, 2009; Vitak & Ellison, 2012; Vitak,
Ellison, & Steinfield, 2011). Most of these studies followed Putnam’s (2000) classification of
bridging and bonding social capital. This classification partially reflects the functional
dimensions of social support. Bridging social capital, which refers to the benefits of loosely
connected interpersonal ties (i.e., acquiring novel information or new perspectives), captures
information and network-related support. Bonding social capital, the resources exchanged
between individuals in a close relationship (i.e., exchanging emotional support or tangible aids)
captures the emotional and instrumental functions of social support. Typically, both types of
social capital were studied as outcomes that were positively associated with individuals’ online
social networking activities (Ellison et al., 2011; Steinfield et al., 2008; Valenzuela et al., 2009).
Although these studies have reported various support-related benefits that social network
sites offer, such as ease of widely distributing support-seeking messages or providing quick
responses to friends who need support (Vitak & Ellison, 2012), only a few studies have
examined how people actually solicit help on social network sites. A study that examined 20,000
Facebook status updates containing requests for action showed that, on average, status updates
mobilizing resources contained more socially oriented words (i.e., words activating a specific
member of one’s network to provide help) than did non-mobilizing updates. In addition, the
request status updates showed more emotional expressions and question marks than did the nonrequesting status updates (Ellison et. al., 2013). When asking for a small favor on social network
sites, offering incentives increased the likelihood of receiving responses from SNS friends (Jung
et al., 2013). Revealing affiliations and combining multiple strategies (e.g., demonstrating
gratitude and revealing incentives or affiliations at the same time) also increased the number of
SNS friends who responded to the help-seeking message. These studies addressed the
10

characteristics of support-seeking messages prevalent on social network sites and discussed their
effectiveness in acquiring desired responses. However, the support-seeking strategies found in
these studies were more exploratory than theory-grounded; the strategies were not hypothesized
from theories but rather drawn from post-hoc analyses of strategies reported by study
respondents. Therefore, there is a need to develop a theory-driven framework to understand the
support-seeking behavior on social network sites and predict features of the messages people
post on those sites to activate emotional support from their SNS friends.
Self-Disclosure on Social Network Sites
Since one must disclose a problem or feeling to activate supportive communication, the
self-disclosure literature is relevant to understanding the different motivations and processes
underlying support-seeking behavior and the disclosure of personal matters. Self-disclosure
refers to the communication about oneself including personal characteristics, opinions, feelings,
and experiences (Chelune, 1975). Previous studies on self-disclosure on social network sites
found that users tended to disclose personal information associated with socially desirable
qualities such as popularity, social connectedness, or an out-going personality (e.g., Zhao et al.,
2008). Studies have also shown that SNS users present positive events or feelings more
frequently than negative ones (Bazarova, Taft, Choi, & Cosley, 2013; Qiu et al., 2012). The idea
of selective self-presentation on CMC theoretically supports such disclosure patterns by claiming
that people carefully select information about themselves to be shared online based on the goals
that they would like to emphasize, and CMC technologies allow tactical presentation of self
(Ellison, Heino, & Gibbs, 2006; Gibbs, Ellison, & Heino, 2006; Rosenberg & Egbert, 2011;
Walther, 2007). On the other hand, another line of research has shown that people also disclose

11

negative feelings (Al-Saggaf & Nielsen, 2014; Lee et al., 2013; Moreno et al., 2011) and receive
social support through honest self-disclosure (Kim & Lee, 2011).
Recently, Bazarova and Choi (2014) proposed an integrative view of self-disclosure on
social network sites. This framework, known as the functional model of self-disclosure on social
network sites, claims that SNS affordances amplify the importance of certain communication
goals and further shape patterns of self-disclosure. SNS affordances can be defined as the
intersection of SNS features and the utilities users derive from those features (Treem &
Leonardi, 2012). For example, when sharing personal stories via public SNS channels such as
wall postings or status updates, people tend to prioritize the impression management goal and
disclose selectively edited versions of self. On the other hand, when using private SNS channels
to interact with a specific communication partner, people are more likely engage in an intimate
form of communication and willingly disclose private, sensitive information.
Although the functional model of self-disclosure on social network sites provides a
general overview of self-disclosure on social network sites, it does not explain how the general
motivation amplified by the communication context interacts with the support-oriented goal
people pursue when they disclose stressful situations to solicit emotional support on social
network sites. The present study seeks to address these questions by examining whether people
pursue an impression management goal when they seek emotional support on social network
sites, whether the communication context, especially the publicness of the support-seeking
behavior, intensifies the importance of the impression management goal, and how the goal
importance shapes the composition of the support-seeking CMC messages. The next section
provides a synthesis of previous studies on supportive communication and discusses the
theoretical implications of investigating support-seeking behavior on social network sites.
12

Support Providers, Support Seekers, and Supportive Communication
A considerable body of literature on supportive communication has suggested the
characteristics of providers, seekers, and situations that are associated with accomplishing
desirable support outcomes (e.g., Barbee et al., 1990; Burleson, 2003; Goldsmith, 2000, 2004;
High & Dillard, 2012; Jones, 2004; MacGeorge et al., 2002, 2003; Mortenson, 2009). Most of
these studies, however, focused on either understanding communication elements of effective
comfort or exploring provider characteristics associated with the support-providing behavior.
They seldom examined whether and how support seekers strategically address the need for
support in order to accomplish their support-seeking goals effectively. This section summarizes
the mainstream research on supportive communication and highlights the scarcity of academic
attention to the support seeker and the support-seeking behavior.
Features of Effective Supportive Communication
Whereas studies of social support have involved a broad range of tangible and intangible
resources, communication-oriented studies on social support have focused on the communication
elements of supportive interactions that help the recipient feel better. Treating communication as
the core element of social support, many communication scholars have devoted their research to
understanding the features of effective support-conveying communication (e.g., Burleson, 2003;
Burleson, Holmstrom, & Gilstrap, 2005; High & Dillard, 2012; Jones, 2004; Jones & Wirtz,
2006; Lehman, Ellard, & Wortman, 1986; Lemieux & Tighe, 2004; MacGeorge et al., 2002).
Studies on supportive communication were largely based on experiments in which
participants rated the comforting messages either directly or indirectly across different stressful
situations. For example, in Jones’s (2004) experiment, participants disclosed their upsetting
event to a confederate, reported their emotional state after the conversation with the confederate,
13

and evaluated the supportive attempt (i.e., helper’s motivation, supportiveness, and
expressiveness) after they received a comforting message from the confederate. In MacGeorge et
al.’s (2002) experiment, participants read a hypothetical scenario about a person facing a
stressful situation (i.e., having failed a midterm exam) and evaluated the quality of the different
comforting messages (i.e., effectiveness, helpfulness, appropriateness, and sensitivity). In these
experimental studies, person-centered messages, which explicitly acknowledge, elaborate, and
grant legitimacy to thoughts and feelings of the distressed (Burleson & Goldsmith, 1998, p. 251)
accomplished positive supportive outcomes, and were evaluated positively by the recipients or
observers (see High & Dillard, 2012 for a meta-analysis). On the other hand, supportive
communication was evaluated as unhelpful when the support provider criticized or invalidated
the seeker’s feelings (Burleson & Samter, 1985; Lehman et al., 1986), asserts how the seeker
should feel (Goldsmith, 1994), focuses on the provider’s own feelings (Lehman et al., 1986), and
becomes overly involved in the problem (Dunkel-Schetter, Blasband, Feinstein, & Herbert,
1992). These unhelpful behaviors often failed to achieve positive support outcomes and
negatively affected the relationship between the provider and the seeker (Burleson, 2003).
Support Providers
Along with the studies conducted to identify the key elements of effective supportproviding communication, a large number of studies have explored various provider-oriented
characteristics associated with support intentions and the quality of support. Although helping
behaviors are often automatically triggered by prosocial impulses or unconscious action rules
(e.g., picking up a pencil when someone drops it), they are also constrained by conflicting goals
(Macrae & Johnston, 1998) or internal states (e.g., being exhausted) (Dewall, Baumeister,
Gailliot, & Maner, 2008).
14

Communication-oriented studies on supportive behavior have treated comforting or
support-providing communication as a goal-driven act (Kellermann, 1992). Following that
assumption, studies have revealed different motivating factors associated with the provision of
emotional support. One of the strongest motivating factors may be the perceived responsibility of
the target (Dunkel-Schetter & Skokan, 1990; Jones & Burleson, 1997; Jung, 1988, also see
Weiner, 2000). Previous studies found that perceived responsibility negatively influences
intention to support the target and the quality of the comforting message (e.g., Jones & Burleson,
1997; Jung, 1988; MacGeorge, 2001; Otten, Panner, & Waugh, 1988). Providers also consider
their own ability to provide effective support (MacGeorge et al., 2002). Feeling empathy
(Burleson, 1983), being in a good mood (Cunningham, Shaffer, Barbee, Wolff, & Kelley, 1990;
Egbert, 2003), and being familiar with the stressful event faced by the recipient (Egbert, 2003)
were found to enhance the likelihood of providing effective support.
Support providers further differ in the level of skills or cognitive abilities needed for
providing sophisticated support (Burleson, 2003; MacGeorge et al., 2003). Whereas motivations
predict support behavior or supportive effort, skill variables are more closely related to the actual
quality of the comforting message (Burleson, 2003). For example, in several experimental
studies, the level of provider’s cognitive complexity (i.e., the ability to differentiate between
different constructs and deal with problems in abstract terms) positively influenced the provision
of effective comforting messages (e.g., Burleson, 1983; Samter, 2002).
Support Seekers
While a wealth of studies were conducted to understand support-providing
communication, only a few attempts were made to understand support seekers and their supportsoliciting communication. Although support is more likely to be given to those who actively seek
15

it (Ballantine & Stephenson, 2011; Eastin & LaRose, 2005; Newman, Lauterbach, Munson,
Resnick, & Morris, 2011; Wright et al., 2013), existing studies treated support seekers as either
passive processors of supportive messages or ineffective communicators whose support-seeking
attempts are often hindered by other concerns (e.g., impression management, relationship
maintenance etc.).
The majority of existing studies on support seekers were outcome-oriented, in that they
focused on the recipient’s psychological outcome following the interaction or cognitive
processing of the supportive message (Bodie, 2013; Bodie & Burleson, 2008; Burleson, 2009;
Holmstrom et al., 2013; Jones &Wirtz, 2006). For example, the dual-process theory of
supportive communication outcomes explains recipient outcomes of supportive interaction
(Bodie & Burleson, 2008; Burleson, 2009). It posits that the effect of supportive communication
is determined not only by features of the message but also by how the recipient cognitively
processes the message. When recipients are motivated and have the ability to carefully process
the provider’s message, the effectiveness of the supportive attempts would be determined by the
quality of the message itself (Bodie & Burleson, 2008; Burleson, 2009). When recipients are not
motivated or are unable to process the message, their evaluation of the message would be
influenced by other easy-to-process cues such as source characteristics (Burleson, 2009) or
relational closeness between the provider and the recipient (Holmstrom et al., 2013).
A relatively small number of studies have investigated the motivations or concerns
behind support-seeking behavior. Sensitive Interaction Systems Theory (SIST), developed by
Barbee and her colleagues (Barbee et al., 1990, 1998; Barbee & Cunningham, 1995), is one of
the few attempts that sought to theorize support-seeking behavior and determine situational
factors associated with it. This theory delineates how emotional conflicts and ambivalence might
16

influence supportive interactions and outcomes. A part of this framework is oriented to support
seekers; it explains how discrete emotions (i.e., sadness, hope, anxiety, anger, embarrassment,
disgust) generate expectations about helpful support and determine support-seeking behaviors.
For example, when experiencing an uncontrollable event (e.g., loss), a person may feel sad and
try to seek solace from others through engaging in direct nonverbal support-activating behavior
such as crying. When experiencing a negative event that is internally caused and not alterable,
the person might feel embarrassed and try to escape from such a feeling by using an indirect and
nonverbal support-activating strategy such as sighing or fidgeting (Barbee et al., 1998, p. 286).
Although several empirical studies seem to support this framework (e.g., Derlega,
Winstead, Oldfield, & Barbee, 2003; Mortenson, 2009), this theory has been criticized by other
scholars due to its lack of specification in describing each strategy and its weak theoretical
explanations of the relationships suggested by the framework (e.g., Burleson, 2003; Goldsmith,
1994). Moreover, the framework involves nonverbal support-activating behaviors (i.e., cry, sigh)
and presumes that those behaviors are also driven by strategic consideration, which in fact are
largely automatic and uncontrollable. More fundamentally, this framework focuses on how event
characteristics and emotions hinder support-seeking behavior, and therefore, it provides limited
information on how support seekers can address situational constraints and effectively activate
supportive responses from others when they are motivated to do so.
To address gaps in the existing support-seeking literature, the present study examines
how individuals compose a support-seeking message, and whether the communicative context
(i.e., social network sites) of seeking emotional support triggers certain communication goals and
shapes the content of the support-seeking message. By viewing support seeking behavior as a
specific form of communication, this study reveals the underlying motivations and goals of such
17

communication efforts and explores how support seekers endeavor to address multiple goals
involved in the support-seeking situation on social network sites. The next section introduces the
main theoretical framework utilized in this study to explain the presence and the role of
communication goals in support-seeking behavior on social network sites.

18

Multiple Goals of Seeking Emotional Support on Social Network Sites
The present study employs a theoretical framework of multiple goals to explain how
other concerns and motivations shape the messages people use to seek emotional support on
social network sites. Support-seeking communication is a specific type of interpersonal
communication that conveys multiple goals involving social support and impression
management (Albrecht & Goldsmith, 2003). Support-seeking behavior is likely to be initiated by
a desire to receive support, but the ways of accomplishing the support-receiving goal can be
altered when people become motivated to serve other interpersonal goals.
Multiple Goal Framework
Communication goals are “desired end states for which individuals strive” during
interaction (Berger, 2007, p. 50). Multiple goal perspectives stem from the premise that
communication is inherently a goal-driven act (Berger, 2007; Caughlin, 2010; Dillard et al.,
1989; Kellermann, 1992; O’Keefe, 1988; Wilson, 2007). Multiple goal frameworks also share
the common idea that individuals tend to pursue multiple communication goals consciously or
unconsciously during an interaction and that sometimes those goals conflict with and constrain
each other (Caughlin, 2010).
Dillard’s (1990) Goals-Planning-Action (GPA) model provides a goal-based explanation
of interpersonal communication. This model distinguishes between primary goals and secondary
goals. The primary goal is an influence goal guided by “the desire to bring about behavioral
change in a target person” (p. 20). Secondary goals, which are objectives that “derive directly
from more general motivations that are recurrent in a person’s life” (p. 20), become temporarily
salient by realizing the primary goal one attempts to achieve through communication. In short,
primary goals motivate and initiate an interaction, and secondary goals shape features of the
19

message content (Dillard et al., 1989; Dillard, 1990). The primary goal is the dominant goal not
in the sense that it is the most important goal that overwhelms other goals but rather because it
defines and determines what the interaction is about (Caughlin, 2010; Dillard et al., 1989).
Dillard et al. (1989) identified six secondary goals that are commonly used in the goal
literature: (a) identity goal, (b) interaction goal, (c) relational goal, (d) personal resource goal,
and (e) arousal management goal (Dillard, 1990; Schrader & Dillard, 1998). Identity goals
involve one’s ethical and normative standards whereas interaction goals are oriented to
impression management and social approval. Relational resource goals focus on relationship
management whereas personal resource goals focus on acquiring tangible resources. Finally,
arousal management goals focus on managing one’s anxiety during the interaction.
The idea of multiple goals has been widely applied in the domain of interpersonal
communication and social relationships (e.g., Caughlin, 2010; MacGeorge, 2001; Meyer, 2002,
2012; Samp & Solomon, 1998, 1999, 2005). Studies have identified how different secondary
goals arise across various communicative situations (MacGeorge, 2001; Meyer, 2002; Samp,
2006) and further revealed how those goals influence communication behavior (Samp &
Solomon, 1998; 1999; 2005). Samp and Solomon (2005), for example, found evidence that
people pursue multiple relationship goals when communicating problematic events to their
romantic partner. In their study, the relative importance and relationships among various goals
influenced message embellishment, which is the message feature of pursuing certain goals
through repeating or rephrasing goal-relevant themes or concepts (Samp & Solomon, 2005).
Politeness Theory also suggests appropriate facework across various communication
situations to achieve secondary goals of protecting positive face (i.e., gaining social approval)
and/or negative face (i.e., maintaining autonomy) of the communicators (Brown, 1987).
20

Empirical studies conducted within this framework revealed that communicators often pursue the
secondary goals of mitigating anticipated threats to face and compose a message in a way that
reduces such threats (Meyer, 2002; Wilson, Aleman, & Leatham, 1998). For example, Wilson et
al. (1998) found that people concern about maintaining positive face for both parties when
providing advice to their friends. When people pursue the face-mitigating secondary goal, they
are more likely to provide reasons to justify their advice so that the friend voluntarily follows the
advice without feeling face-threatened (Wilson et al., 1998).
The multiple goal framework would be particularly useful in understanding how people
compose a CMC message to accomplish multiple goals. Although this framework is not immune
from the criticism that communication often occurs automatically without deliberate planning or
a conscious effort to accomplish goals (Caughlin, 2010), it can perhaps explain the role of
motivations in computer-mediated communication on social network sites more accurately than
it does for face-to-face communication. According to the functional model of self-disclosure on
social network sites (Bazarova & Choi, 2014), affordances of SNS channels activate different
communication goals, which then shape the content of the CMC message. For example, in a
survey study of Facebook users, the affordance of status updates activated users’ social
validation goals (i.e., validation of self-concept and social approval) more strongly than did those
of wall posts or private messages (Bazarova & Choi, 2014). Relationship development goals
were more prominent in private messages and wall posts than they were in status updates. These
goals further mediated the relationships between the SNS affordances and disclosure intimacy.
Although the present study is primarily based on the multiple goal framework, this study
is expected to provide new insight to the functional model of self-disclosure on social network
sites. The functional model of self-disclosure is informative, as it provides a broad overview of
21

how SNS affordances activate general communication goals (e.g., relationship development,
social validation, information sharing) and further determine disclosure intimacy. But this
framework has yet to provide a context-specific understanding of whether those general goals
triggered by the SNS affordances shape or interact with the primary goal of communication that
initiates the conversation on social network sites. Employing the concept of multiple goals, the
present study examines how general (secondary) communication goals prompted by SNS
affordances interact with the primary goal (i.e., receiving emotional support) and shape the
content of support-seeking CMC messages. By doing so, this study would contribute to the
advancement of the functional model of self-disclosure, particularly by enhancing its
applicability in a specific communication context of seeking emotional support on social network
sites. The next section integrates the multiple goal framework with research on social support to
understand how the importance of the primary goal changes according to the nature of the
stressful situation.
Primary Goal of Support-Seeking Communication
As described earlier, the present study examines whether and how people pursue multiple
goals when they solicit emotional support on social network sites. This study proposes that when
people desire emotional support from others, they will attempt to influence others’ feelings and
thoughts to provoke supportive responses. In this case, the support-receiving goal would be the
primary goal that initiates and guides the communication effort.
Primary goals vary in their importance (Schrader & Dillard, 1998). In the context of
seeking social support, the severity of the problematic situation would be the key factor
motivating individuals to seek social support. In other words, as the level of stress increases,
people would desire for emotional support from others. The existing support literature indirectly
22

supports this claim by showing a positive relationship between problem severity and supportseeking intention (Dunkel-Schetter, Folkman, & Lazarus, 1987; Flett et al., 1995; Spendelow &
Jose, 2010; Wilcox & Birkel, 1983).
Findings relevant to the processing of supportive messages suggest that people become
more upset and motivated to thoughtfully process supportive messages when they face
moderately serious problems as opposed to mildly serious problems (e.g., Bodie, 2013; Bodie,
Burleson, & Jones, 2012; Bodie, Burleson, Holmstrom, et al., 2011; Holmstrom et al., 2013).
Although these studies focus on when people process comforting messages more or less
carefully, the results indirectly inform us that problem severity influences motivation to mitigate
negative feelings. Studies on negative emotions have also suggested that the more intense the
emotion, the more motivated the person feeling it becomes to regulate it (e.g., Barrett, Gross,
Christensen, & Benvenuto, 2001). However, as described earlier, an extreme level of stress
caused by major life events would sometimes impede the cognitive functions needed for
regulating negative emotions and stress (Arnsten, 2009; Shansky & Lipps, 2013). So focusing on
problematic events that produce mild-to-moderate levels of stresses, Hypothesis 1 is proposed to
test the relationship between severity of the problem (mild vs. moderate) and importance of the
support-receiving goal.
Hypothesis 1: When seeking support on a social network site, individuals will perceive
the support-receiving goal as more important when they communicate about a stressful
event that is moderately severe as opposed to one that is mildly severe.

23

Secondary Goal of Support-Seeking Communication on Social Network Sites
In addition to the nature of the problem, the present study proposes that the
communicative context of seeking support would trigger other communication goals. According
to the multiple goal framework, adoption of the primary goal causes consideration of secondary
goals (Schrader & Dillard, 1998). Since support-seeking behavior is inherently face-threatening
(Goldsmith, 1992), there is a risk that support seekers’ positive self-image can be threatened
through their support-activating attempt. Support seekers, however, would not always
consciously be aware of or motivated to manage their impression when they seek support.
Perceived importance of impression management would vary depending on the communicative
context (Schrader & Dillard, 1998). The present study proposes that social network sites would
provide a communication environment that prompts people with the impression management
concern; in particular, expected publicity of the support-seeking message would motivate people
to pursue the impression management goal.
Public vs. Private Communication on Social Network Sites
Social network sites provide various applications and functions that people can utilize to
network with others. Burke, Kraut, and Marlow (2011) distinguished between three types of
social activities that people engage in on social network sites: (1) directed communication, (2)
broadcasting, and (3) passive consumption.
Directed communication includes activities oriented to an individual SNS friend that are
more “personal and one-on-one” (p. 2). Directed communication occurs through emails, instant
messages, private chatting, and other channels that allow person-directed communication.
Directed communication often involves the rich content of supportiveness, positivity, and selfdisclosure. Such intimate interaction contributes to the strength of the relationship. Directed
24

communication can occur through both public and private SNS channels, but the recipient often
perceives personal disclosure as more intimate when it occurs through private SNS channels
(Bazarova, 2012).
The second type of activity is “broadcasting,” which targets a large audience. The
broadcasting message is not directed at a particular person but highly visible to a large number of
audience (Bazarova, 2012). The news feed application on Facebook, for example, provides
features such as status updates, profiles, and photos that allow users to disclose and share their
stories with an untargeted mass audience. Broadcasted messages are more likely to be shared
through public SNS channels. Although the content shared through public channels is less
intimate and in-depth, such open conversation also contributes to relationship building and
maintenance (Burke et al., 2011).
The last type of activity is mere consumption of content provided by others without
responding to it. This type of activity is similar to “lurking,” which is also passive consumption
of online content provided by other users (Kollock & Smith, 1996). Such activity occurs in
response to the second type of activity (broadcasting). This type of SNS activity also contributes
to relationship building indirectly through understanding and forming impressions about SNS
friends.
Publicity of Support-Seeking Message and Pursuit of Impression Management Goal
Since support-seeking attempts are enacted with communication behaviors, they should
be delivered through either directed/private messages or public postings. The functional model of
self-disclosure in social network sites posits that communication applications on social network
sites provide a complicated set of communication modalities involving different levels of
directedness and visibility (i.e., publicity) (Bazarova & Choi, 2014). For example, as described
25

above, SNS users may use the status update function on Facebook for public communication (no
directedness and high in visibility) or send a private message to their friend (highly directed and
low in visibility).
As described earlier, self-disclosure through different SNS channels serves different
motivations and goals as users construct different audiences in social network sites (Bazarova &
Choi, 2014). More specifically, public self-disclosure directed at general others addresses the
desire for self-validation, self-expression, and social approval. Private self-disclosure, on the
other hand, is more often made to serve relationship-oriented goals. The multiple goal
perspective also aligns with this idea as both argue that goals people pursue vary depending on
the context of communication (Schrader & Dillard, 1998).
Early studies on help-seeking behavior suggested that the behavior’s visibility could be
an importance determinant in soliciting help (Shapiro, 1978; Williams & Williams, 1983).
Although most of these studies were conducted in the context of instrumental help-seeking such
as feedback or advice, they provided evidence that the publicity level of the help-seeking attempt
influences whether and how a person solicits help from others. These studies argued that people
worry about feeling embarrassment, humiliation, or public shame by making their personal
problems public (Depaulo & Fisher, 1980; Williamson, 1974). Seeking help is more likely when
it is done privately as opposed to publicly, and face-saving tend to inhibit help-seeking behavior
(Tassler & Schwartz, 1972). In one study, the number of bystanders influenced help-seeking
behavior; those surrounded by three people were more reluctant to seek help than those who
were with just one other person (Williams & Williams, 1983).
A recent study outlining various benefits and costs of help-seeking behavior also
emphasized threat to one’s public image as an important inhibitor of support-seeking behavior
26

(Chan, 2013). Fear of public stigma (Eisenberg, Downs, Golberstein, & Zivin, 2009; Vogel,
Wade, & Hackler, 2007) and threat to public images of competence (e.g. Newman & Goldin,
1990) were the major concerns that inhibited people from seeking help publicly. To avoid
damaging their public image, some people chose to fail their tasks rather than solicit help (Chan,
2013). Although these studies were not conducted in the context of seeking emotional support or
comfort, the findings indirectly suggest that people want to maintain a positive and competent
public image, and this desire becomes greater as their behavior becomes more visible to a large
audience.
Understanding the role of impression management is particularly important in studying
support-seeking behavior on social network sites. Social network sites provide a unique
communication environment where individuals can “broadcast” their story to a whole network of
diverse friends, ranging from close friends to complete strangers (Gilbert & Karahalios, 2009).
The earlier help-seeking studies were conducted either in a face-to-face context with
homogeneous groups of strangers or acquaintances, or in CMC, which allows communicators to
remain completely anonymous. These earlier studies might not fully inform the dynamics of
different communication goals involving the diversity of audiences and the varying levels of
visibility that influence support-seeking communication on social network sites.
Moreover, self-disclosure on social network sites does not follow the traditional principle
of gradual self-disclosure (i.e., start by revealing casual information and incrementally move
toward revealing central information) or disclosure intimacy (i.e., people share intimate
information privately but not publicly). SNS users post personal and sometimes sensitive
information on social network sites and make it viewable to a large audience (Bazarova & Choi,
2014). So the questions arise: Does public support-seeking communication to a large and diverse
27

audience group on social network sites amplify the goal of impression management? If so, how
do people communicate to address the impression management goal involved in support-seeking
behavior on social network sites?
As a first step toward answering these questions, the present study proposes that the
visibility or the publicity of the CMC message on social network sites motivates individuals to
pursue the impression management goal. Self-presentation researchers have posited that the
publicity of one’s behavior (i.e., the probability that one’s behavior will be observed by others or
the number of others who might learn about the behavior) is central to impression management
(Leary & Kowalski, 1990, p. 38). Because SNS audiences are “no longer bounded by space and
time” as well as being “invisible and ill-defined” (Bazarova, Taft, Choi, & Cosley, 2013, pp.
122-123), the publicity of the message would motivate individuals to pursue the impression
management goal as audiences for the public modes (e.g., status updates and wall posts) can
easily expand beyond intended targets (Bazarova & Choi, 2014; Bazarova et al., 2013).
In sum, although SNS users expect supportive response from their SNS friends when they
disseminate a stressful story through a public channel on social network sites, they would also
anticipate that a larger audience beyond their circle of friends could also see their supportseeking message. The more public one’s behavior, the more likely one is to be concerned about
how behavior appears to others, and the more motivated one will be to manage one’s impression.
Therefore, Hypothesis 2 is proposed as follows:
Hypothesis 2: Individuals will perceive the impression management goal as more
important when they communicate about a stressful event via a public SNS channel as
opposed to a private SNS channel.

28

Production of Support-Seeking CMC Message on Social Network Sites
Production of Support-Seeking CMC Message
The present study proposes that CMC would allow individuals to construct a more
sophisticated support-seeking CMC message when they are motivated to do so than when they
are not. According to the hyperpersonal model of computer-mediated communication (Walther,
1996, 2007), CMC can lead to a more desirable interaction outcome than face-to-face
communication. This is partly because people can spend more time on message production and
editing, and they can focus solely on the message itself without managing other nonverbal
communication elements (e.g., facial expression). The asynchronous CMC tools further
encourage careful reflection of goals in the communication process as users are not forced by the
situation to communicate immediately (Duthler, 2006; Walther, 2007). Although this framework
was developed to explain interpersonal communication and relationship development on CMC,
the underlying assumption is that CMC offers features that allow individuals to achieve
communication goals more effectively.
These features, however, benefit users differently depending on users’ motivation and
interaction goals (Walther, 2007). In other words, people utilize CMC features to best serve their
goals only when they are motivated to do so. For example, users of online dating sites utilize
tools offered from the sites to present themselves in a more positive light. Their motivation for
visiting those sites is to find a dating partner, and they try to accomplish this by making
themselves look attractive (Ellison et al., 2006; Gibbs et al., 2006). Certain CMC features also
help users effectively address their motivation. In one study, individuals’ motivation to be polite
was better reflected in the communication message when they used an asynchronous medium as
opposed to a synchronous medium to send a request (Duthler, 2006). Social network sites also
29

provide a wide variety of communication tools (i.e., photos, wall posts, liking, commenting etc.)
that allow individuals to disclose personal information and develop social relationships. The
utilities of these features, however, depend on users’ willingness to disclose on those sites
(Trepte & Reinecke, 2013).
In a support-seeking context, the perceived importance of the support-seeking goal and
the impression management goal would influence support seekers’ motivation to manage those
goals. Goal complexity, which is a function of the number of different goals that are temporarily
dominant (Samp & Solomon, 2005; Schrader, 1999) will be greater when people pursue multiple
goals as opposed to a single goal. By having a complex set of multiple goals in mind, people will
become motivated to carefully tailor their message to their purpose (O’Keefe, 1988; Samp &
Solomon, 2005; Schrader, 1999). With the unlimited time they can use to compose a message
(Walther, 2007), individuals would put more effort into content selection and message
production when they are motivated to address more than one goal in their message.
In sum, when people use a public SNS channel to seek emotional support, the severity of
the problem will influence the amount of effort people put into the message as people will
become motivated to pursue both the support-seeking and the impression management goals.
They will spend more time to optimize their message for serving multiple goals effectively. In
contrast, when using a private SNS channel to seek emotional support, people will exert less
effort on composing a message than when they use a public SNS channel because people can
select and control the audience of their support-seeking message. With the impression
management concern being relatively low, people would only concern about receiving emotional
support, and thus, not become motivated to address multiple goals. Reflecting such reasoning,
Hypothesis 3 is proposed as follows:
30

Hypothesis 3: There is an interaction effect of the severity of the problem and the
publicity of the message on CMC message composition time.
Hypothesis 3-1: When using a public SNS channel to communicate about a stressful
event, individuals will spend more time on CMC message composition when the problem
is moderately severe as opposed to mildly severe.
Hypothesis 3-2: When using a private SNS channel to communicate about a stressful
event, CMC message composition time will not be influenced by the severity of the
problem.

Sophistication of Support-Seeking CMC Message: Message Design Logic Approach
In the context of seeking emotional support, the present study explores whether
affordances provided by social network sites amplify certain communication goals, and whether
people successfully accomplish their communication goals by carefully constructing a supportseeking CMC message to be posted on social network sites. What remains to be tested is whether
users’ efforts to compose support-seeking CMC messages are conveyed in the actual messages.
Do those who spend more time composing and editing their messages produce support-seeking
messages that are more sophisticated and goal-reflective than those who spend relatively less
time on message composition? This question is crucial, as examining the actual message is one
of the unique contributions expected from the present study.
The present study employs the message design logic approach to analyze the supportseeking messages produced by individuals on social network sites. The message design logic
framework is appropriate for the present study because its logical hierarchy is based on the
communication goals a message serves to address. According to O’Keefe (1988), messages
differ in their goal structure and design logic. Goal structure is determined by the number and
31

type of goals a person pursues when producing a message (Hullman, 2004; O’Keefe, 1988;
Schrader & Dillard, 1998; Scott, Caughlin, Donovan-Kicken, & Mikucki-Enyat, 2013). Message
design logic, then, refers to rules the message design a producer follows to accomplish goals
(O’Keefe, 1988). O’Keefe (1988) identified three types of design logic people use to convey
goals: expressive design logic, conventional design logic, and rhetorical design logic.
Expressive design logic is based on the premise that language is a medium for expressing
thoughts and feelings (O’Keefe, 1988). So individuals using the expressive design logic view
messages as simply conveying their thoughts or feelings without attending to or addressing
contextual demands while they communicate. They are aware of the hearers but the hearers are
assumed to receive and understand the message straightforwardly as communicated by the
speaker. Expressive messages are, therefore, self-oriented and seldom addresses goals that are
relevant in the situation (Hullman, 2004). In the present study, expressive messages would
simply convey seekers’ negative feelings (e.g., anger, sadness) or thoughts (e.g., complaints)
without explaining their cause.
Conventional design logic uses a more sophisticated design of messages than does
expressive design logic (O’Keefe, 1988). It is based on the premise that language achieves a
certain social effect by following socially appropriate rules and procedures. Using this logic,
individuals pursue a socially appropriate goal and select communication messages that
coherently contribute to accomplishing their goal (O’Keefe, 1988). Although rules governing the
accomplishment of communication goals might differ depending on the context of interaction,
they share core functions such as exchanging good news, promoting relationships, sharing social
support, and avoiding conflicts (Bryant & Marmo, 2012). Furthermore, the communicative
context of social network sites highlights particular rules pertaining to facework and public
32

image, especially when people engage in public communication (Bryant & Marmo, 2012).
Intimate disclosures, therefore, are often viewed as inappropriate in public as opposed to private
communication (Bazarova, 2012). Since conventional messages are goal-driven, such messages
emphasize the goals of the communicator and the techniques needed to achieve those goals
(Hullman, 2004). In the present study, conventional messages would address one salient goal that
would be either the support-receiving goal or the impression management goal.
Rhetorical design logic is the most sophisticated message design. This logic is based on
the idea that communication creates and balances social selves and situations. Messages
produced using rhetorical design logic are proactive (O’Keefe, 1988). Rhetorical messages
redefine a situation to address multiple goals involved in the communication (O’Keefe, 1988).
Rhetorical messages are typically longer and more complex, as this logic requires more
deliberate communicative efforts (Hullman, 2004). In a support-seeking context, rhetorical
messages would address multiple goals (i.e., support-receiving goal and impression management
goal) and try to redefine the negative event in a positive light or focus on a lesson learned
through the experience.
The framework of message design logic proposes that certain goals are conventionally
understood to be salient in certain situations. Although messages following expressive or
conventional logics are not inherently less valuable than those following the rhetorical logic, they
are often evaluated by the partner or the third party as less effective or appropriate than the
rhetorical logic (Hullman, 2004; Lambert & Gillespie, 1994; O’Keefe & McCornack, 1987;
Scott et al., 2013).
The present study proposes that asynchronous and editable features of CMC would allow
individuals to employ a sophisticated design logic addressing multiple goals when they are
33

motivated to do so. For example, when a person is highly motivated to pursue a support-seeking
goal and an impression management goal, the person will put more time and effort into
composing a message that serves both goals effectively. Although Walther (2007) found
relationships between motivation and message composition efforts, few studies were conducted
to examine whether these efforts are conveyed in the actual message. Following the message
design logic framework, Hypothesis 4 proposes that the amount of effort individuals put into
composing a support-seeking message is positively associated with the level of message
sophistication. In other words, as individuals spend more time on message production and
editing, their support-seeking message is more likely to include specific message elements that
reflect the rhetorical design logic.
Hypothesis 4: Message composition time is positively associated with the level of
sophistication in the support-seeking CMC message.
In sum, the present study aims to explore whether people pursue multiple goals when
they solicit emotional support on social network sites, and whether consideration of multiple
goals shapes the content of their support-seeking messages. More specifically, the study tests
whether the severity of a stressful situation motivates people to pursue the goal of receiving
support (H1), and whether the expected publicity of the support-seeking message further
motivates people to pursue an impression management goal in addition to a support-receiving
goal (H2). This study also examines whether individuals put more time and effort into CMC
message composition when they are motivated to pursue multiple goals (H3), and whether such
effort leads to the production of sophisticated CMC messages that respond to those goals (H4).
The body of existing knowledge about supportive communication has focused on
understanding what constitutes desirable supportive communication and identifying provider
34

characteristics associated with producing effective support. Scarce attention has been given to
understanding how people solicit emotional support to cope with a stressful situation and how
impression-relevant concerns associated with sharing a stressful story shape their supportseeking message production. Moreover, although prior studies on social network sites have
praised the beneficial outcomes of engaging in online social networking and emphasized social
support as one of the key benefits online social networking provides, no studies have employed a
communication-oriented theoretical framework to understand the mechanisms of the supportseeking CMC message production.
The present study would contribute to the theoretical advancement of social support and
supportive communication by providing a theory-grounded explanation of support-activating
behavior on social network sites. By examining how multiple goals shape the way people present
stressful situations and seek emotional support on social network sites, this study will contribute
to a better understanding of how support seekers can influence the quality of supportive
interaction. Since no study has discussed how people share their problems and stressful situations
on social network sites, findings from the present study will contribute to knowledge of how
people utilize social network sites to maintain psychological well-being. Finally, exploring how
people represent stressful events on social network sites and how affordances of social network
sites shape and highlight certain communication goals may help us better understand the unique
opportunities and constraints of social network sites in delivering needed support.

35

CHAPTER 3
METHOD

Participants
College students taking classes in the Communication department of a large Midwestern
university were recruited for extra credit. College students account for the largest percentage of
SNS users in the United States (Duggan & Smith, 2013) and thus are suitable subjects for this
study. A total of 178 students (ages 18–29) were recruited to participate in a laboratory
experiment for extra credit. Since there were no prior studies that could directly inform this study
on its typical effect size, the number of subjects was estimated by a priori power analysis using
G-Power 3.1 (Faul, Erdfelder, Buchner, & Lang, 2009) based on existing experimental studies
conducted in similar contexts that showed medium-to-strong effect sizes (see High & Dillard,
2012, for meta-analysis). Participants were screened out if they did not have a Facebook account.
Data from two participants were excluded as their emotional valence was extremely negative,
resulting in a total sample size of 176. Among 176 participants, 121 (68.8%) were male and 55
(31.3%) were female. Their ages ranged from 18 to 28 with a mean age of 20.86 (SD = 1.45). No
significant difference was observed among participants across the four conditions (Sex: χ2 (3) =
2.15, p = .54; Age: F (3, 172) = .69, p = .56). Finally, the number of SNS friends ranged from 8
to 4,000 (M = 837.96, SD = 733.12). Since the number of SNS friends was not normally
distributed (Skewness = 2.15, Kurtosis = 7.38), it was square-root transformed for further
analyses (transformed Mean = 26.48, SD = 11.72, Skewness = .60, Kurtosis = .81). Appendix A
provides probability-probability (p-p) plots of the raw and transformed data. Neither the original
nor the transformed number of SNS friends differed across the four conditions (original number
36

of SNS friends: F (3, 172) = .82, p = .48; transformed number of SNS friends: F (3, 172) = .60, p
= .62). The variances of the number of SNS friends were homogeneous across the four
conditions as well (original number of SNS friends: F (3, 172) = .76, p = .52; transformed
number of SNS friends: F (3, 172) = .41, p = .75). Table 1 summarizes the demographic
characteristics of the participants across the experimental conditions.

Table 1. Participants’ Demographic Characteristics and Number of Their SNS Friends
Public Channel

Total
Gendera
Male
Female
Ageb
SNS friendsb

Private Channel

Moderately
Severe

Mildly
Severe

Moderately
Severe

Mildly
Severe

Total

45 (25.6%)

41 (23.3%)

44 (25.0%)

46 (26.1%)

176 (100%)

28 (62.2%)
31 (75.6%)
29 (65.9%)
33 (71.1%)
17 (37.8%)
10 (24.4%)
15 (34.1%)
13 (28.3%)
21.04 (1.78)
20.68 (1.21)
20.98 (1.50)
20.72 (1.22)
982.78 (953.07) 764.27 (543.75) 781.39 (708.24) 816.09 (656.00)

121 (68.8%)
55 (31.3%)
20.86 (1.45)
837.96 (733.12)

Notes: a. Number of cases with percentages in the parentheses. b. Mean values with standard
deviations in the parentheses.

37

Procedure
The present study employed a two (problem severity: moderate vs. mild) by two
(publicity of message: public vs. private) between-subjects design. The experiment took place in
a computer lab with a forced exposure manipulation. Once participating students arrived at the
lab, they were instructed by a female experimenter who was blinded to the hypotheses. The
participants were asked to sit in front of a computer and self-administer the experiment by
following instructions that appeared on the computer screen. The involvement of the
experimenter was minimized during the experiment. Following completion of an informed
consent form, student participants were exposed to six randomly selected neutral images
developed by Lang, Bradley, and Cuthbert (2008) to neutralize and soothe their mood and
emotional state. The neutral images were pretested before they were used in the main study (see
Appendix B for example images and Appendix B for the arousal scores measured in the pretest).
Participants’ emotions were measured using the Self-Assessment Manikin (SAM) scale, which is
a picture-oriented instrument that directly assesses pleasure, arousal, and dominance in response
to an object or event (Lang, 1980). Images in the SAM scale are scalable vector drawings, and
emotions measured with this scale can be treated as continuous. This picture-based scale has
been widely validated across different psychological studies (see Bradley & Lang, 1994 for its
comparison with the multi-item emotion scale). The present study used two sets of SAM pictures
to examine participants’ emotional states associated with emotional valence and emotional
arousal. No significant difference was observed across the four experimental conditions in terms
of participants’ emotional arousal (F (3, 170) = .18, p = .18). Emotional valence across the four
groups showed a significant F-score (F (3, 172) = 2.89, p < .05), suggesting between-group
differences in participants’ emotional valence. Tukey’s post-hoc analysis, however, did not show
38

a significant difference between individual groups at the 95% significance level. Thus, the
neutralization treatment was successful in stabilizing participants’ emotional state. See Appendix
C for detailed SAM scores across the four conditions.
The present study employed a randomized block design embedded in the Qualtrics survey
software.1 After viewing the neutral images, participants were randomly assigned to read one
upsetting scenario and asked to assume that they were involved in the situation. Two problematic
scenarios were used, each of which was manipulated in two conditions to represent either a
mildly stressful or a moderately stressful version of the situation. Participants were then asked to
share a story about the event on Facebook, either by posting a status update about it or by
sending a private message to one of their SNS friends. The present study used Facebook as a
study platform due to its popularity among young American adults (Duggan & Smith, 2013).
Participants then wrote an actual message that they would either post on their status update or
send privately to one of their SNS friends in order to activate an emotional support. Participants
also responded to questions examining (a) the importance of the support-receiving goal and (b)
the importance of the impression management goal. These goal importance questions randomly
appeared either before or after the message was composed to randomize the effect of the
possibility that the questions would prompt participants to attend to these goals when composing
their message. Participants then provided their demographic information, answered questions
developed for manipulation checking, and were debriefed about the purpose of the study.

1

http://www.qualtrics.com/

39

Stimuli
Two types of stimuli were used in the experiment: (a) relaxing images, to neutralize
participants’ emotional state before they underwent the main experimental process, and (b)
written stressful scenarios, as manipulation of the problem severity variable.
This study used neutral photos from the International Affective Picture System (IAPS)
inventory developed by Lang and colleagues (2008). IAPS has a large set of standardized
emotion-provoking color photos developed to evoke a wide range of emotional categories.
Fourteen neutral images were selected for a pretest. These images contained natural scenes
consisting of green, blue, and yellow colors. Participants for the pretest were recruited by the
same means as for the main study. In the pretest, participants (N = 48) viewed seven randomly
selected images and reported their arousal level. Six semantic differential items with a 7-point
scale were used to measure participants’ arousal level (Bradley & Lang, 1994).2 Two pictures
(IAPS #5030, #7036) were excluded as their arousal scores were higher than the mid-point
(IAPS#5030: M = 4.03, SD = .82; IAPS#7036: M = 4.43; SD = .83; see Appendix B for detailed
scores.).
The scenario-based form of stimulus is widely used in empirical studies on supportive
communication and is known to produce a strong effect (see High & Dillard, 2012, for metaanalysis). This study adopted 12 scenario samples representing mild-to-moderate problematic
situations developed by Bodie and colleagues to be used in their studies on supportive
communication (Bodie, Burleson, Gill-Rosier et al., 2011; Bodie, Burleson, Holmstrom, et al.,
2011; Holmstrom et al., 2013). Bodie and colleagues originally created these scenarios to reflect

Relaxed – stimulated; calm – excited; sluggish – frenzied; dull – jittery; sleepy – wide awake; unaroused –
aroused. Cronbach’s alpha reliability scores ranged from .76 to .89.
2

40

stressful situations that are commonly experienced by college students (e.g., getting a parking
ticket, failing a quiz, losing a part-time job, relationship conflicts etc.) and are widely used in
previous studies in the context of emotional support (Bodie, Burleson, Gill-Rosier et al., 2011).
The present study pretested these scenarios in order to select one scenario pair that is (a) the most
realistic and (b) accurately reflects differences in problem severity. College students recruited for
pretesting the IAPS images were also asked evaluate six randomly chosen scenarios in terms of
perceived realism and severity of the problem. Appendix D provides the results of the first
pretest. Three scenario pairs (i.e., relationship issue, car trouble, scholarship loss) showed good
realism scores and significant differences in their severity scores, and so they were pretested one
more time using the perceived realism and severity tests, as well as the likelihood that people
would seek emotional support through Facebook for each situation. Participants (N =31) were
again recruited by the same means. This time, participants were instructed to read and evaluate
three random scenarios. After reading each scenario, they were also asked to compose a supportseeking message that they would either post on their Facebook status update or send to their SNS
friend using Facebook private message. This process was done to check if participants would
successfully produce an actual support-seeking message in an imaginary support-seeking
situation. The actual support-seeking messages produced by the participants were also used for
the coder-training purpose. Appendix E provides the results of the second pretest. The scenario
pair selected for the main study was the “car trouble” scenario, as it showed the highest scores
for perceived realism (M = 5.66, SD = 1.32) and likelihood of seeking support (M = 3.03, SD =
1.87). The car trouble scenario pair also showed a significant difference between the two
scenarios in terms of perceived severity of the situation (mildly severe scenario: M = 4.61, SD =
1.18; moderately severe scenario: M = 5.41, SD = 1.08; t (30) = -.3.55, p < .01). In this scenario
41

pair, both mildly and moderately severe conditions ask participants to imagine that they went to a
party with their car left in a university garage, and failed to retrieve the car the next morning
before the 7 a.m. ticketing time. The scenario outcome for the mild condition is a $25 parking
ticket, whereas the outcome for the moderately severe condition involves the car being towed
and a fine over $400. Appendix F shows the actual scenarios used in the main experiment.
Manipulation Check
After reading a scenario representing either a moderately or a mildly severe situation,
participants evaluated the severity of the situation using a 7-point semantic differential scale.
Participants also indicated their emotional states after they were exposed to the stressful
scenarios. The severity question appeared at the end of the experiment to ensure that it would not
contaminate the other results. The manipulation was successful: participants in the moderately
severe condition perceived the severity of the situation portrayed in the scenario as significantly
more serious than those in the the mildly severe condition (moderately severe condition: M =
5.52, SD = .94; mildly severe condition: M = 4.55, SD = 1.25, F (1, 171) = 34.46, p < .001). The
publicity of the message (public vs. private) did not influence the perceived severity of the
scenario (F (1, 171) = 1.45, p = .23), nor did the interaction of the scenario and the message
publicity (F (1, 171) = 1.37, p = .24). As for participants’ emotional states, pairwise t-tests
revealed that the participants felt significantly less pleasant but more aroused after they read a
stressful scenario regardless of its problem severity (emotional valence: Mean Difference = 3.97, SD = 1.70, t (172) = -30.79, p < .001; emotional arousal: Mean Difference = 3.16, SD =
2.55, t (173) = 16.33, p < .001). More specifically, participants who read the moderately severe
scenario felt significantly less pleasant than those who read the mildly severe scenario
(moderately severe scenario: M = 1.76, SD = .93; mildly severe scenario: M = 2.61, SD = 1.19,
42

F (1, 172) = 27.91, p < .001). In addition, those in the moderately severe condition were
significantly more aroused than those in the mildly severe condition (moderately severe
condition: M = 7.63, SD = 1.68; mildly severe condition: M = 6.37, SD = 1.87, F (1, 172) =
22.74, p < .001). Thus, manipulation was successful in upsetting participants and creating
differential levels of emotional states in individuals between different groups while not
confounding publicity. Appendix G provides the perceived severity scores as well as emotional
states across the experimental conditions and the results of the ANOVA tests.
Order Effect
In the main experiment, the order of the goal importance questions and the message
composition task was randomized to minimize the possibility that the goal importance questions
influence participants’ message composition behavior by prompting them with certain
communication goals. In order to check this possibility, the effects of the task order on the two
outcome variables (i.e., time spent on CMC message composition and level of message
sophistication) were examined. As a result, the task order did not significantly influence either
time spent on CMC message composition (with original time: t (174) = .16, p = .88; with logtransformed time: t (174) = .75, p = .46) or message sophistication (χ2 (3) = 1.96, p = .58).

43

Measures
Items for Pretest and Manipulation Check
Perceived severity of the problem was measured in both the pretest and the main study.
Four semantic items with a 7-point scale were used to measure perceived severity: not at all
serious—very serious; not at all severe—very severe; not at all upsetting—very upsetting; and
not at all stressful—very stressful. Cronbach’s alpha values were .78 and .93 in the pretests, and
.84 in the main experiment.
Perceived realism of the scenario was measured in the pretests in order to select the
scenario pair that would be the most likely to happen in the participants’ everyday lives. Three
semantic items with a 7-point scale were used to measure perceived realism: not at all realistic—
very realistic; not at all believable—very believable; and not at all likely to happen—very likely
to happen. Cronbach’s alpha scores were .67 and .95. These items were drawn from Holmstrom
et al. (2013).
Likelihood of seeking support from SNS friends in response to the scenario was
measured only in the second pretest. It was measured with the likelihood of (a) sharing negative
feelings with, (b) seeking to lean on, and (c) seeking comfort from Facebook friends using a 7point scale (Flett et al., 1995). Cronbach’s alpha reliability scores ranged from .92 to .98.
Goal Importance
Measures of the importance of primary and secondary goals were adopted from Dillard et
al. (1989), as their measures yielded reliable results in previous studies. Items were measured
with a 6-point Likert scale ranging from 1 (strongly disagree) to 6 (strongly agree). The items
measuring the importance of the primary goal yielded Cronbach’s alpha values of .85 and .87 in
the original study (Dillard et al., 1989). These items are general in nature, in order to encompass
44

various primary goals involved in different communication situations. Thus, they were modified
to more specifically measure the importance of the support-receiving goal. For example, the
item, “It was very important to me to convince this person to do what I wanted him or her to do,”
was modified to, “It is very important to me to convince my Facebook friends (this person) and
activate supportive response after they (he or she) read (s) my posting.”
Of the different secondary goals proposed by Dillard et al. (1989), the present study
focused on the impression management goal, which Dillard labeled as the interaction goal,
defined as the extent to which one desires to manage his or her impression and stay socially
appropriate. Therefore, Dillard et al.’s five items measuring the importance of the interaction
goal were used to measure the importance of the impression management goal. These items
showed Cronbach’s alpha values of .71 and .72 in the original study (Dillard et al., 1989). The
items were also modified to fit this study context more specifically. An example item is, “I am
concerned with putting myself in a “bad light” in this message.”
Confirmatory factor analysis was conducted with AMOS 22.0 (IBM Corp., 2013) to
validate the structure of the two multi-item variables. After fitting a CFA model, two items were
deleted from the model due to their low factor loading scores (i.e., “I am very concerned about
getting supportive response with this message” and “I am concerned with making (or
maintaining) a good impression in this message.”). After deleting these two items from the
model, the CFA model showed a good fit (χ2 (19) = 27.53, p = .09, GFI = .96, CFI = .99, TLI =
.99, NFI = .97, RMSEA = .05, SRMR = .04). The reliability score was .86 for the importance of
the support-receiving goal and .90 for the importance of the impression management goal. Factor
loadings ranged from .69 to .87 for the importance of the support-receiving goal, and from .60 to
.94 for the importance of the impression management goal. Composite scores of the two
45

variables were calculated by weighing each item by its unstandardized factor loading score.3 The
composite scores of the two variables were both normally distributed (Skewness = -.96, -.29,
Kurtosis = .27, -1.25, respectively). The mean value of the importance of the support-receiving
goal was 4.32, and its standard deviation was 1.21. The mean value of the importance of the
impression management goal was 3.74 with the standard deviation of 1.56. Table 2 shows
descriptives, reliabilities, and factor loading scores of the two multi-item variables.

3

Importance of Support-Receiving Goal = Goal1 × (1/4.317) + Goal3 × (1.246/4.317) + Goal4 × (0.985/4.317) +
Goal5 × (1.086/4.317); Importance of the Impression Management Goal = Goal7 × (1/3.591) + Goal8 ×
(1.015/3.591) + Goal9 × (0.959/3.591) + Goal10 × (0.617/3.591).

46

Table 2. Descriptive, Reliability, and Validity of Key Variables
Mean (SD)
Importance of Support-Receiving Goal

Factor
Loading

Alpha

4.32 (1.21)

It is very important to me to convince my Facebook friends
(this person) and activate supportive response after they
(s/he) read (s) my posting.
I really care whether my Facebook friends (s/he) comfort (s)
me.
The outcome of this message will have an important
consequence for me.
I really want my Facebook friends (this person) to provide
me with necessary support.
Importance of Impression Management Goal
I will be careful to avoid saying things, which are socially
inappropriate.
I will be very conscious of what is appropriate and
inappropriate in this message.
I am concerned with putting myself in a “bad light” in this
message.
I don’t want to look stupid in this message.
Time spent on CMC Message Composition
Original Timea
Log-Transformed Time

Notes: N = 176. a. Originally measured in seconds.

47

4.33 (1.41)

.74

4.35 (1.48)

.87

4.17 (1.47)

.69

4.43 (1.38)

.82

.86

3.74 (1.56)
3.60 (1.81)

.89

3.77 (1.74)

.94

3.55 (1.77)

.88

4.19 (1.66)

.60

107.37 (56.48)
1.97 (.23)

.90

Time Spent on Message Composition
The time spent on composing a support-seeking message was measured by recording the
time participants stayed on the page where they were asked to compose a message to be either
posted on their Facebook status update or shared through a private message. The survey tool
provided by Qualtrics allows researchers to record the time participants stay on each web page.
Table 2 provides means and standard deviations of the time scores across the four conditions.
The time participants spent on composing the CMC message ranged from 21.01 to 311.43
seconds (M = 107.37, SD = 56.48). Normality of the distribution was checked, and the skewness
value suggested a slight deviation of the data from the normal distribution (Skewness = 1.09,
Kurtosis = 1.36). Log-transformation provided the best distribution of the data (Skewness = -.21,
Kurtosis = -.30); for this reason, results from the original data are provided along with those
using the log-transformed data. Appendix A provides p-p plots of the original and the
transformed data.
Level of Message Sophistication
The level of message sophistication can be operationalized as the extent to which the
message reflects a more sophisticated design logic. O’Keefe (1988) proposed a theoretically
valid hierarchy of sophistication in the three design logics: a person must be able to express
internal thoughts and feelings (expressive logic) before being able to communicate them in an
appropriate manner (conventional logic), and a person must learn appropriate goals in a given
situation before being able to address multiple goals by redefining the situation (rhetorical logic)
(Scott et al., 2013). Later studies also confirmed that messages reflecting the rhetorical logic are
the most effective compared to messages following the conventional logic or expressive logic
(Hullman, 2004; Lambert & Gillespie, 1994; O’Keefe & McCornack, 1987; Scott et al., 2013).
48

Messages employing the expressive design logic serve no goal and are often evaluated as the
least sophisticated.
The present study therefore rated messages following the rhetorical design logic as the
most sophisticated, followed by messages reflecting the conventional design logic, and finally
those following the expressive design logic. Coding was done using a holistic approach, by
treating the entire message as one unit (O’Keefe & McCornack, 1987); this approach has been
widely used in previous studies on message sophistication and has been shown to be reliable.4
Two pre-trained coders coded each message as follows: no logic (0), expressive logic (1),
conventional logic (2), and rhetorical logic (3). The variable was treated as ordinal, and so there
was no assumption that the intervals between different levels of message logic were equal.
Although the message features defining each design logic differ across studies depending
on the context of the study (Caughlin et al., 2008; Peterson & Albrecht, 1996; Scott et al., 2013),
they share underlying themes, that expressive messages address no goal, that conventional
messages address an immediate goal, and that rhetorical messages serve multiple goals. Since the
present study used the message design logic approach to determine the level of message
sophistication, the logic reflected in each message was coded by following the common rules
used in previous studies. More specifically, a message was coded as “expressive” if it contained
explicit expressions of negative emotions or thoughts without explaining the context or serving a
goal. A message was coded as “conventional” if it served one immediate goal. A message
pursuing a support-receiving goal would contain (a) a direct request for support, (b) an
expression of the seeker’s expectations or need for support, or (c) a statement seeking empathy

4

Cohen’s Kappa was .83 in the original study (O’Keefe, 1988). Later studies also showed good reliabilities:
Cohen’s Kappa of .90 and 1.00 in Peterson and Albrecht (1996); intra-class correlations of .80 in Scott et al.
(2013) and .91 in Caughlin et al. (2008).

49

for the seeker’s emotions or stressful situation. A message pursuing the impression management
goal would contain (a) an explicit statement asking others not to evaluate the seeker negatively,
or (b) an implicit effort at impression management by showing positive aspects about oneself,
highlighting situational causes, or emphasizing the responsibility of others (Caughlin et al., 2008;
Scott et al., 2013). A message was coded as “rhetorical” if it contained an attempt to pursue both
support-receiving and impression management goals. Rhetorical messages further redefined or
reframed the stressful situation, the seeker, and the communication context in a positive light or
provided in-depth interpretation of the problem (Hullman, 2004; O’Keefe, 1988; Scott et al.,
2013). Finally, a message that only contained simple facts about the scenario without any
expression of thoughts or feelings was coded as having “no logic” and treated as the least
sophisticated.
Two pre-trained coders, each of whom was blind to the research hypotheses, analyzed
176 messages composed by the research participants. A coding protocol was developed
following the three-part approach (Riffe, Lacy, & Fico, 1998), which includes (a) the goals of the
study and definitions of the concepts, (b) the specification of the content to be analyzed, and (c)
the operational definitions of all the subcategories specifying how the coders assign values to
each case, including the messages selected for checking inter-coder reliability. Appendix H
provides the codebook, with a detailed coding protocol.
After an initial draft of the coding protocol was developed, the two coders underwent a
series of training sessions. Following the training process proposed by Riffe et al. (1998), the two
coders first spent time familiarizing themselves with the content. Next, the coders reviewed the
coding protocol and discussed ambiguities in the instructions. After the familiarization and initial
clarification process, the two coders independently coded 30 support-seeking message samples
50

based on the protocol. These sample messages were produced by the participants of the second
pretest. The coding was done electronically on an Excel spreadsheet. In order to further
systemize the coding process, the two coders first coded according to whether each posting
contained a statement serving the impression management goal and/or the support-receiving
goal. Each coder then highlighted the specific statements or passages reflecting those goals for
later comparison. This process facilitated identification and clarification of dissensus between the
two coders, especially about the goals reflected in the message, which was the key to message
sophistication. The coding protocol was then revised based on the discussion of disagreements,
ambiguities, and potential problems. Examples were added to the protocol to help the coders
better understand the definition of each category. After the coding protocol was finalized, intercoder reliabilities were checked using 62 randomly sampled messages (Lacy & Riffe, 1996).5
The present study used Krippendorff's alpha (KALPHA) coefficient to establish inter-coder
reliability, as this coefficient is particularly useful where the level of measurement of
classification data is higher than nominal or ordinal (Krippendorff, 1980). Hayes’ SPSS macro
was used to compute the KALPHA estimate for the level of message sophistication (Hayes &
Krippendorff, 2007). The reliability score of the message sophistication was .87.6
Among 176 support-seeking CMC messages, 9 (5.1%) messages were coded as
conveying no logic, 37 (21.0%) messages as conveying the expressive logic, 97 (55.1%)
messages as conveying the conventional logic and 33 (18.8%) messages as conveying the
rhetorical logic. Table 3 provides the numbers and percentages of the messages in each category.
5

The number of sample messages for reliability checking was calculated using the formula proposed by Lacy and
Riffe (1996): n = {(N-1)(SE)2 + PQN}/{(N-1)(SE)2+PQ} with the 5% confidence interval and 95% probability
level. The minimal agreement level was 85% and the assumed probability level (PQ) was 90%.
6
In order to systemize the main coding procedure, the two coders coded whether each posting contained a statement
serving the impression management goal (KALPHA = .86) and/or the support-receiving goal (KALPHA = .82),
prior to deciding the level of message sophistication.

51

Table 3. Content Analysis of Support-Seeking CMC Messages: Presence of Goal-Oriented
Content, Level of Message Sophistication, and Inter-Coder Reliability
Public Channel
Moderately
Severe

Private Channel

Mildly
Severe

Moderately
Severe

Mildly
Severe

Total

KALPHA
.82

Presence of Support-Oriented Content
Included

29 (64.4%)

5 (12.2%)

40 (90.9%)

26 (56.5%)

100 (56.8%)

Not included

16 (35.6%)

36 (87.8%)

4 (9.1%)

20 (43.5%)

76 (43.2%)

Presence of Impression-Oriented Content
Included

28 (62.2%)

26 (63.4%)

4 (9.1%)

12 (26.1%)

106 (60.2%)

Not included

17 (37.8%)

15 (36.6%)

40 (90.9%)

34 (73.9%)

70 (39.8%)

.86

Level of Message Sophistication
No logic

2 (4.4%)

4 (9.8%)

1 (2.3%)

2 (4.3%)

9 (5.1%)

Expressive

7 (15.6%)

12 (29.3%)

3 (6.8%)

15 (32.6%)

37 (21.0%)

Conventional

15 (33.3%)

23 (56.1)

37 (84.1)

22 (47.8%)

97 (55.1%)

Rhetorical

21 (46.7%)

2 (4.9%)

3 (6.8%)

7 (15.2%)

33 (18.8%)

45 (100%)

41 (100%)

44 (100%)

46 (100%)

176 (100%)

Total

Note: Values are number of cases with their percentages in the parentheses.

52

.87

Analytic Strategy
SPSS 22 (IBM Corp., 2013) was utilized for data analysis. Hypotheses 1–2 were tested
with two separate independent sample t-tests. In Hypothesis 1, the severity of the problem
(mildly severe vs. moderately severe) was a nominal independent variable, and the importance of
support-receiving goal was a continuous dependent variable. In Hypothesis 2, the publicity of the
message (public vs. private) was a nominal independent variable, and the importance of the
impression management goal was a continuous dependent variable. In the present study, Welch’s
formula of calculating t-values and degrees of freedom was utilized to account for group
differences in variances and cell sizes (Welch, 1947).7
Hypothesis 3 was tested using a two-way between-subjects Analysis of Variance
(ANOVA): the two manipulated factors, the problem severity and the publicity of the message,
were categorical independent variables, while CMC message composition time was a continuous
dependent variable. Levene’s F-test was conducted prior to running each t-test and ANOVA to
check the homogeneity assumption that the variance of the dependent variable should be equal
across groups (Levene, 1960).
Finally, Hypothesis 4 was tested with an ordinal regression analysis involving CMC
message composition time as an independent variable predicting the level of message
sophistication, which was treated as an ordinal variable. Before examining the ordinal regression
model, the proportional odds assumption was checked to ensure that CMC message composition
time would have an identical effect at each level of sophistication (Brant, 1990). A one-tailed

7

2

2

2

2 2

𝑆
𝑆
𝑆
𝑆
𝑊𝑒𝑙𝑐ℎ′ 𝑠 𝑡 = ̅̅̅
𝑋1 − ̅̅̅
𝑋2 / √ 1 + 2 , 𝑣 ≈ ( 1 + 2 ) /
𝑁1

𝑁2

𝑁1

𝑁2

𝑠14
𝑁12 𝑣1

+

𝑠24
𝑁22 𝑣2

53

significance test at the .05 level was used to test the relationships predicted in Hypotheses 1–4
except for Hypothesis 3 and 3-2, as they were non-directional hypotheses.

54

CHAPTER 4
RESULTS

Effect of Problem Severity on Importance of Support-Receiving Goal
Hypothesis 1 predicted that when seeking support on a social networking site, individuals
would perceive the support-receiving goal as more important when they were communicating
about a stressful event that is moderately severe, as opposed to one that is mildly severe. An
independent sample t-test was conducted to test the effect of the severity of the problem on the
perceived importance of the support-receiving goal. Levene’s F-test indicated a possibility that
the variances of the two groups were slightly different (F (1, 174) = 3.83, p = .05). With unequal
sample sizes and a possibility of unequal variances of the two groups, Welch’s unpaired t-test
was performed as this formula of defining t-value accounts for inequality of the two groups in
their variances and sample sizes (Welch, 1947). As shown in Table 4, Hypothesis 1 is not
supported. At the .05 significance level, there was no significant group difference in the
perceived importance of the support-receiving goal (t (163) = 1.37, p = .09, one-tailed). Although
participants facing a moderately severe problem perceived greater importance of the supportreceiving goal than did those facing a mildly severe problem, as predicted, that difference was
not statistically significant at the .05 significance level.

55

Table 4. Means and Standard Deviations of Perceived Importance of Support-Receiving
Goal between Individuals Facing Moderately vs. Mildly Severe Problems
N

Mean (SD)a

Total

176

4.32 (1.21)

Moderately Severe

89

4.45 (1.06)

Mildly Severe

87

4.20 (1.34)

t-value (163) b

1.37

Notes: a. Values are mean and standard deviations of the perceived importance of the supportreceiving goal. b. Welch’s t-value. *p < .05, **p < .01, ***p < .001, one-tailed.

56

Effect of Message Publicity on Importance of Impression Management Goal
Hypothesis 2 was proposed to test whether individuals would perceive the impression
management goal as more important when they solicit emotional support through a public SNS
channel as opposed to a private SNS channel. Since Levene’s F-test suggested a possibility of
slight violation of the homogeneity assumption (F (1, 174) = 3.76, p = .05), a Welch’s unpaired
t-test was again conducted to test the effect of message publicity on the importance of the
impression management goal (Welch, 1947).
As shown in Table 5, Hypothesis 2 is supported. There was a significant difference in the
perceived importance of the support-receiving goal between individuals in the public sharing
condition and the private-sharing condition, such that people perceived the impression
management goal as more important when they were sharing the negative story using the public
as opposed to the private SNS channel (public: M = 4.50, SD = 1.27; private: M = 3.01; SD =
1.47; t (172) = 7.21, p < .001, one-tailed).

Table 5. Means and Standard Deviations of Perceived Importance of Impression
Management Goal between Individuals Seeking Emotional Support through Public vs.
Private SNS Channels
N

Mean (SD)a

Total

176

3.74 (1.56)

Public Channel

86

4.50 (1.27)

Private Channel

90

3.01 (1.47)

t-value (172)

7.21***

Notes: a. Values are means and standard deviations of the perceived importance of the supportreceiving goal. b. Welch’s t-value. *p < .05, **p < .01, ***p < .001, one-tailed.

57

Effect of Problem Severity and Message Publicity on Message Composition Time
Hypothesis 3 predicted an interaction effect of the severity of the problem and the
publicity of the message on support-seeking CMC message composition time. More specifically,
it predicted that when using a public SNS channel to communicate a stressful event, individuals
would spend more time on CMC message composition when the problem is moderately severe as
opposed to mildly severe (H3-1). It was also predicted that when using a private SNS channel to
communicate a stressful event, CMC message composition time would not be influenced by the
severity of the problem (H3-2). A two-way ANOVA was conducted to test the interaction effect
of the problem severity and message publicity on CMC message composition time. The results
of using the original time and the log-transformed time spent on CMC message composition are
both presented in this section, as the original data showed a slight deviation from the normal
distribution. Levene’s F-test suggested no significant difference in variances of the four groups
(original time: F (3, 172) = .81, p = .50; log-transformed time: F (3, 172) = .90, p = .44). Because
of the unequal cell sizes, Type III sums of squares were used to calculate the F-values in the twoway ANOVA (Keppel & Wickens, 2004). Since this inequality occurred randomly rather than
intentionally, the result was very unlikely to be biased by the differences in the sample sizes of
the four groups.
As shown in Table 6, there was a significant interaction between the problem severity
and the message publicity in both the original and the transformed data (with original time: F (1,
172) = 4.31, p <. 05, partial η2 = .02; with log-transformed time: F (1, 172) = 4.99, p < .05,
partial η2 = .03). When using a public SNS channel to communicate about a stressful event,
participants spent significantly more time on CMC message composition when the problem was
moderately severe (M = 142.45, SD = 59.77) as opposed to mildly severe (M = 89.27, SD =

58

51.89). A pairwise comparison was made with a Bonferroni adjustment for multiple
comparisons, and the result suggested the significance of such difference (p < .001 for both the
original and the log-transformed time). Thus, Hypothesis 3-1 is supported.
When using a private SNS channel to communicate a stressful event, participants spent
more time on CMC message composition when the problem was moderately severe (M =108.48,
SD = 52.67) as opposed to mildly severe (M = 88.12, SD = 44.24). However, a pairwise
comparison with a Bonferroni p-value adjustment suggested a non-significant difference between
the two groups (with original time: p = .10; with log-transformed time: p = .14). Thus,
Hypothesis 3-2 was supported. Figure 1 shows the pattern of interaction.

Table 6. Main and Interaction Effects of Problem Severity and Message Publicity on Time
Spent on CMC Message Composition
Original Time Spent on
CMC Message Composition

Problem Severity
Message Publicity
Problem Severity ×
Message Publicity
Error

F

Partial 2

Log-Transformed Time Spent on
CMC Message Composition
MS

MS

1
1

59376.43
13544.43

21.63***
4.93*

.11
.03

1.05
.15

22.65***
3.21

.12
.02

1

11820.84

4.31*

.02

.23

4.99*

.03

172

2745.67

.05

*

p < .05, **p < .01, ***p < .001, two-tailed.

59

F

Partial 2

df

Table 7. Means and Standard Deviations of Time Spent on CMC Message Composition
across Four Experimental Conditions
Message Publicity

Problem Severity

n

Mean (SD)a

Public

Total

86

117.10 (61.89)

Moderately Severe

45

142.45 (59.77)

Mildly Severe

41

89.27 (51.89)

Total

90

98.07 (49.34)

Moderately Severe

44

108.48 (52.67)

Mildly Severe

46

88.12 (44.24)

Total

176

107.37 (56.48)

Moderately Severe

89

126.66 (58.60)

Mildly Severe

87

88.66 (47.71)

Private

Total

Note: a. Values are means and standard deviations for the time spent on message composition.

Public

Private

160
140
120
100
80
60

Mildly Severe

Moderately Severe

Figure 1. Main and Interaction Effects of Problem Severity and Message Publicity on Time
Spent on CMC Message Composition8

8

The Y-axis is time spent on CMC message composition in seconds whereas the X-axis is perceived severity of the
problem. The separate lines are publicity of the support-seeking CMC message.

60

Effect of Message Composition Time on Sophistication of Support-Seeking CMC Message
Hypothesis 4 proposed a positive relationship between message composition time and the
level of sophistication in the support-seeking CMC message. Table 8 shows correlation
coefficients among the key variables including the time spent on CMC message composition and
the level of sophistication in the support-seeking message. An ordinal regression model was
estimated to test the relationship between CMC message composition time and the level of
message sophistication. The proportional odds assumption was not violated (Brant, 1990), as the
slope coefficient was not significantly different across levels of sophistication (with original
time: χ2 (2) = .34, p = .56; with log-transformed time: χ2 (2) = .69, p = .71).
The result of the ordinal regression analysis is shown in Table 9, which indicates a
significant relationship between CMC message composition time and the level of sophistication
in the message. More specifically, with a one-second increase in CMC message composition
time, there would be a .02 increase in the ordered log odds of the message having a higher level
of sophistication (with original time: OR = 1.02, p < .001, McFadden’s Pseudo r2 = .10; with
log-transformed time: OR = 107.07, p < .001, McFadden’s Pseudo r2 = .12). Hypothesis 4,
therefore, is supported.

61

Table 8. Correlations among Key Variables
1

2

3

4

5

1.00
1.00
.46***

1.00
.46***

1.00

1.00
.95***
.42***

1.00
.47***

1.00

Spearman’s Nonparametric Correlations
1. SS Goal

1.00

2. IM Goal
3. Time
4. Log-transformed Time
5. Sophistication

-.01
.11
.11
.11

1.00
.17*
.17*
.05

Pearson’s Zero-Order Correlations
1. SS Goal
1.00
2. IM Goal

.03

1.00
*

3. Time
4. Log-transformed Time
5. Sophistication

.19*
.20**
.05

.15
.20**
.13*

Notes: Listwise deleted, N = 176. 1. Importance of Support-Receiving Goal; 2. Importance of
Impression Management Goal; 3. Time Spent on CMC Message Composition; 4. LogTransformed Time spent on CMC Message Composition; 5. Level of Message Sophistication.
*
p < .05, **p < .01, ***p < .001, one-tailed.

Table 9. Ordinal Regression Analyses Examining Effect of Time Spent on CMC Message
Composition on Sophistication of Support-Seeking CMC Message
Predictors
Time
Log-Transformed Time

B

S.E.

Wald χ2

df

OR [95% CI]

.02

.003

31.37

1

1.02 [1.01, 1.02]***

4.67

.74

40.46

1

107.07 [25.37, 451.91]***

Notes: Effects of the original and the log-transformed time are estimated using separate ordinal
regression models. *p < .05, **p < .01, ***p < .001, one-tailed.

62

CHAPTER 5
DISCUSSION

The present study investigates whether people pursue multiple goals when they seek
emotional support on social network sites, and whether multiple goal considerations shape their
support-seeking messages. More specifically, the study tests whether the severity of a stressful
situation motivates people to more strongly pursue the goal of receiving support, and whether the
expected publicity of the support-seeking message further motivates people to pursue the
impression management goal in addition to the support-receiving goal. This study also examines
whether individuals put more time and effort into CMC message composition when they are
motivated to pursue multiple goals, and whether such effort leads to the production of
sophisticated CMC messages that respond to those goals.
Importance of Support-Receiving and Impression Management Goals
The first hypothesis predicted that people would more strongly pursue the goal of
receiving emotional support when they were confronting a moderately stressful situation as
opposed to a mildly stressful situation. The result of this study, however, failed to support this
prediction. Although participants in the moderately severe condition evaluated the scenario as
significantly more severe than did those in the mildly severe condition, participants’ perceptions
on the importance of the support-receiving goal did not differ between those in the moderately
stressful condition and the mildly stressful condition. This result seems to suggest that
confronting a more severe problem does not necessarily create a stronger motivation to solicit
emotional support. However, before concluding a non-significant relationship between the

63

severity of the problem and the importance of the support-receiving goal, an alternative
explanation needs further exploration.
First, there is a possibility that participants’ psychological reactions were somehow
different within the same stressful condition. People have different thresholds in their stress
resistance, as well as in their capabilities to cope with psychological difficulties (Vodenicharov,
Dyakova, & Shipkovenska, 2005). As the result of Levene’s F-test suggested, the variability of
the importance of the support-seeking goal was larger in the mildly severe condition than in the
moderately severe condition. This result suggests that for some individuals, getting a $25 ticket
was a very stressful situation that required emotional support whereas others did not feel stressed
by such an event. A significant and positive relationship between perceived severity and the
importance of the support-receiving goal (r = .28, p < .001) indirectly supports this explanation.
Since previous studies have consistently shown a significant relationship between problem
severity and motivation to seek or process emotional support (e.g., Bodie, 2013; Bodie et al.,
2011, 2012; Dunkel-Schetter et al., 1987; Flett et al., 1995; Holmstrom et al., 2013; Spendelow
& Jose, 2010; Wilcox & Birkel, 1983), future studies should measure and account for individual
differences in stress thresholds in a variety of stressful situations.
Given the non-significant relationship between problem severity and the importance of
the support-receiving goal, a post-hoc analysis was conducted to test another possible
explanation that the communication context (i.e., message publicity) had an impact on people’s
primary goal perceptions. The result showed that participants pursued the support-receiving goal
more strongly when they were asked to seek emotional support using a private SNS channel (M
= 4.77, SD = .83) as opposed to a public SNS channel (M = 3.85, SD = 1.36), t (174) = 5.45, p <
.001.

64

The present study deliberately assigned participants to either the public or the private
sharing condition. Thus, being consciously aware of the audience of the support-seeking message
could have shaped the importance of the primary goal for participants. Since this study did not
provide participants with an opportunity to select the channel that they would use to share their
stressful story, it is possible that participants’ communication goals were already determined
based on the channel and audience assigned to their conditions. Previous studies on
communication channel selection have suggested that channel preferences change as people
pursue different communication goals (Feaster, 2010; O’Sullivan, 2000). The functional
disclosure framework also posits that people pursue different strategies when they self-disclose
on social network sites depending on media affordances (Bazarova & Choi, 2014). In the present
study, participants’ perceptions of the media affordances of private messages on Facebook, such
as intimate and personal communication, could have shaped their communication goals.
Future studies should consider testing if the severity of the problem or other features of a
stressful situation influence people’s selection of a particular SNS channel or preference for a
specific media affordance to share negative stories with others. Future studies might also
consider assigning people into multiple communication contexts and examine how supportrelated perceptions and communication goals change across different communication channels
and audiences.
Following the theoretical framework of multiple goals (Schrader & Dillard, 1998), the
functional model of self-disclosure (Bazarova & Choi, 2014), and the idea that support-seeking
behavior is inherently face-threatening (Goldsmith, 1992), the present study predicted that the
communicative context would activate a particular secondary communicative goal; that is, the
expected publicity of a support-seeking message would prompt people with an impression

65

management concern and make them behave in accordance with the impression management
goal. The result supported this prediction. Participants pursued the impression management goal
more strongly when they shared a stressful story with their SNS friends using the public SNS
channel (i.e., Facebook status update) as opposed to the private SNS channel (i.e., Facebook
private message).
Previous studies on self-disclosure and communication goals suggest that the publicity of
the CMC message motivates users to pursue an impression management goal (Leary &
Kowalski, 1990) and such motivation also exists when people "broadcast” personal information
on social network sites (Bazarova et al., 2013; Bazarova & Choi, 2014). The findings of the
present study supports this idea, as participants in this study perceived the impression
management goal as more important when they expected greater publicity for their supportseeking message.
The present study was conducted in a communication context that can be considered
either public or private. However, certain other communication tools exist along the continuum
of the two extremes. For example, Facebook Groups allow users to form groups and post content
that can only be shared among group members (Facebook, 2014). By posting a supportive
message on selected Facebook Groups, SNS users can share a stressful situation with a larger
number of people than they could manage with private messages, while having a greater control
over the audience than they would have with status updates. Moreover, SNS channels are
different not only in their visibility but also in their directedness (Bazarova & Choi, 2014). For
example, every Facebook user owns a wall, and friends can post public messages that are
directed to the owner of the wall. Although communication on Facebook walls and status
updates are both public and visible to a large audience, they might serve to address different

66

goals due to their differences in directedness. Future studies would benefit from examining goals
of supportive communication on different SNS channels that contain a wide variety of
characteristics (e.g., visibility, directedness etc.) and exploring how affordances of other SNS
channels prompt users with other communication goals when seeking emotional support through
those channels.
Pursuit of Multiple Goals and Production of Support-Seeking CMC Message
The present study further examined whether the pursuit of multiple goals influences
people’s message production behavior. The hyperpersonal model of CMC suggests that CMC
can lead to more desirable interaction outcomes than face-to-face communication, partly due to
the asynchronous nature of CMC tools, which encourage careful reflection of goals during the
communication process (Duthler, 2006; Walther, 2007). When people are motivated to pursue a
certain goal through computer-mediated communication, they are more likely to utilize CMC
features to accomplish it (Duthler, 2006; Ellison et al., 2006; Gibbs et al., 2006). In the present
study, people spent more time on developing a support-seeking CMC message when they were
sharing a stressful story that was moderately severe as opposed to mildly severe. The difference
was dependent on the expected publicity of the message, such that the difference in the time
people spent on composing a support-seeking message was much greater when they used a
public SNS channel as opposed to a private SNS channel. This finding paralleled those in
previous research that highlighted the role of motivation in computer-mediated communication,
in the sense that people put more effort into CMC message composition when they are motivated
to accomplish certain goals (Walther, 2007) such as looking attractive (Ellison et al., 2006;
Gibbs et al., 2006) or being polite (Duthler, 2006).

67

Moreover, this finding seems to support the idea of goal complexity (O’Keefe, 1988;
Samp & Solomon, 2005; Schrader, 1999), which suggests that pursuing multiple goals increases
complexity of goal structures, and individuals may embellish their messages to address multiple
goals simultaneously. When the problem they encounter is serious and when the publicity of the
message is high, people are more likely to develop a complex set of goal structures, and become
further motivated to carefully tailor their messages to these purposes. In the present study,
participants who shared a moderately severe problem using a public SNS channel spent a longer
time on developing a support-seeking CMC message, whereas those sharing a mild problem
through a private SNS channel spent less time on the task.
However, it is unclear whether such a result is directly attributable to the pursuit of
multiple goals, because the importance of the support-receiving goal was not significantly
greater when people were confronting the moderately severe problem as opposed to the mildly
severe problem. If the composition time discrepancy is not directly due to goal complexity, what
other factors would explain such discrepancy? One possible explanation might be that
participants not only developed perceptions of the importance of certain goals, but also
developed expectations on how likely or how easily they could achieve those goals. Goals do not
automatically motivate behaviors (Bandura, 2001); instead, people engage in an evaluative selfengagement process that is influenced by different goal characteristics, as well as judgments on
their own capabilities of engaging in goal-directed behaviors (Bandura, 1986, 2001). In
communication research, the concept of goal challenge—a judgment on how challenging the
goal one wants to accomplish is (Samp & Solomon, 2005)—has also been found to influence
people’s communication behavior. This influence appears in addition to that of goal complexity,

68

such that people facing a greater goal challenge tend to expend more efforts on communication,
reflected by using more clauses and repetitions of information (Samp & Solomon, 1999, 2005).
In the present study, participants evaluated receiving emotional support as more
important when they were sharing a negative story privately with one of their SNS friends than
when they were posting their story using a public channel that addresses a larger audience.
However, it is also likely that they perceived the support-receiving goal as easier to attain in
private discussion than in a public forum, such as sharing the story with all their SNS friends at
once. Public communication on social network sites is often unpredictable, and audiences of
public SNS messages are often “ill-defined,” and therefore “ill-intended,” as the nature of the
technology makes the presence of audiences invisible and the postings persistent (Bazarova et
al., 2013; boyd, 2008). Therefore, it is likely that people expected a greater degree of challenge
or uncertainty about accomplishing the support-receiving goal when they were asked to share the
stressful story through the public status update, and therefore put more time and effort into their
message composition in considering how to elicit a supportive response from others in their SNS
network.
In the present study, the number of SNS friends did not influence the amount of time
spent on message composition among those in the public-sharing condition, nor did it alter the
relationships tested in Hypothesis 3.9 This result also supports the claim above that audiences of
public SNS messages often go beyond one’s circle of friends, and therefore one’s perception of
the audience is not bounded by the number of SNS friends. Although such an interpretation is

9

In the public sharing condition, there was no significant relationship between the number of Facebook friends and
time spent on message composition (With original values: r = -.17, p =.11; With transformed values: r = -.15, p
= .17). Moreover, the number of Facebook friends did not emerge as a significant covariate in the two-way
ANOVA testing Hypothesis 3 (With original values: F (1, 171) = 1.31, p = .25; With transformed values: F (1,
171) = 1.87, p = .17).

69

theoretically valid and indirectly supported in this study, future investigation is warranted to
empirically and directly test this speculation.
The present study used time as a proxy for the amount of effort people put into CMC
message composition. The level of effort or effort intensity can be defined as the amount of
resources a person expends to accomplish a goal (Porter & Lawler, 1968; Yeo & Neal, 2004).
Although this approach has been widely utilized in psychological studies (e.g., Blau, 1993;
Brown & Peterson, 1994; Rieh, Kim, & Market, 2012), it has also been criticized for its
limitations and inability to capture the actual amount of mental effort a person allocates to the
task (Fisher & Ford, 1998). Effort is difficult to measure because it is an “invisible, internal,
hypothetical construct that is not directly observable” (Yeo & Neal, 2004, p. 231). Moreover,
effort can further be explained in terms of its persistency, quality, and direction (Blau, 1993; Yeo
& Neal, 2004), and time might be a better indicator of duration than it is of quality or direction of
effort.
Moreover, in the present study, people who spent more time on message production were
more likely to produce longer messages. The correlation between time spent on message
composition and the number of words was positive and statistically significant (with original
time: r = .46, p < .001; with log-transformed time: r = .45, p < .001). This result seems to suggest
that people simply spent more time in order to produce a longer message but not necessarily
invested more effort into it. However, it is also found that the time per word was much greater
when people used a public SNS channel as opposed to a private SNS channel (public: M = 4.75,
SD = 3.05; private: M = 2.60, SD = 1.59, t (174) = 5.88, p < .001). This result implies that people
spent more time selecting a word as the publicity of the message increased. For an accurate
understanding of the effort put into message production, future studies should integrate different

70

approaches for operationalization (e.g., quality, orientation, direction etc.) and measurement
(e.g., self-report measure, observation, or other physiological measures to capture cognitive
functioning) of effort and examine how they change as people seek support in different
communication environments.
It should also be noted that the relationship between the severity of the problem and the
message composition time in the private-sharing condition (H3-2) could have approached
statistical significance at a .05 significance level if a directional hypothesis was used to test the
difference. The multiple goal framework originally suggested that the primary goals vary in their
strength as well as attractiveness (Dillard et al., 1989). Later studies further revealed that the
importance a person attaches to achieving a single goal influences the characteristics of the
message such as the focus or the embellishment in the message (Samp & Solomon, 1998, 1999,
2005). The present study also found that individuals who faced a moderately severe problem
spent a greater time than did those who confronted a mildly severe problem regardless of the
publicity of the message (moderately severe condition: M = 126.66, SD = 58.60; mildly severe
condition: M = 88.66, SD = 47.71, F (1, 172) = 21.63, p < .001, η2= .11, See Table 6 and Table 7
for more details). A significant correlation between problem severity and time spent on CMC
message composition also supports the positive relationship between the problem severity and
the amount of effort individuals put into CMC message composition (with original time: r = .15
p < . 05; with log-transformed time: r = .20, p < .01, See Table 8). The magnitude of this
relationship, however, was much greater among individuals in the public-sharing condition as
opposed to the private-sharing condition. This finding suggests that the importance of the
primary goal might be sufficient to influence communication behavior, but the strength of such
influence can be affected by the context of communication and the importance of other goals.

71

Sophistication of Supportive-Seeking CMC Message
Finally, this study found that spending more time on composing a message increased the
likelihood of producing a more sophisticated support-seeking CMC message that conveyed
multiple goals simultaneously. CMC technologies provide users with opportunities to carefully
tailor messages to their purposes and accomplish desirable forms of communication between
users (Walther, 1996, 2007). The findings from the present study also suggest that SNS users can
develop more sophisticated messages that effectively accomplish their goals, not only by
addressing their primary goal but also by responding to situational demands. Although
participants were instructed to create a SNS message that would activate supportive responses
from their SNS friends, their messages reflected their impression management concerns by
highlighting positive aspects of their own behavior or attributing the cause of the problem to
external factors.
What do the findings from this study reveal about supportive communication occurring
on social network sites? Previous studies on supportive communication have focused on support
providers (e.g., Jones & Burleson, 1997; Jung, 1988; MacGeorge, 2001; MacGeorge et al., 2002;
Macrae & Johnston, 1998; Otten et al., 1988) and qualities of supportive messages (e.g.,
Burleson, 2003; Burleson et al., 2005; High & Dillard, 2012; Jones, 2004; Jones & Wirtz, 2006;
Lehman et al., 1986; Weiner, 2000). Support-seekers’ communication behavior, however, has
been a relatively understudied area. Although scholars have investigated various aspects of
support providers and support-giving messages that lead to desirable support outcomes, they
have overlooked the fact that support seekers can themselves shape the dynamics of supportive
interaction by changing ways of representing their stressors.

72

The present study provides a new insight into existing lines of research on supportive
communication. The finding suggests that support seekers can also influence the quality of
supportive interaction by framing their problem differently, which would determine how the
problematic situation is perceived by the support providers. The same stressors can be framed
differently, depending on the communication context, and through balancing and prioritizing
different communication goals arising from the interaction between the nature of the problem
and the context of communication.
One important factor that people consider when providing support to distressed others is
perceived responsibility for the problem, such that people are more likely to comfort others in
stress when they perceive the problem as being caused by factors that are external to the supportseeker (Jones & Burleson, 1997; Jung, 1988; MacGeorge, 2001; Otten et al., 1988). The present
study, however, found that support seekers can also frame their story to minimize representation
of their own responsibility for the problem, by selectively presenting information and attributing
causes to external factors. For example, when publicly sharing a story about getting a parking
ticket, participants tended to omit the fact that they failed to arrive in the parking garage before
the ticketing time, but instead emphasized how the towing company, the police, and the
university were greedy and coldhearted. Participants also rationalized their feelings or
dramatized their problem to activate empathic reactions from their SNS friends. Although they
did not misrepresent their problem, participants selectively presented factual information
surrounding the problematic situation. Intentional omission of and emphasis on certain pieces of
information are not uncommon in everyday conversation, but may influence the perceptions of
the support providers and their intentions to comfort the support seekers.

73

This finding offers implications for future research on supportive interaction. It suggests
that one must understand the motivations and goals of each interactant and examine how these
different goals are reflected in communication elements in order to understand the full picture of
supportive communication. Although the present study only examined the message production
behavior of support seekers, future study would benefit from studying the motivations and
behaviors of both support seekers and support providers, and also by testing how representations
of the problems made by support seekers influences providers’ reactions to the situation and thus
the quality of the supportive interaction.
Opportunities and Constraints of Supportive Communication on Social Network Sites
One might question why support-seeking behavior on social network sites is unique, and
why investigating support-seeking messages on social network sites is theoretically important.
New communication technologies and affordances now available on various communication
platforms provide different functions, utilities, and conveniences to users. These new
technologies and affordances, however, do not “fundamentally alter the theoretical bounds of
human interaction; such interaction continues to be governed by basic human tendencies” (Yzer
& Southwell, 2008, p. 8). Rather, what we should seek to understand or learn from research on
new communication technologies is “their ability to provide new or previously rare contexts for
information expression and engagement” (p. 8).
What new opportunities and constraints do social network sites provide? Many SNS
scholars have advocated the benefits of online social networking associated with general usage
of social network sites, and believed in the unique contribution of online social networking in
promoting psychological well-being (Burke et al., 2010; Ellison et al., 2011; Oh et al., 2013;
2014; Pua, 2013; Steinfield et al., 2008; Valenzuela et al., 2009). The present study provides

74

empirical evidence suggesting the benefits of social network sites pertain to sharing everyday
stressors. Social network sites might provide users with instant opportunities to publicly share
their everyday hassles and mild-to-moderate problems with a large audience. Although daily
stressful events have a less significant impact upon individuals than negative major life events
(Brantley et al., 1987; Flett et al., 1995), they occur more frequently and therefore are often the
primary determinants of mood and psychological well-being (Blankstein & Flett, 1992;
DeLongis et al., 1982, 1988). Daily hassles and negative feelings are also less visible to others,
and therefore elicit less support from others, unless people actively express their need for support
(Flett et al., 1995).
Unlike traditional CMC support groups that are oriented to particular illnesses or
problems, people can share a wide range of daily stressful stories in a way that satisfies their
needs and concerns using social network sites. Broadcasting through social network sites also
allows users to share their small moments and minor stressors with a large and diverse audience,
which in the old days was only possible through mass media (i.e., appearing on national
television). Minor hassles and difficulties can be publicly communicated on social network sites
in a more casual way without overburdening the audience who might feel obligated to comfort
their distressed friend if they were communicating about it privately or face-to-face.
A diversity of audiences on social network sites, ranging from close family members to
loosely connected acquaintances, further provides support seekers with an opportunity to receive
emotional support that is more optimal than in CMC or face-to-face. According to Thoits (2012),
significant and distanced others both contribute to emotional sustenance of the distressed person.
Significant others such as family members and close friends are good sources of providing love,
caring, and concern. On the other hand, emotional support from distanced others is also effective

75

when it comes from those who had a similar experience as they are likely to express “empathic
understanding” and “validation of feelings” (Thoits, 2012, p. 153).
This opportunity to easily and casually reaching a large, diverse, and “ill-defined”
audience, on the other hand, can also constrain users’ support-seeking behavior. On social
network sites, the characteristics or the size of the audience is less likely to be certain as they are
no longer identifiable by space and time (Bazarova et al., 2013). Thus, SNS users are likely to
form expectations and perceptions of the audience based on the channel and affordance they
select for communication, and this also happens when people seek emotional support on social
network sites. In the present study, the number of SNS friends did not emerge as a critical
variable that shaped support-seeking behavior when the participants were using a public SNS
channel to seek emotional support. Instead, the context of public communication triggered
impression-relevant concerns and made the participants carefully produce a support-seeking
message. Unlike intimate contexts of sharing emotions, such as crying or yelling at a friend
about how stressed and upset one is, public representation of a stressful event is a more complex
task requiring more care, particularly in times of stress. This may lead people to reframe their
story, rationalize their feelings, and in turn, makes their message more “sophisticated,” which, in
turn, increases the likelihood of receiving encouragement and comfort they need to cope with
everyday stress.
New social technologies create new dynamics of communication not because they
theoretically alter human interaction and the motivation behind it, but because they shape how
communication goals are enacted and accomplished (Walther, 2007). The present study
employed the multiple goal framework to provide a new insight into the mechanism of supportseeking behavior on social network sites, especially in terms of how new opportunities and

76

constraints of social network sites shape support-seeking communication on those sites by
creating a dynamic between the primary and the secondary goals. The finding from this study is
expected to shed light on understanding supportive interaction on social network sites. It
suggests that support seekers are not just passive recipients who await comforts from goodhearted others. Instead, they are communicators whose communication goals and messages vary,
depending on the nature of the problem and the context of communication. Future research
should continue the effort to view the whole mechanism of supportive interaction on social
network sites by understanding communication goals of both support seekers and support
providers, and also by examining how affordances of social network sites complicate ways they
pursue multiple communication goals.
Limitations
Several limitations of the present study deserve attention. The first limitation involves the
characteristics of the participants recruited for the laboratory experiment. The present study was
conducted among the college student sample. The selection of the college student sample was
inevitable, as the scenario approach used in the present experiment required a narrow-range of
participants who would share similar experiences, and thereby show similar emotional reactions
(i.e., stress levels) to a certain problematic event. However, the finding suffers from the external
validity issue, as evidence suggests that findings from student and nonstudent samples are
sometimes different in both the direction and the size of effects (Gordon, Slade, & Schmitt,
1986; Henrich, Heine, & Norenzayan, 2010; Notani, 1998; Peterson, 2001; Sears, 1986). A study
on stress and psychological adjustment also found a significant difference between younger and
older adults in stress perception: older adults were more optimistic than younger adults, whereas
younger adults experienced greater stress and showed poorer psychological adjustment than did

77

older adults (Chang, 2002). A study also showed an age difference in the relationship pattern
among perceived psychological difficulties, social support, and well-being (Kafetsios &
Sideridis, 2006).
Moreover, students and nonstudents are different in their communication and intellectual
characteristics; college students are equipped with stronger cognitive abilities, especially those
related to performing tasks and answering questions, than nonstudent populations (Sears, 1986).
College education also influences the development of communication competence (Rubin,
Graham, & Mignerey, 1990). Since cognitive abilities are likely to influence skills or
competence related to supportive communication (Burleson, 2003; MacGeorge et al., 2004;
Samter, 2002), support-seeking messages analyzed in this study could be more sophisticated than
those written by nonstudents. Future study should replicate this study with nonstudent samples
with varying levels of age and education to enhance the external validity of the relationships
hypothesized in the present study.
Moreover, in the present study, males accounted for a larger proportion of participants
than did females (see Table 1). Previous studies have suggested some gender differences in the
way individuals react to and cope with stressors (e.g., Wang et al., 2007; Watson, Goh, &
Sawang, 2011), communicate support (e.g., Ashton & Fuehrer, 1993; Burleson et al., 2005;
Holmstrom et al., 2005), and disclose themselves on social network sites (e.g., Al-Saggaf &
Nielsen, 2014; Kuo, Tseng, Tseng, & Lin, 2013; Mehdizadeh, 2010). Although gender did not
emerge as a significant predictor of the main outcomes or a moderator of the proposed

78

relationships,10 it would be worthwhile in future studies to explore whether characteristics shared
by the same gender group influence support-seeking behaviors on social network sites.
Thirdly, the study employed a forced experimental design by creating a hypothetical
stressful situation and forcing participants to use either a public or private SNS channel to solicit
emotional support. Although the manipulation was successful in engendering differential
affective states among participants, it should be noted that people were still responding to a
fictitious situation, and therefore the findings of the present study are again affected by issues
related to external validity. Moreover, because the scenario approach relies on participants’
imaginary reactions to the hypothetical situation, creating the same level of stress in all
participants is unlikely, as the same event could create different levels of stress across
individuals (Alspaugh, Stephens, Townsend, Zarit, & Greene, 1999). Future studies might
consider employing a direct measure of stress or physiological measures to more accurately
capture emotions and stress levels. They might also consider simulating or using real events that
upset people and generate stresses.
Stressors are different not only in their intensity but also in their nature (i.e.,
controllability, past experiences etc.) Originally, Cutrona and Russell (1990) identified several
characteristics of life events, including desirability, controllability, duration, and life domain
affected by the event. The present study focused on desirability of an event by manipulating the
negative consequences associated with the event and looked at its influence on goal perception
and support-seeking behavior. However, it is also likely that other characteristics of the stressful
10

There was no significant gender difference in the importance of the support-receiving goal (t (174) = 1.67, p
= .10) or the impression management goal (t (174) = .56, p = .58). Additionally, there was no main effect of
gender on time spent on CMC message composition (F (1, 168) = .02, p = .88) or interaction effects (with problem
severity: F (1, 168) = .04, p = .85; with message publicity: F (1, 168) = .73, p = .39; three-way interaction: F (1,
168) = .46, p = .50). Finally, levels of message sophistication were not different between males and females (χ2 (3)
= 1.11, p = .78).

79

event (i.e., controllability, duration or past experience, life domain) can influence coping
strategies and the types of support needed. For example, Cutrona and Russell’s (1990) study
found that individuals were more likely to seek comfort when they faced an uncontrollable and
negative situation. When the occurrence or consequence of a stressful situation is controllable,
individuals were more likely to pursue instrumental support to overcome the situation. Future
studies should consider replicating the present study in other stressful situations to reflect various
dimensions of stressors and see how they influence the pursuit of different coping strategies and
support-seeking behaviors.
Fourthly, the present study analyzed the content of support-seeking CMC messages based
on the goals reflected in the messages. Although the coding scheme was developed based on
previous studies that examined how different goals are reflected in communication (Caughlin et
al., 2008; Peterson & Albrecht, 1996; Scott et al., 2013), the intention or goal underlying each
message can only be speculated upon based on prior knowledge and assumptions. It should be
acknowledged that the same message can serve different goals depending on how
communicators employ it. For example, the coders in the present study treated minimizing one’s
own fault in the content of the message as reflecting an impression management goal, but it is
also likely that some participants were trying to receive empathic responses from their SNS
friends by portraying themselves as innocent victims. Future studies will benefit from
diversifying the qualitative approach to the experiment, and by exploring the true intentions
underlying different CMC messages and the outcomes people anticipate from them.
Finally, the findings from the present study should be discussed in relation to the
fundamental nature of human behavior, especially in times of stress. It should be noted that the
purpose of this study was not to preclude the automaticity of human behavior in response to

80

stressors. Humans often show uncontrolled emotional responses to stressors, and therefore
intentional expression of emotions is often difficult, as it lies beyond people’s conscious choice
(Wegner & Bargh, 1998; Zajonc, 1998). Having an extreme level of stress can impair the ability
to process information and regulate thoughts (Arnsten, 2009; Shansky & Lipps, 2013), which
further hinders one’s capability and motivation to carefully formulate a support-seeking message.
Goal-oriented perspectives view communication as inherently goal-driven (Berger, 2007;
Caughlin, 2010; Dillard et al., 1989; O’Keefe, 1988; Wilson, 2007), but they also acknowledge
that the communicator is not always conscious of the goal that the message pursues (Bargh &
Chartrand, 1999; Fitzsimons & Bargh, 2003; Kellermann, 1992). Instead, selection of
communication symbols often occurs as a result of unconscious consideration of goals and
automatic adjustment to situational constraints (Kellermann, 1992), which are defined in the
multiple goal framework as “secondary goals” (Dillard, 1990; Dillard et al., 1989). In the present
study, the impression management goal was the secondary goal that constrained people’s
support-seeking behavior on social network sites. However, it is beyond the scope of the present
study to determine whether such a process occurred at a conscious or unconscious level.
Additionally, the present study was conducted in the context of mild-to-moderate levels
of stress, and impression management concern was found to regulate support-seeking behaviors
to a certain extent. It is, however, unclear whether such a regulation process also prevails when
people are confronting and suffering from an extreme level of stress. Future studies should
explore the mechanisms through which stressors influence certain communicative behaviors, and
the roles of automatic and conscious processes in formulation of goals and messages in times of
stress. Future studies should also examine whether the communication context and different
affordances of social network sites trigger people to consciously or unconsciously pursue

81

particular goals, and whether their level of stress or other external factors influence the selfregulation process of balancing various needs.
The use of self-report measures should also be discussed in conjunction with the
nonconsciousness of goal formation. The present study measured the importance of two goals,
the support-receiving goal and the impression management goal, using self-report measures.
Although the validity and the reliability of the measures were rigorously tested, they are not
immune to self-report bias, as some goals are likely to be formed on an unconscious level. It
should be acknowledged that people’s judgments on selections of symbols and languages can
also occur without awareness of the sets of rules they pursue, because people often communicate
and behave according to automatic scripts embedded in their everyday routine (Kellermann,
1992). Participants could provide their best guesses on how and why they communicated in a
particular way or what goals they pursued, but those guesses might deviate from their true
intentions and goals (Kellermann, 1992). Furthermore, there might be other communication
goals, beyond the support-receiving and impression management goals, that participants pursued
while they were composing their support-soliciting messages. Future studies should consider
measuring various types of communication goals and exploring ways of measuring those goals
on an implicit or unconscious level.

82

Conclusion
Humans are social animals. People socialize and communicate with others to satisfy their
needs, and more than often the satisfaction of those needs requires actions of others (Kellermann,
1992). The present study was conducted to answer a simple question: How do people
communicate their needs for support on social network sites? The findings from this study
suggest that people develop messages that not only address their primary needs (i.e., receiving
comfort), but also respond to constraints arising from the context of communication (i.e.,
message publicity). Support-seekers invest more time and effort into CMC message production
when they are motivated to address multiple goals, which increases the likelihood that they will
develop more sophisticated messages that effectively reflect their needs for support and
impression management. Supportive communication on social network sites is unique in that it
provides a different combination of opportunity and constraint to users, but fundamentally, it
does not alter the role of supportive human communication and true motivation underlying such
communication. As discussed by Yzer and Southwell (2008) and further elaborated by Walther
(2014), studies on communicative behaviors on social network sites are meaningful and
theoretically interesting, not because they provide opportunities to develop new theories and
predictions, but because they offer new insights and perspectives on old ideas and theories about
the very nature of human beings.

83

APPENDICES

84

APPENDIX A. Distribution of Original, Log-Transformed, and Square Root-Transformed
Number of Facebook Friends and Time Spent on Message Composition

Figure 2. Probability-Probability Plots of Number of SNS Friends and Time Spent on
CMC Message Composition

85

APPENDIX B. Results of Neutral Stimuli Pretest

Table 10. Pretest of IAPS Stimuli: Means and Standard Deviations of Arousal Scores
IAPS#

N

M

SD

IAPS#

N

M

SD

p5000
p5020
p5030a
p5130
p5250
p5390
p5551

24
24
24
24
24
24
24

3.22
3.31
4.03
3.76
3.44
2.98
3.16

.95
1.16
.82
1.07
.91
.93
1.14

p5711
p5731
p5740
p5750
p5800
p5875
p7036a

24
24
24
24
24
24
24

3.06
3.67
3.22
3.68
3.03
3.63
4.43

.99
1.23
1.05
1.31
.93
1.08
.83

Note: a. Deleted in the main experiment.

86

APPENDIX C. Emotional State after Neutralization Treatment

Table 11. Means and Standard Deviations of SAM Scores after Neutralization Treatment
Public
Moderately
Severe

Private
Mildly
Severe

Moderately
Severe

Mildly
Severe

F

M

SD

M

SD

M

SD

M

SD

Emotional Valence

5.89

1.33

6.27

1.16

5.93

1.48

6.59

1.15

2.89*

Emotional Arousal

4.13

1.84

3.31

1.45

3.98

2.06

3.96

1.84

1.64

*

p < .05, *p < .01, ***p < .001, two-tailed.

87

APPENDIX D. Results of First Pretest

Table 12. First Pretest of Twelve Scenarios: Means and Standard Deviations of Perceived
Reality and Perceived Severity
Perceived Reality

Perceived Severity

Scenarios
Relationship Issue
Mild
Moderate
Roommate Issue
Mild
Moderate
Job Loss
Mild
Moderate
Car Troublea
Mild
Moderate
Academic Failure
Mild
Moderate
Scholarship Lossa
Mild
Moderate

a

N

M

SD

48
24
24
48
24
24
48
23
25
48
24
24
48
24
24
42
22
20

5.13
5.17
5.10
4.90
5.46
4.35
4.94
6.16
3.81
5.54
5.88
5.21
5.42
5.74
5.11
5.07
5.29
4.83

1.50
1.43
1.59
1.62
1.39
1.68
1.81
0.86
1.73
1.41
1.03
1.66
1.32
1.34
1.23
1.36
1.23
1.49

t
.16

2.50*

5.87***

1.67

1.68

-.49

N

M

SD

48
24
24
48
24
24
48
23
25
48
24
24
48
24
24
42
22
20

5.02
4.19
5.85
5.30
4.31
6.29
5.80
5.59
5.99
5.35
4.74
5.96
5.86
5.60
6.13
6.08
5.91
6.27

1.47
1.38
1.04
1.51
1.27
1.00
1.04
0.93
1.12
1.29
1.28
0.98
1.09
1.13
0.99
1.07
1.02
1.13

Note: a. Selected for the second pretest. *p < .05, *p < .01, ***p < .001, two-tailed.

88

t
-4.68***

-6.03***

-1.31

-3.71**

-1.72

-2.43*

APPENDIX E. Results of Second Pretest
Table 13. Second Pretest of Six Scenarios: Means and Standard Deviations of Perceived
Reality, Perceived Severity, and Likelihood of Seeking Support on Facebook
Scenarios
Relationship
Mild
Severe
Scholarship loss
Mild
Severe
Car trouble
Mild
Severe

Perceived Reality

Perceived Severity

M

SD

M

5.10
5.27
4.92
5.10
5.22
4.98
5.66
6.09
5.23

.99
1.43
1.02
1.25
1.36
1.44
1.32
1.20
1.77

t

1.27

1.05

3.25**

SD

4.75
4.25
5.26
5.44
5.11
5.77
5.10
4.61
5.41

1.00
1.44 -4.53***
0.84
0.84
1.11 -3.34**
0.89
0.99
1.18 -3.55**
1.08

Note: N = 31. *p < .05, *p < .01, ***p < .001, two-tailed.

89

t

Seeking Support on FB
M

SD

1.88
1.66
2.10
2.97
2.90
3.03
3.03
2.97
3.10

1.54
1.50
1.85
1.98
2.18
2.21
1.87
2.16
1.94

t
-1.78

-.38

-.42

APPENDIX F. Stressful Scenarios Used in Main Experiment

Mildly Severe Version
Imagine that you have driven to campus to meet with a group after dark. Since you may
park in any available space and it is raining, you decide to park in a university garage. After your
meeting ends, a friend asks if you want to go to a party. Since you don’t have any major
commitments the next day you agree, leaving your car parked in the garage. Later that night you
get a ride home with your friend, with your car still in the garage. Although you planned to get
your car before the 7 AM ticketing time, your alarm does not wake you up and you arrive at the
garage at 7:05 to find a police officer writing you a $25 parking ticket.
Moderately Severe Version
Imagine that you have driven to campus to meet with a group after dark. Since you may
park in any available space and it is raining, you decide to park in a university garage. After your
meeting ends, a friend asks if you want to go to a party. Since you don’t have any major
commitments the next day you agree, leaving your car parked in the garage. Later that night you
get a ride home with your friend, with your car still in the garage. Although you planned to get
your car before the 7 AM ticketing time, your alarm does not wake you up and you arrive at the
garage at 7:05 to find a tow truck driving away with your vehicle. To get your car back you will
have to pay the $80 towing fee, all of your unpaid parking tickets, and an additional fine for a
total of $350.

90

APPENDIX G. Results of Manipulation Check
Table 14. Means and Standard Deviations of Perceived Severity and SAM Scores after Reading Scenario
Public
Moderately
Severe

Perceived Severitya

Moderately
Severe

Total

Mildly
Severe

Total
Moderately
Severe

Total

Mildly
Severe

Total

5.52 (1.00)

4.34 (1.27) 4.95 (1.28)

5.52 (.89)

4.74 (1.21) 5.12 (1.13)

5.52 (.94)

4.55 (1.25) 5.04 (1.20)

b

1.80 (.97)

2.76 (1.26) 2.26 (1.21)

1.71 (.90)

2.48 (1.13) 2.11 (1.09)

1.76 (.93)

2.61 (1.19) 2.19 (1.15)

c

7.65 (1.73)

6.05 (1.99) 6.88 (2.01)

7.61 (1.65)

6.65 (1.74) 7.12 (1.75)

7.63 (1.68)

6.37 (1.87) 7.01 (1.88)

Emotional Valence
Emotional Arousal

Mildly
Severe

Private

Notes: Values are means and standard deviations in the parentheses. a. N = 175; b. N = 173; c. N = 176

Table 15. ANOVA Results of Perceived Severity and SAM Scores after Reading the Scenario
Perceived Severity
df
Problem Severity
Message Publicity
Severity × Publicity
Error

MS

1 41.82
1.76
1
1.66
1
171 1.21

Emotional Valence

Emotional Arousal

F

Partial 2

df

MS

F

Partial 2

df

MS

F

Partial2

34.46***
1.45
1.37

.17
.01
.01

1
1
1
172

32.16
1.48
.37
1.15

27.91***
1.29
.32

.14
.01
.002

1
1
1
172

71.79
3.60
4.42
3.16

22.74***
1.14
1.40

.12
.01
.01

*

p < .05, **p < .01, ***p < .001, two-tailed.

91

APPENDIX H. Codebook for Content-Analysis of Message Sophistication of Support-Seeking
CMC Messages

Introduction
This protocol is aimed at assessing sophistication of the support-seeking CMC messages composed
by participants. The level of message sophistication is operationalized as the extent to which the
message reflects a more sophisticated design logic.
General Rules
1. Coding occurs at the level of the message, based on a holistic assessment of its overall
characteristics.
2. Coding occurs based on the explicit content but not on the implicit intent implied in the
message.
3. Coding occurs based on the highest level of the logic appear in the message. (e.g., if a message
contains an expression of sadness and a support-seeking request, it must be coded as
conventional.)
4. Coding of “no logic” will occur only when the message contains the factual information about
the scenario without further interpretation or expression.

No Logic (0)
Operational Definition

Examples

Factual information about the scenario without further interpretation or
expression.
“Got tickets and car towed in a night....”
“just got my car towed. $350”
“Aibo I got up late this morning, so I received a parking ticket.”

Expressive Design Logic
General principles
Communication is a medium for expressing feelings.
Clarity, openness, honesty, unimpeded expression.
Goal is self-expression.
Operational definition
and features that
guarantee expressive
code

 Explicit blame, criticism, complaints, regrets without explanation of
contexts or rationale.
 Automatic or reflexive reaction to the situation.
 Explicit expressions of negative emotions or mental state without
explaining the cause.

Examples

“Dammit, I got my car towed.”
“Really, 5 minutes late and I have a ticket!!! >:(”
“OMGOMGOMG PACE … I JUST GOT A TICKET LITERALLY 5
MINUTES AFTER THE TIME WAS UP.”

92

Conventional Design Logic
General principles
Communication is a game played cooperatively by social rules.
Appropriateness, some control over messages to serve immediate goals.
Given the dominant goal (s) in this context, a message must address at
least one goal either explicitly or implicitly to qualify as conventional.
Operational definition
and features that
guarantee conventional
code

Statements addressing a support-receiving goal
 Direct request of support.
 Seeking for empathy by legitimizing one’s emotions or feelings.
 Expressions of seeker’s support-oriented expectations or needs.
Statements addressing an impression management goal
 Showing positive aspects of oneself (e.g., describing how one tries
to cope with the situation; how one is capable of dealing with the
situation; how good or nice one is).
 Minimizing the responsibility of oneself in causing the problem or
attributing the cause of the problem to external factors.
 Justifying one’s behavior.

Examples

Support-receiving goal
“…Could I be more damned? …anyone can tell me how to appeal for
this?”
“…80 bucks for the towing fee and 350 for the tickets…
#someonepayformyticketplease”
“…my baby got towed... I feel so upset...better talk to someone.”
“i'm so pissed. I got a prking ticket … can you believe this?! what
would you do?”
Impression management goal
“…I feel like we shouldn't be reprimanded for being smart and not
drinking and driving.”
“…How annoying is that? I was just trying to be safe and do the right
thing.”
“I thought drinking responsibly gets rewarded? I parked my car at
Com arts building so that i wouldnt drink and drive... of course i got
a ticket!”

Rhetorical Design Logic
General principles
Communication is the creation and negotiation of social selves
and situations.
Flexibility, symbolic sophistication, depth of interpretation
Communication addresses multiple goals.
Operational definition
and features that
guarantee rhetorical
code




Inclusion of statements serving both support-receiving and
impression management goals.
Redefinition or elaboration of the situation in a positive light.

93


Examples

In-depth interpretation of the situation.
“…I was asked to go out, and of course since I got everything done
I said yes…I guess "balancing" your fun and work is impossible
these days…”
“…Oh well, just a bad luck…my day will only get better from here.”
“… the university should be a little bit more lenient with
students..after all they're constantly saying we're the ones with no
money in our bank accounts..”

94

APPENDIX I. Consent Form, Survey, and Debriefing Form

CONSENT FORM
You are being asked to participate in a research study conducted to learn about how people seek social
support on social network sites. You have been selected as a possible participant in this study as (a) you
are an adult aged between 18 and 29, and (b) you have an active Facebook account. From this study, the
researchers hope to learn adults’ support-seeking behavior on social network sites and what goals they
have when they post a stressful story on social network sites.
This lab study will take about 30 minutes. You will read a hypothetical scenario about a stressful event
and imagine that you are actually experiencing the stressful situation described in the scenario. You will
then be asked to post a support-seeking message on your Facebook. Either before or after composing a
message to be posted on your Facebook, you will be asked questions about goals and concerns you would
have when sharing the story with your Facebook friends. At the end of the study, you will report your
demographic information.
There is no potential risk of participating in this study. All the information you provide during study
participation will remain confidential to the maximum extent allowable by law. Data will be stored in a
password-protected computer at Michigan State University for at least three years. No one other than the
researchers, research staff, and the MSU IRB will have access to the data. No identifying information will
be stored with the data. Additionally, your email address will be kept separately from the main dataset
and will not be linked with the data. The results of this study may be published or presented at
professional meetings, but identities of the participants will remain anonymous.
Your participation in this study is voluntary. Refusal to participate will have no effect on your grade or
evaluation. You may choose not to answer specific questions or to stop participation at any time without
any consequences.
You will receive 1 research credits through the SONA system for your participation in this study per the
agreement with your professor/instructor. If you do not want to participate in the research study but still
wish to receive extra-course credit, you will be given the opportunity to complete an alternative
assignment. The alternative assignment will take about the same time and effort as does the survey.
If you have any concerns or questions about this study, such as scientific issues, how to do any part of it,
or to report an injury, please contact Hyun Jung Oh (ohhyun4@msu.edu), Department of Advertising +
Public Relations, Michigan State University, 309 Communication Arts and Sciences, East Lansing, MI
48824.
If you have questions or concerns about your role and rights as a research participant, would like to obtain
information or offer input, or would like to register a complaint about this study, you may contact the
Michigan State University’s Human Research Protection Program at (517) 355-2180, Fax (517) 4324503, email irb@msu.edu, or regular mail at 207 Olds Hall, MSU, East Lansing, MI 48824. Any contact
you make may be done so anonymously.
By clicking the next button, you are voluntarily agreeing to participate in this research study.

95

SQ. Do you have a Facebook account?
a. Yes
b. No (end the survey)

Part 1. Neutralization Treatment
Before starting the main experiment, we will show you some pictures. Please try to concentrate on the
images while you view the pictures.
[NOTE: Participants will view six random neutral pictures.]
Now, you will be answering some questions asking your current emotional state.
Q. Please select an image that best reflects your current emotional valence.

Figure 3. SAM Scale – Emotional Valence
Q. Please select an image that best reflects your current emotional state.

Figure 4. SAM Scale – Emotional Arousal

96

Part 2. Manipulation
Now, you will read a hypothetical scenario depicting a stressful situation. Please read the scenario as
thoroughly as possible and imagine that the scenario is actually happening to you right now.
[NOTE: Participants will read one of the following scenarios depending on the condition assigned.]
[Mildly Severe Condition] Imagine that you have driven to campus to meet with a group after dark. Since
you may park in any available space and it is raining, you decide to park in a university garage. After your
meeting ends a friend asks if you want to go to a party. Since you don’t have any major commitments the
next day you agree, leaving your car parked in the garage. Later that night you get a ride home with your
friend, with your car still in the garage. Although you planned to get your car before the 7 AM ticketing
time, your alarm does not wake you up and you arrive at the garage at 7:05 to find a police officer writing
you a $25 parking ticket.
[Moderately Severe Condition] Imagine that you have driven to campus to meet with a group after dark.
Since you may park in any available space and it is raining, you decide to park in a university garage.
After your meeting ends a friend asks if you want to go to a party. Since you don’t have any major
commitments the next day you agree, leaving your car parked in the garage. Later that night you get a ride
home with your friend, with your car still in the garage. Although you planned to get your car before the 7
AM ticketing time, your alarm does not wake you up and you arrive at the garage at 7:05 to find a tow
truck driving away with your vehicle. To get your car back you will have to pay the $80 towing fee, all of
your unpaid parking tickets, and an additional fine for a total of $350.

Q. Please imagine that the situation you have just read is actually happening to you. What would you feel
if you actually experience this situation? Please select an image that best depicts your emotional valence.
[See Figure 3]
Q. Again, please select an image that best depicts your feeling if this situation actually happens to you.
[See Figure 4]

97

Part 3. Importance of Communication Goals
[NOTE: Part 3 and Part 4 will appear in a random order.]
[Public Sharing Condition]
Now, please tell us what goals or concerns you would have in mind when you post about this situation on
your Facebook Status Update. Indicate how much you agree or disagree with each of the following
statements.
[Private Sharing Condition]
Now, imagine that you are sending a Facebook Private Message to one of your Facebook friends about
the stressful situation to receive some comforting from him/her. Choose a Facebook friend that you would
share the story with. Provide the first name of the friend and his/her relationship to you below.
First name: ______________
Relationship: _____________

Q. Now, please tell us what goals or concerns you would have in mind when you share this situation
publicly with your Facebook friends via Facebook Status Update (privately with OOOO via Facebook
Private Message). Indicate how much you agree or disagree with each of the following statements.
Strongly
Disagree
It is very important to me to convince my Facebook friends (this
person) and activate supportive response after they (he or she)
read(s) my posting.
I am very concerned about getting supportive response with this
message.
I really care whether my Facebook friends (he or she) comfort(s)
me.
The support-receiving outcome of this message will have an
important consequence for me.
I really want my Facebook friends (this person) to provide me
with necessary support.
I am concerned with making (or maintaining) a good impression
in this message.
I will be careful to avoid saying things which are socially
inappropriate.
I will be very conscious of what is appropriate an inappropriate in
this message.
I am concerned with putting myself in a “bad light” in this
message.
I don’t want to look stupid in this message.

98

Strongly
Agree

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

○

Part 4. Message Composition
[Public Condition] Now, imagine that you are posting something about this situation on your Facebook
Status to seek some comforting from your Facebook friends. In the blank below, please compose the
exact message that you would post on your Facebook Status. You may spend as much as time you need
on this page.
[Private Condition] Now, imagine that you are sending a Facebook Private Message to [the name of the
friend appears here] about the stressful situation to receive some comforting from him/her. In the blank
below, please compose the exact message that you would send to your friend using Facebook Private
Message to get some comforting from him/her. You may spend as much as time you need on this page.
[NOTE: Time spent on this page will be recorded.]
Part 5. Manipulation Check
Q. This section will ask your personal thought on the scenario you read. Please click the circle that best
describes your thought. Refer to the examples below.
If the situation was not at all serious, you may check the circle that is closest to “Not at all serious”.
Not at all serious
●
○
○
○
○
○
○
Very serious
If you think the situation was moderately serious, you can check the box that is in the middle.
Not at all serious
○
○
○
●
○
○
○
Very serious

The situation I read is...
Not at all serious
Not at all severe
Not at all stressful
Not at all upsetting

○
○
○
○

○
○
○
○

○
○
○
○

○
○
○
○

○
○
○
○

Part 6. About Me
Q. In what year were you born? _____________
Q. What is your sex?
Male
Female
Refuse to answer
Q. How many Facebook friends do you have? ________________

99

○
○
○
○

○
○
○
○

Very serious
Very severe
Very stressful
Very upsetting

DEBRIEFING FORM
Thank you for taking part in the study. The purpose of this form is to give you more information about the
research study and its goal.
For this study, we are interested in how people share a stressful story on Facebook especially, whether
they consider different goals and such consideration influences their message production. We are also
interested in whether the severity of the problematic situation and the communication channel (public vs.
private) influence perceived goal importance and message production.
Please do not share any information about this study with anyone who could potentially participate in the
study. Sharing even the littlest detail could have negative impacts on the accuracy of our research
findings.
If you have any questions about the study and its results, we encourage you to ask. Also, since you are
now given additional information about the study that you were unaware during the initial consent
process, you now have the option to withdraw consent to use your data. Please let us know, if you wish to
withdraw your consent. You may contact the investigators, Hyun Jung Oh (ohhyun4@msu.edu) and
Robert LaRose (larose@msu.edu).
Thank you very much for your participation.
Best wishes,

100

BIBLIOGRAPHY

101

BIBLIOGRAPHY

Albrecht, T. L., & Goldsmith, D. J. (2003). Social support, social networks, and health. In T. L.
Thompson, A. M. Dorsey, K. I. Miller & R. Parrott (Eds.), Handbook of health
communication (pp. 263-284). Mahwah, NJ: Erlbaum.
Al-Saggaf, Y., & Nielsen, S. (2014). Self-disclosure on Facebook among female users and its
relationship to feelings of loneliness. Computers in Human Behavior, 36, 460-468. doi:
10.1016/j.chb.2014.04.014
Alspaugh, M. E. L., Stephens, M. A. P., Townsend, A. L., Zarit, S. H., & Greene, R. (1999).
Longitudinal patterns of risk for depression in dementia caregivers: Objective and
subjective primary stress as predictors. Psychology and Aging, 14, 34-43. doi:
10.1037/0882-7974.14.1.34
Amichai-Hamburger, Y., Kingsbury, M., & Schneider, B. H. (2013). Friendship: An old concept
with a new meaning? Computers in Human Behavior, 29, 33-39. doi:
10.1016/j.chb.2012.05.025
Arnsten, A. T. (2009). Stress signaling pathways that impair prefrontal cortex structure and
function. Nature Reviews Neuroscience, 10, 410-422. doi: 10.1038/nrn2648
Ashton, W. A., & Fuehrer, A. (1993). Effects of gender and gender role identification of
participant and type of social support resource on support seeking. Sex Roles, 28, 461476. doi: 10.1007/BF00289608
Ballantine, P. W., & Stephenson, R. J. (2011). Help me, I’m fat! Social support in online weight
loss networks. Journal of Consumer Behaviour, 10, 332-337. doi: 10.1002/cb.374
Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of
Psychology, 52, 1-26. doi: 10.1146/annurev.psych.52.1.1
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84, 191-215. doi: 10.1037/0033-295X.84.2.191
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice Hall.
Barbee, A. P., & Cunningham, M. R. (1995). An experimental approach to social support
communications: Interactive coping in close relationships. Communication Yearbook, 18,
381-413.
Barbee, A. P., Gulley, M. R., & Cunningham, M. R. (1990). Support seeking in personal
relationships. Journal of Social and Personal Relationships, 7, 531-540. doi:
10.1111/j.1540-4560.1993.tb01175.x

102

Barbee, A. P., Rowatt, T. L., & Cunningham, M. R. (1998). When a friend is in need: Feelings
about seeking, giving, and receiving social support. In P. A. Andersen & L. K. Guerrero
(Eds.), Handbook of communication and emotion: Research, theory, applications, and
contexts (pp. 281-301). San Diego, CA: Academic.
Bargh, J. A., & Chartrand, T. L. (1999). The unbearable automaticity of being. American
Psychologist, 54, 462-479. doi: 10.1037/0003-066X.54.7.462
Barrett, L. F., Gross, J., Christensen, T. C., & Benvenuto, M. (2001). Knowing what you’re
feeling and knowing what to do about it: Mapping the relation between emotion
differentiation and emotion regulation. Cognition & Emotion, 15, 713-724. doi:
10.1080/02699930143000239
Bazarova, N. N. (2012). Public intimacy: Disclosure interpretation and social judgments on
Facebook. Journal of Communication, 62, 815-832. doi: 10.1111/j.14602466.2012.01664.x
Bazarova, N. N., & Choi, Y. H. (2014). Self‐disclosure in social media: Extending the functional
approach to disclosure motivations and characteristics on social network sites. Journal
of Communication, 64, 635-657. doi: 10.1111/jcom.12106
Bazarova, N. N., Taft, J. G., Choi, Y. H., & Cosley, D. (2013). Managing impressions and
relationships on Facebook: Self-presentational and relational concerns revealed through
the analysis of language style. Journal of Language and Social Psychology, 32, 121-141.
doi: 10.1177/0261927X12456384
Berger, C. R. (2007). Communication: A goal-directed, plan-guided process. In D. RoskosEwoldsen & J. Monahan (Eds.), Communication and social cognition: Theories and
methods (pp. 47-70). Mahwah, NJ: Erlbaum.
Blankstein, K. R., & Flett, G. L. (1992). Specificity in the assessment of daily hassles: Hassles,
locus of control, and adjustment in college students. Canadian Journal of Behavioural
Science, 24, 382-398. doi: 10.1037/h0078738
Blau, G. (1993). Operationalizing direction and level of effort and testing their relationships to
individual job performance. Organizational Behavior and Human Decision Processes,
55, 152-170. doi:10.1006/obhd.1993.1028
Bodie, G. D. (2013). The role of thinking in the comforting process: An empirical test of a dualprocess framework. Communication Research, 40, 533-558. doi:
10.1177/0093650211427030
Bodie, G. D., & Burleson, B. R. (2008). Explaining variations in the effects of supportive
messages: A dual-process framework. Communication Yearbook, 32, 355-398.
Bodie, G. D., Burleson, B. R., Gill-Rosier, J., McCullough, J. D., Holmstrom, A. J., Rack, J. J., .
. . Mincy, J. (2011). Explaining the impact of attachment style on evaluations of

103

supportive messages: A dual-process framework. Communication Research, 38, 228-247.
doi: 10.1177/0093650210362678
Bodie, G. D., Burleson, B. R., Holmstrom, A. J., McCullough, J. D., Rack, J. J., Hanasono, L.
K., & Rosier, J. G. (2011). Effects of cognitive complexity and emotional upset on
processing supportive messages: Two tests of a dual-process theory of supportive
communication outcomes. Human Communication Research, 37, 350-376. doi:
10.1111/j.1468-2958.2011.01405.x
Bodie, G. D., Burleson, B. R., & Jones, S. M. (2012). Explaining the relationships among
supportive message quality, evaluations, and outcomes: A dual-process approach.
Communication Monographs, 79, 1-22. doi: 10.1080/03637751.2011.646491
boyd, D. (2008). Taken out of context: American teen sociality in networked publics.
Unpublished doctoral dissertation, University of California, Berkeley, CA. Retrieved
from http://www.danah.org/papers/TakenOutOfContext.html
Bradley, M. M. and P. J. Lang (1994). Measuring emotion: The self-assessment manikin and the
semantic differential. Journal of Behavior Therapy and Experimental Psychiatry, 25,
49-59. doi: 10.1016/0005-7916(94)90063-9
Brant, R. (1990). Assessing proportionality in the proportional odds model for ordinal logistic
regression. Biometrics, 46, 1171-1178.
Brantley, P. J., Waggoner, C. D., Jones, G. N., & Rappaport, N. B. (1987). A daily stress
inventory: Development, reliability, and validity. Journal of Behavioral Medicine, 10,
61-73. Doi: 10.1007/BF00845128
Brashers, D. E., Neidig, J. L., & Goldsmith, D. J. (2004). Social support and the management of
uncertainty for people living with HIV or AIDS. Health Communication, 16, 305-331.
doi: 10.1207/S15327027HC1603_3
Brewin, C. R., Andrews, B., & Valentine, J. D. (2000). Meta-analysis of risk factors for
posttraumatic stress disorder in trauma-exposed adults. Journal of Consulting and
Clinical Psychology, 68, 748-766. doi: 10.1037/0022-006X.68.5.748
Brown, P. (1987). Politeness: Some universals in language usage. Cambridge, UK: Cambridge
University Press.
Brown, S. P., & Peterson, R. A. (1994). The effect of effort on sales performance and job
satisfaction. The Journal of Marketing, 58, 70-80. doi: 10.2307/1252270
Bryant, E. M., & Marmo, J. (2012). The rules of Facebook friendship: A two-stage examination
of interaction rules in close, casual, and acquaintance friendships. Journal of Social and
Personal Relationships, 29, 1013-1035. doi: 10.1177/0265407512443616
Burke, M., Kraut, R., & Marlow, C. (2011). Social capital on Facebook: Differentiating uses
and users. Paper presented at the CHI, Vancouver, BC, Canada.
104

Burke, M., Marlow, C., & Lento, T. (2010). Social network activity and social well-being.
Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp.
1909-1912). New York, NY: ACM. doi: 10.1145/1753326.1753613
Burleson, B. R. (1983). Social cognition, empathic motivation, and adults’ comforting strategies.
Human Communication Research, 10, 295-304. doi: 10.1111/j.14682958.1983.tb00019.x
Burleson, B. R. (2003). Emotional support skill. In J. O. Greene & B. R. Burleson (Eds.),
Handbook of communication and social interaction skills (pp. 551-594). Mahwah, NJ:
Lawrence Erlbaum Associates.
Burleson, B. R. (2009). Understanding the outcomes of supportive communication: A dualprocess approach. Journal of Social and Personal Relationships, 26, 21-38. doi:
10.1177/0265407509105519
Burleson, B. R., & Goldsmith, D. J. (1998). How the comforting process works: Alleviating
emotional distress through conversationally induced reappraisals. In P. A. Anderson & L.
K. Guerrero (Eds.), Handbook of communication and emotion: Research, theory,
applications, and contexts (pp. 245-280). San Diego, CA: Academic Press.
Burleson, B. R., Holmstrom, A. J., & Gilstrap, C. M. (2005). “Guys can’t say that to guys”: Four
experiments assessing the normative motivation account for deficiencies in the emotional
support provided by men. Communication Monographs, 72, 468-501. doi:
10.1080/03637750500322636
Burleson, B. R., & Samter, W. (1985). Consistencies in theoretical and naive evaluations of
comforting messages. Communications Monographs, 52, 103-123. doi:
10.1080/03637758509376099
Caughlin, J. P. (2010). A multiple goals theory of personal relationships: Conceptual integration
and program overview. Journal of Social and Personal Relationships, 27, 824-848. doi:
10.1177/0265407510373262
Caughlin, J. P., Brashers, D. E., Ramey, M. E., Kosenko, K. A., Donovan-Kicken, E., & Bute, J.
J. (2008). The message design logics of responses to HIV disclosures. Human
Communication Research, 34, 655-684. doi: 10.1111/j.1468-2958.2008.00336.x
Chan, M. E. (2013). Antecedents of instrumental interpersonal help‐seeking: An integrative
review. Applied Psychology, 62, 571-596. doi: 10.1111/j.1464-0597.2012.00496.x
Chang, E. C. (2002). Optimism–pessimism and stress appraisal: Testing a cognitive interactive
model of psychological adjustment in adults. Cognitive Therapy and Research, 26, 675690. doi: 0.1023/A:1020313427884
Chelune, G. J. (1975). Self-disclosure: An elaboration of its basic dimensions. Psychological
Reports, 36, 79-85. doi: 10.2466/pr0.1975.36.1.79

105

Cobb, S. (1976). Social support as a moderator of life stress. Psychosomatic Medicine, 38, 300314.
Cohen, S., & Hoberman, H. (1983). Positive events and social supports as buffers of life change
stress. Journal of Applied Social Psychology, 13, 99-125. doi: 10.1111/j.15591816.1983.tb02325.x
Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis.
Psychological Bulletin, 98, 310-357.
Cunningham, M. R., Shaffer, D. R., Barbee, A. P., Wolff, P. L., & Kelley, D. J. (1990). Separate
processes in the relation of elation and depression to helping: Social versus personal
concerns. Journal of Experimental Social Psychology, 26, 13-33. doi: 10.1016/00221031(90)90059-U
Cutrona, C. E., & Russell, D. W. (1990). Type of social support and specific stress: Toward a
theory of optimal matching. In B. R. Sarason, I. G. Sarason & G. R. Pierce (Eds.), Social
support: An interactional view (Vol. XVI, pp. 319-366). Oxford, England: John Wiley &
Sons.
Cutrona, C. E., & Suhr, J. A. (1992). Controllability of stressful events and satisfaction with
spouse support behaviors. Communication Research, 19, 154-174. doi:
10.1177/009365092019002002
DeLongis, A., Coyne, J. C., Dakof, G., Folkman, S., & Lazarus, R. S. (1982). Relationship of
daily hassles, uplifts, and major life events to health status. Health Psychology, 1, 119136. doi: 10.1037/0278-6133.1.2.119
DeLongis, A., Folkman, S., & Lazarus, R. S. (1988). The impact of daily stress on health and
mood: Psychological and social resources as mediators. Journal of Personality and Social
Psychology, 54, 486-495. doi: 10.1037/0022-3514.54.3.486
DePaulo, B. M., & Fisher, J. D. (1980). The cost of asking for help. Basic and Applied Social
Psychology, 1, 23-35. doi: 10.1207/s15324834basp0101_3
Derlega, V. J., Winstead, B. A., Oldfield III, E. C., & Barbee, A. P. (2003). Close relationships
and social support in coping with HIV: A test of sensitive interaction systems theory.
AIDS and Behavior, 7, 119-129.
DeWall, C. N., Baumeister, R. F., Gailliot, M. T., & Maner, J. K. (2008). Depletion makes the
heart grow less helpful: Helping as a function of self-regulatory energy and genetic
relatedness. Personality and Social Psychology Bulletin, 34, 1653-1662. doi:
10.1177/0146167208323981
Dillard, J. P. (1990). A goal-driven model of interpersonal influence. In J. P. Dillard (Ed.),
Seeking compliance: The production of interpersonal influence messages (pp. 41-56).
Scottsdale, AZ: Gorsuch Scarisbrick.

106

Dillard, J. P., Segrin, C., & Harden, J. M. (1989). Primary and secondary goals in the production
of interpersonal influence messages. Communications Monographs, 56, 19-38. doi:
10.1080/03637758909390247
Donovan, E. E., LeFebvre, L., Tardif, S., Brown, L. E., & Love, B. (2014). Patterns of social
support communicated in response to expressions of uncertainty in an online community
of young adults with cancer. Journal of Applied Communication Research, 42, 432-455.
doi: 10.1080/00909882.2014.929725
Duggan, M., & Smith, A. (2013). Social media update 2013. Retrieved from
http://www.pewinternet.org/2013/12/30/social-media-update-2013/
Dunkel-Schetter, C., Blasband, D. E., Feinstein, L. G., & Herbert, T. B. (1992). Elements of
supportive interactions: When are attempts to help effective? In S. Spacapan & S.
Oskamp (Eds.), Helping and being helped: Naturalistic studies. The Claremont
Symposium on Applied Psychology (pp. 83-114). Thousand Oaks, CA: Sage.
Dunkel-Schetter, C., Folkman, S., & Lazarus, R. S. (1987). Correlates of social support receipt.
Journal of Personality and Social Psychology, 53, 71-80. doi: 10.1037/00223514.53.1.71
Dunkel-Schetter, C., & Skokan, L. A. (1990). Determinants of social support provision in
personal relationships. Journal of Social and Personal Relationships, 7, 437-450. doi:
10.1177/0265407590074002
Duthler, K. W. (2006). The politeness of requests made via email and voicemail: Support for the
hyperpersonal model. Journal of Computer-Mediated Communication, 11, 500-521. doi:
10.1111/j.1083-6101.2006.00024.x
Eastin, M. S., & LaRose, R. (2005). Alt. support: Modeling social support online. Computers in
Human Behavior, 21, 977-992. doi: 10.1016/j.chb.2004.02.024
Egbert, N. (2003). Support provider mood and familiar versus unfamiliar events: An
investigation of social support quality. Communication Quarterly, 51, 209-224. doi:
10.1080/01463370309370151
Eisenberg, D., Downs, M. F., Golberstein, E., & Zivin, K. (2009). Stigma and help seeking for
mental health among college students. Medical Care Research and Review, 66, 522-541.
doi: 10.1177/1077558709335173
Ellison, N. B., Gray, R., Vitak, J., Lampe, C., & Fiore, A. T. (2013). Calling all Facebook
friends: Exploring requests for help on Facebook. Paper presented at the International
AAAI Conference on Weblogs and Social Media, Boston, MA.
Ellison, N. B., Heino, R., & Gibbs, J. (2006). Managing impressions online: Self-presentation
processes in the online dating environment. Journal of Computer-Mediated
Communication, 11, 415-441. doi: 10.1111/j.1083-6101.2006.00020.x

107

Ellison, N. B., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook “friends”: Social
capital and college students’ use of online social network sites. Journal of ComputerMediated Communication, 12, 1143-1168. doi: 10.1111/j.1083-6101.2007.00367.x
Ellison, N. B., Steinfield, C., & Lampe, C. (2011). Connection strategies: Social capital
implications of Facebook-enabled communication practices. New Media & Society, 13,
873-892. doi: 10.1177/1461444810385389
Ellison, N. B., Vitak, J., Gray, R., & Lampe, C. (2014). Cultivating social resources on social
network sites: Facebook relationship maintenance behaviors and their role in social
capital processes. Journal of Computer‐Mediated Communication. 64, 635-657. doi:
10.1111/jcom.12106
Facebook. (2014). Group Basics. Facebook Help Center. Retrieved from
https://www.facebook.com/help/162866443847527/
Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2009). Statistical power analyses using
G*Power 3.1: Tests for correlation and regression analyses. Behavior Research
Methods, 41, 1149-1160. doi: 10.3758/BRM.41.4.1149
Feaster, J. C. (2010). Expanding the impression management model of communication channels:
An information control scale. Journal of Computer‐Mediated Communication, 16, 115138. doi: 10.1111/j.1083-6101.2010.01535.x
Fisher, S. L., & Ford, J. K. (1998). Differential effects of learner effort and goal orientation on
two learning outcomes. Personnel Psychology, 51, 397-420. doi: 10.1111/j.17446570.1998.tb00731.x
Fitzsimons, G. M., & Bargh, J. A. (2003). Thinking of you: Nonconscious pursuit of
interpersonal goals associated with relationship partners. Journal of Personality and
Social Psychology, 84, 148-164. doi: 10.1037/0022-3514.84.1.148
Flett, G. L., Blankstein, K. R., Hicken, D. J., & Watson, M. S. (1995). Social support and helpseeking in daily hassles versus major life events stress. Journal of Applied Social
Psychology, 25, 49-58. doi: 10.1111/j.1559-1816.1995.tb01583.x
Gibbs, J. L., Ellison, N. B., & Heino, R. D. (2006). Self-presentation in online personals the role
of anticipated future interaction, self-disclosure, and perceived success in Internet dating.
Communication Research, 33, 152-177. doi: 10.1177/0093650205285368
Gilbert, E., & Karahalios, K. (2009). Predicting tie strength with social media. Proceedings of
the SIGCHI Conference on Human Factors in Computing Systems (pp. 211-220). New
York, NY: ACM. doi: 10.1145/1518701.1518736
Goffman, E. (1959). The presentation of self in everyday life. New York, NY: Anchor Book.

108

Goldsmith, D. (1992). Managing conflicting goals in supportive interaction: An integrative
theoretical framework. Communication Research, 19, 264-286. doi:
10.1177/009365092019002007
Goldsmith, D. J. (1994). The role of facework in supportive communication. In B. R. Burleson,
T. L. Albrecht & I. G. Sarason (Eds.), Communication of social support: Messages,
interactions, relationships, and community (pp. 29-49). Thousand Oaks, CA: Sage.
Goldsmith, D. J. (2000). Soliciting advice: The role of sequential placement in mitigating face
threat. Communications Monographs, 67, 1-19. doi: 10.1080/03637750009376492
Goldsmith, D. J. (2004). Communicating social support. Cambridge, UK: Cambridge University
Press.
Gordon, M. E., Slade, L. A., & Schmitt, N. (1986). The “science of the sophomore” revisited:
From conjecture to empiricism. Academy of Management Review, 11, 191-207. doi:
10.5465/AMR.1986.4282666
Hampton, K., Goulet, L. S., Rainie, L., & Purcell, K. (2011). Social networking sites and our
lives. Washington, DC: Pew Research Center’s Internet & American Life Project.
Hayes, A. F., & Krippendorff, K. (2007). Answering the call for a standard reliability measure
for coding data. Communication Methods and Measures, 1, 77-89. doi:
10.1080/19312450709336664
Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world?.
Behavioral and Brain Sciences, 33, 61-83.doi: 10.1017/S0140525X0999152X
Herbert, T. B., & Cohen, S. (1993). Stress and immunity in humans: A meta-analytic review.
Psychosomatic Medicine, 55(4), 364-379.
High, A. C., & Dillard, J. P. (2012). A review and meta-analysis of person-centered messages
and social support outcomes. Communication Studies, 63, 99-118. doi:
10.1080/10510974.2011.598208
Hirsch, J. K., & Barton, A. L. (2011). Positive social support, negative social exchanges, and
suicidal behavior in college students. Journal of American College Health, 59, 393-398.
doi: 10.1080/07448481.2010.515635
Hobfòll, S. E. (1989). Conservation of resources: A new attempt at conceptualizing stress.
American Psychologist, 44, 513-524. doi: 10.1037/0003-066X.44.3.513
Holmstrom, A. J., Bodie, G. D., Burleson, B. R., McCullough, J. D., Rack, J. J., Hanasono, L.
K., & Rosier, J. G. (2013). Testing a dual-process theory of supportive communication
outcomes: How multiple factors influence outcomes in support situations.
Communication Research [Online Version]. doi: 10.1177/0093650213476293

109

Holmstrom, A. J., Burleson, B. R., & Jones, S. M. (2005). Some consequences for helpers who
deliver “cold comfort”: Why it’s worse for women than men to be inept when providing
emotional support. Sex Roles, 53, 153-172. doi: 10.1007/s11199-005-5676-4
Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social relationships and mortality risk: A
meta-analytic review. PLoS Medicine, 7, e1000316. doi: 10.1371/journal.pmed.100031
House, J. S. (1981). Work stress and social support. MA: Addison-Wesley.
House, J. S., & Kahn, R. L. (1985). Measures and concepts of social support. In S. Cohen & S.
L. Syme (Eds.), Social support and health (pp. 83-108). Orlando, FL: Academic Press.
Hullman, G. A. (2004). Interpersonal communication motives and message design logic:
Exploring their interaction on perceptions of competence. Communication Monographs,
71, 208-225. doi: 10.1080/0363775042000250411
IBM Corp. (2013). IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp.
Jones, S. (2004). Putting the person into person-centered and immediate emotional support
emotional change and perceived helper competence as outcomes of comforting in helping
situations. Communication Research, 31, 338-360. doi: 10.1177/0093650204263436
Jones, S. M., & Burleson, B. R. (1997). The impact of situational variables on helpers’
perceptions of comforting messages: An attributional analysis. Communication Research,
24, 530-555. doi: 10.1177/009365097024005004
Jones, S. M., & Wirtz, J. G. (2006). How does the comforting process work? An empirical test of
an appraisal-based model of comforting. Human Communication Research, 32, 217-243.
doi: 10.1111/j.1468-2958.2006.00274.x
Jung, J. (1988). Social support providers: Why do they help? Basic and Applied Social
Psychology, 9, 231-240. doi: 10.1207/s15324834basp0903_5
Jung, Y., Gray, R., Lampe, C., & Ellison, N. (2013). Favors from Facebook friends: Unpacking
dimensions of social capital. Proceedings of the SIGCHI Conference on Human Factors
in Computing Systems (pp. 11-20). New York, NY: ACM. doi:
10.1145/2470654.2470657
Kafetsios, K., & Sideridis, G. D. (2006). Attachment, social support and well-being in young and
older adults. Journal of Health Psychology, 11, 863-875. doi:
10.1177/1359105306069084
Kanner, A. D., Coyne, J. C., Schaefer, C., & Lazarus, R. S. (1981). Comparison of two modes of
stress measurement: Daily hassles and uplifts versus major life events. Journal of
Behavioral Medicine, 4(1), 1-39.
Kellermann, K. (1992). Communication: Inherently strategic and primarily automatic.
Communications Monographs, 59, 288-300. doi: 10.1080/03637759209376270Keppel,
110

G., & Wickens, T. D. (2004). Design and analysis. A researcher’s handbook (4th ed.).
Upper Saddle River, NJ: Pearson Education International.
Kim, J., & Lee, J. E. R. (2011). The Facebook paths to happiness: Effects of the number of
Facebook friends and self-presentation on subjective well-being. Cyberpsychology,
Behavior, and Social Networking, 14, 359-364. doi: 10.1089/cyber.2010.0374
Kollock, P., & Smith, M. (1996). Managing the virtual commons: Cooperation and conflict in
computer communities. In S. C. Herring (Ed.), Computer-mediated communication:
Linguistic, social, and cross-cultural perspectives (pp. 109-128). Philadelphia, PA: John
Benjamins Publishing Company
Kraaij, V., Arensman, E., & Spinhoven, P. (2002). Negative life events and depression in elderly
persons: A meta-analysis. The Journals of Gerontology Series B: Psychological Sciences
and Social Sciences, 57, P87-P94. doi: 10.1093/geronb/57.1.P87
Kraut, R., Patterson, M., Lundmark, V., Kiesler, S., Mukophadhyay, T., & Scherlis, W. (1998).
Internet paradox: A social technology that reduces social involvement and psychological
well-being? American Psychologist, 53, 1017-1031. doi: 10.1037//0003-066X.53.9.1017
Krippendorff, K. (1980). Content analysis: An introduction to its methodology. Beverly Hills,
CA: Sage.
Kuo, F. Y., Tseng, C. Y., Tseng, F. C., & Lin, C. S. (2013). A study of social information control
affordances and gender difference in Facebook self-presentation. Cyberpsychology,
Behavior, and Social Networking, 16, 635-644. doi:10.1089/cyber.2012.0345
Lacy, S., & Riffe, D. (1996). Sampling error and selecting intercoder reliability samples for
nominal content categories. Journalism & Mass Communication Quarterly, 73, 963973. doi: 10.1177/107769909607300414
Lambert, B. L., & Gillespie, J. L. (1994). Patient perceptions of pharmacy students’ hypertension
compliance-gaining messages: Effects of message design logic and content themes.
Health Communication, 6, 311-325. doi: 10.1207/s15327027hc0604_6
Lang, P. J. (1980). Behavioral treatment and bio-behavioral assessment: Computer applications.
In J. B. Sidowski, J. H. Johnson, & T. A. Williams (Eds.), Technology in mental health
care delivery systems (pp. 119-l37). Norwood, NJ: Ablex.
Lang, P.J., Bradley, M.M., & Cuthbert, B.N. (2008). International affective picture system
(IAPS): Affective ratings of pictures and instruction manual. Technical Report A-8.
University of Florida, Gainesville, FL.
LaRose, R., Eastin, M. S., & Gregg, J. (2001). Reformulating the Internet paradox: Social
cognitive explanations of Internet use and depression. Journal of Online Behavior, 1(2),
1092-4790.

111

Leary, M. R., & Kowalski, R. M. (1990). Impression management: A literature review and twocomponent model. Psychological Bulletin, 107, 34-47. doi: 10.1037//0033-2909.107.1.34
Lee, K.-T., Noh, M.-J., & Koo, D.-M. (2013). Lonely people are no longer lonely on social
networking sites: The mediating role of self-disclosure and social support.
Cyberpsychology, Behavior, & Social Networking, 16, 413-418. doi:
10.1089/cyber.2012.0553
Lehman, D. R., Ellard, J. H., & Wortman, C. B. (1986). Social support for the bereaved:
Recipients’ and providers’ perspectives on what is helpful. Journal of Consulting and
Clinical Psychology, 54, 438-446. doi: 10.1037/0022-006X.54.4.438
Lemieux, R., & Tighe, M. R. (2004). Attachment styles and the evaluation of comforting
responses: A receiver perspective. Communication Research Reports, 21, 144-153. doi:
10.1080/08824090409359976
Levene, H. (1960). Robust testes for equality of variances. In I. Olkin (Ed.), Contributions to
probability and statistics (pp.278-292). Palo Alto, CA: Stanford University Press.
Lim, V. K., Teo, T. S., & Zhao, X. (2013). Psychological costs of support seeking and choice of
communication channel. Behaviour & Information Technology, 32, 132-146. doi:
10.1080/0144929X.2010.518248
McFadden, D. (1974) Conditional logit analysis of qualitative choice behavior. In P. Zarembka
(Ed.), Frontiers in econometrics (pp. 105-142). Waltham, MA: Academic Press.
MacGeorge, E. L. (2001). Support providers’ interaction goals: The influence of attributions and
emotions. Communication Monographs, 68, 72-97. doi: 10.1080/03637750128050
MacGeorge, E. L., Gillihan, S. J., Samter, W., & Clark, R. A. (2003). Skill deficit or differential
motivation? Testing alternative explanations for gender differences in the provision of
emotional support. Communication Research, 30, 272-303. doi:
10.1177/0093650203030003002
MacGeorge, E. L., Lichtman, R. M., & Pressey, L. C. (2002). The evaluation of advice in
supportive interactions: Facework and contextual factors. Human Communication
Research, 28, 451-463. doi: 10.1093/hcr/28.3.451
Macrae, C. N., & Johnston, L. (1998). Help, I need somebody: Automatic action and inaction.
Social Cognition, 16, 400-417. doi: 10.1521/soco.1998.16.4.400
Mehdizadeh, S. (2010). Self-presentation 2.0: Narcissism and self-esteem on Facebook.
Cyberpsychology, Behavior, and Social Networking, 13, 357-364.
doi:10.1089/cyber.2009.0257
Meyer, J. R. (2002). Contextual influences on the pursuit of secondary goals in request
messages. Communication Monographs, 69, 189-203. doi: 10.1080/03637750216539

112

Moreno, M. A., Jelenchick, L. A., Egan, K. G., Cox, E., Young, H., Gannon, K. E., & Becker, T.
(2011). Feeling bad on Facebook: Depression disclosures by college students on a social
networking site. Depression and Anxiety, 28, 447-455. doi: 10.1002/da.20805
Mortenson, S. T. (2009). Interpersonal trust and social skill in seeking social support among
Chinese and Americans. Communication Research, 36, 32-53. doi:
10.1177/0093650208326460
Nabi, R. L., Prestin, A., & So, J. (2013). Facebook friends with (health) benefits? Exploring
social network site use and perceptions of social support, stress, and well-being.
Cyberpsychology, Behavior, and Social Networking, 16, 721-727. doi:
10.1089/cyber.2012.0521
Newman, M. W., Lauterbach, D., Munson, S. A., Resnick, P., & Morris, M. E. (2011). It’s not
that I don’t have problems, I’m just not putting them on Facebook: Challenges and
opportunities in using online social networks for health. Proceedings of the ACM 2011
Conference on Computer Supported Cooperative Work (pp. 341-350). New York, NY:
ACM. doi: 10.1145/1958824.1958876
Newman, R. S., & Goldin, L. (1990). Children’s reluctance to seek help with schoolwork.
Journal of Educational Psychology, 82, 92-100. doi: 10.1037/0022-0663.82.1.92
Nixon, A. E., Mazzola, J. J., Bauer, J., Krueger, J. R., & Spector, P. E. (2011). Can work make
you sick? A meta-analysis of the relationships between job stressors and physical
symptoms. Work & Stress: An International Journal of Work, Health & Organisations,
25, 1-22. doi: 10.1080/02678373.2011.569175
Notani, A. S. (1998). Moderators of perceived behavioral controls’ predictiveness in the theory
of planned behavior: A meta-analysis. Journal of Consumer Psychology, 7, 247-271.
doi:10.1207/s15327663jcp0703_02
Oh, H. J., Lauckner, C., Boehmer, J., Fewins-Bliss, R., & Li, K. (2013). Facebooking for health:
An examination into the solicitation and effects of health-related social support on social
networking sites. Computers in Human Behavior, 29, 2072-2080. doi:
10.1016/j.chb.2013.04.017
Oh, H. J., & Lee, B. (2012). The effect of computer-mediated social support in online
communities on patient empowerment and doctor–patient communication. Health
Communication, 27, 30-41. doi: 10.1080/10410236.2011.567449
Oh, H. J., Ozkaya, E., & LaRose, R. (2014). How does online social networking enhance life
satisfaction? The relationships among online supportive interaction, affect, perceived
social support, sense of community, and life satisfaction. Computers in Human Behavior,
30, 69-78. doi: 10.1016/j.chb.2013.07.053
O'Keefe. (1988). The logic of message design: Individual differences in reasoning about
communication. Communication Monographs, 55, 80-103. doi:
10.1080/03637758809376159
113

O’Keefe, B. J., & McCornack, S. A. (1987). Message design logic and message goal structure
effects on perceptions of message quality in regulative communication situations. Human
Communication Research, 14, 68-92. doi: 10.1111/j.1468-2958.1987.tb00122.x
O’Sullivan, P.B. (2000). What you don’t know won’t hurt me: Impression management functions
of communication channels in relationships. Human Communication Research, 26,
403–431. doi: 10.1111/j.1468-2958.2000.tb00763.x
Otten, C. A., Penner, L. A., & Waugh, G. (1988). That’s what friends are for: The determinants
of psychological helping. Journal of Social and Clinical Psychology, 7, 34-41. doi:
10.1521/jscp.1988.7.1.34
Park, N., Kee, K. F., & Valenzuela, S. (2009). Being immersed in social networking
environment: Facebook groups, uses and gratifications, and social outcomes.
Cyberpsychology & Behavior, 12, 729-733. doi: 10.1089/cpb.2009.0003
Perry, B. L. (2011). The labeling paradox stigma, the sick role, and social networks in mental
illness. Journal of Health and Social Behavior, 52, 460-477. doi:
10.1177/0022146511408913
Peterson, L. W., & Albrecht, T. L. (1996). Message design logic, social support, and mixedstatus relationships. Western Journal of Communication, 60, 291-309. doi:
10.1080/10570319609374551
Peterson, R. A. (2001). On the use of college students in social science research: Insights from a
second‐order meta‐analysis. Journal of Consumer Research, 28, 450-461. doi:
10.1086/323732
Phua, J. (2013). Participating in health issue-specific social networking sites to quit smoking:
How does online social interconnectedness influence smoking cessation self-efficacy?
Journal of Communication, 63, 933-952. doi: 10.1111/jcom.12054
Porter, L., & Lawler, E. (1968). Managerial attitudes and performance. Homewood, IL: Irwin
Dorsey
Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community. New
York, NY: Simon & Schuster.
Qiu, L., Lin, H., Leung, A. K., & Tov, W. (2012). Putting their best foot forward: Emotional
disclosure on Facebook. Cyberpsychology, Behavior, and Social Networking, 15, 569572. doi: 10.1089/cyber.2012.0200
Rieh, S. Y., Kim, Y. M., & Markey, K. (2012). Amount of invested mental effort (AIME) in
online searching. Information Processing & Management, 48, 1136-1150.
doi:10.1016/j.ipm.2012.05.001
Riffe, D., Lacy, S., Fico, F. (1998). Analyzing media messages: Using quantitative content
analysis in research. Mahwah, NJ: Erlbaum.
114

Rosenberg, J., & Egbert, N. (2011). Online impression management: Personality traits and
concerns for secondary goals as predictors of self-presentation tactics on Facebook.
Journal of Computer-Mediated Communication, 17, 1-18. doi: 10.1111/j.10836101.2011.01560.x
Rubin, R. B., Graham, E. E., & Mignerey, J. T. (1990). A longitudinal study of college students’
communication competence. Communication Education, 39, 1-14. doi:
10.1080/03634528809378700
Samp, J. A. (2006). Identity-processing orientations and the importance of single and multiple
communication goals for managing problematic events. Communication Research
Reports, 23, 101-110. doi: 10.1080/08824090600668949
Samp, J. A., & Solomon, D. H. (1998). Communicative responses to problematic events in close
relationships I: The variety and facets of goals. Communication Research, 25, 66-95. doi:
10.1177/009365098025001003
Samp, J. A., & Solomon, D. H. (1999). Communicative responses to problematic events in close
relationships II: The influence of five facets of goals on message features.
Communication Research, 26, 193-239. doi: 10.1177/009365099026002005
Samp, J. A., & Solomon, D. H. (2005). Toward a theoretical account of goal characteristics in
micro-level message features. Communication Monographs, 72, 22-45. doi:
10.1080/0363775052000342517
Samter, W. (2002). How gender and cognitive complexity influence the provision of emotional
support: A study of indirect effects. Communication Reports, 15, 5-16. doi:
10.1080/08934210209367748
Schrader, D. C. (1999). Goal complexity and the perceived competence of interpersonal
influence messages. Communication Studies, 50, 188-202. doi:
10.1080/10510979909388486
Schrader, D. C., & Dillard, J. P. (1998). Goal structures and interpersonal influence.
Communication Studies, 49, 276-293. doi: 10.1080/10510979809368538
Sears, D. O. (1986). College sophomores in the laboratory: Influence of a narrow database on
social psychology’s view of human nature. Journal of Personality and Social Psychology
51, 515-530. doi: 10.1037/0022-3514.51.3.515
Scott, A. M., Caughlin, J. P., Donovan-Kicken, E., & Mikucki-Enyart, S. L. (2013). Do message
features influence responses to depression disclosure? A message design logics
perspective. Western Journal of Communication, 77, 139-163. doi:
10.1080/10570314.2012.694007
Segerstrom, S. C., & Miller, G. E. (2004). Psychological stress and the human immune system:
A meta-analytic study of 30 years of inquiry. Psychological Bulletin, 130, 601-630. doi:
10.1037/0033-2909.130.4.601
115

Shansky, R. M., & Lipps, J. (2013). Stress-induced cognitive dysfunction: Hormoneneurotransmitter interactions in the prefrontal cortex. Frontiers in Human Neuroscience,
7, 1-6. doi: 10.3389/fnhum.2013.00123
Shapiro, E. G. (1978). Help seeking: Effects of visibility of task performance and seeking help.
Journal of Applied Social Psychology, 8, 163-173. doi: 10.1111/j.15591816.1978.tb00774.x
Siedlecki, K. L., Salthouse, T. A., Oishi, S., & Jeswani, S. (2014). The relationship between
social support and subjective well-being across age. Social Indicators Research, 117,
561-576. doi: 10.1007/s11205-013-0361-4
Spendelow, J. S., & Jose, P. E. (2010). Does the optimism bias affect help-seeking intentions for
depressive symptoms in young people? The Journal of General Psychology:
Experimental, Psychological, and Comparative Psychology, 137, 190-209. doi:
10.1080/00221301003645277
Steinfield, C., Ellison, N. B., & Lampe, C. (2008). Social capital, self-esteem, and use of online
social network sites: A longitudinal analysis. Journal of Applied Developmental
Psychology, 29, 434-445. doi: 10.1016/j.appdev.2008.07.002
Taylor, S. E., Sherman, D. K., Kim, H. S., Jarcho, J., Takagi, K., & Dunagan, M. S. (2004).
Culture and social support: Who seeks it and why?. Journal of Personality and Social
Psychology, 87, 354-362. doi: 10.1037/0022-3514.87.3.354
Tessler, R. C, & Schwartz, S. H. (1972). Help seeking, self-esteem, and achievement motivation:
An attributional analysis. Journal of Personality and Social Psychology, 27, 318-326.
doi: dx.doi.org/10.1037/h0032321
Thoits, P. A. (2011). Mechanisms linking social ties and support to physical and mental health.
Journal of Health and Social Behavior, 52, 145-161. doi: 10.1177/0022146510395592
Tracy, M., Norris, F. H., & Galea, S. (2011). Differences in the determinants of posttraumatic
stress disorder and depression after a mass traumatic event. Depression and Anxiety, 28,
666-675. doi: 10.1002/da.20838
Treem, J.W., & Leonardi, P. M. (2012). Social media use in organizations: Exploring the
affordances of visibility, editability, persistence, and association. Communication
Yearbook, 36, 143–189.
Trepte, S., & Reinecke, L. (2013). The reciprocal effects of social network site use and the
disposition for self-disclosure: A longitudinal study. Computers in Human Behavior, 29,
1102-1112. doi: 10.1016/j.chb.2012.10.002
Turner, J. W., Grube, J. A., & Meyers, J. (2001). Developing an optimal match within online
communities: An exploration of CMC support communities and traditional support.
Journal of Communication, 51, 231-251. doi: 10.1093/joc/51.2.231

116

Turnley, W. H., & Bolino, M. C. (2001). Achieving desired images while avoiding undesired
images: Exploring the role of self-monitoring in impression management. Journal of
Applied Psychology, 86, 351-360. doi: 10.1037/0021-9010.86.2.351
Ursin, H., & Eriksen, H. R. (2004). The cognitive activation theory of stress.
Psychoneuroendocrinology, 29, 567-592. doi:10.1016/S0306-4530(03)00091-X
Valenzuela, S., Park, N., & Kee, K. F. (2009). Is there social capital in a social network site?:
Facebook use and college students’ life satisfaction, trust, and participation. Journal of
Computer-Mediated Communication, 14, 875-901. doi: 10.1111/j.10836101.2009.01474.x
Vitak, J., & Ellison, N. B. (2012). ‘There’s a network out there you might as well tap’: Exploring
the benefits of and barriers to exchanging informational and support-based resources on
Facebook. New Media & Society, 15, 243-259. doi: 10.1177/1461444812451566
Vitak, J., Ellison, N. B., & Steinfield, C. (2011). The ties that bond: Re-examining the
relationship between Facebook use and bonding social capital. Proceedings of the 2011
44th Hawaii International Conference on System Sciences (pp. 1-10). Washington, DC:
IEEE Computer Society. doi: 10.1109/HICSS.2011.435
Vodenicharov, T., Dyakova, M., & Shipkovenska, E. (2005). Stress as a determinant of health. In
L. Georgieva & G. Burazeri (Eds.), Health determinants in the scope of new public
health: A handbook for teachers, researchers, and health professionals (pp. 162-179).
Lage, Germany: Hans Jacobs Publishing Company.
Vogel, D. L., Wade, N. G., & Hackler, A. H. (2007). Perceived public stigma and the willingness
to seek counseling: The mediating roles of self-stigma and attitudes toward counseling.
Journal of Counseling Psychology, 54, 40-50. doi: 10.1037/0022-0167.54.1.40
Walther, J. B. (1996). Computer-mediated communication impersonal, interpersonal, and
hyperpersonal interaction. Communication Research, 23, 3-43. doi:
10.1177/009365096023001001
Walther, J. B. (2007). Selective self-presentation in computer-mediated communication:
Hyperpersonal dimensions of technology, language, and cognition. Computers in Human
Behavior, 23, 2538-2557. doi: 10.1016/j.chb.2006.05.002
Walther, J. B. (2012). Interaction through technological lenses: Computer-mediated
communication and language. Journal of Language and Social Psychology, 13, 397414. doi:10.1177/0261927X12446610
Walther, J. B., & Boyd, S. (2002). Attraction to computer-mediated social support. In C. A. Lin
& D. Atkin (Eds.), Communication technology and society: Audience adoption and uses
(pp. 153-188). Cresskill, NJ: Hampton Press.

117

Wang, J., Korczykowski, M., Rao, H., Fan, Y., Pluta, J., Gur, R. C., ... & Detre, J. A. (2007).
Gender difference in neural response to psychological stress. Social Cognitive and
Affective Neuroscience, 2, 227-239. doi: 10.1093/scan/nsm018
Watson, S. B., Goh, Y. W., & Sawang, S. (2011). Gender influences on the work-related stresscoping process. Journal of Individual Differences, 32, 39-46. doi: 10.1027/16140001/a000033
Wegner, D. M., & Bargh, J. A. (1998). Control and automaticity in social life. In D. T. Gilbert,
S. T. Fiske, & L. Gardner (Eds.), The Handbook of social psychology (4th ed., pp. 446496). New York, NY: McGraw-Hill.
Weiner, B. (2000). Intrapersonal and interpersonal theories of motivation from an attributional
perspective. Educational Psychology Review, 12, 1-14. doi: 10.1007/978-1-4615-12738_2
Welch, B. L. (1947). The generalization of students’ problem when several different population
variances are involved. Biometrika, 34, 28-35. doi:10.1093/biomet/34.1-2.28
Wilcox, B. L., & Birkel, R. C. (1983). Social networks and the help-seeking process: A structural
perspective. In A. Nadler, J. Fisher & B. M. DePaulo (Eds.), New directions in helping
(Vol. 3, pp. 235-253). New York, NY: Academic Press.
Williams, K. B., & Williams, K. D. (1983). Social inhibition and asking for help: The effects of
number, strength, and immediacy of potential help givers. Journal of Personality and
Social Psychology, 44, 67-77. doi: 10.1037/0022-3514.44.1.67
Williamson, J. B. (1974). The stigma of public dependency: A comparison of alternative forms
of public aid to the poor. Social Problems, 22 (2), 213-238.
Wilson, S. R. (2007). Communication theory and the concept of “goal”. In B. B. Whaley & W.
Samter (Eds.), Explaining communication: Contemporary theories and exemplars (pp.
73-104). Mahwah, NJ: Erlbaum.
Wilson, S. R., Aleman, C. G., & Leatham, G. B. (1998). Identity implications of influence goals:
A revised analysis of face-threatening acts and application to seeking compliance with
same-sex friends. Human Communication Research, 25, 64-96. doi: 10.1111/j.14682958.1998.tb00437.x
Wilson, S. R., & Feng, H. (2007). Interaction goals and message production: Conceptual and
methodological developments. In D. Roskos-Ewoldsen & J. Monahan (Eds.),
Communication and social cognition: Theories and methods (pp. 71-95). Mahwah: NJ:
Lawrence Erlbaum.
Wright, K. B. (2012). Emotional support and perceived stress among college students using
Facebook.com: An exploration of the relationship between source perceptions and
emotional support. Communication Research Reports, 29, 175-184. doi:
10.1080/08824096.2012.695957
118

Wright, K. B., Rosenberg, J., Egbert, N., Ploeger, N. A., Bernard, D. R., & King, S. (2013).
Communication competence, social support, and depression among college students: A
model of Facebook and face-to-face support network influence. Journal of Health
Communication, 18, 41-57. doi: 10.1080/10810730.2012.688250
Yeo, G. B., & Neal, A. (2004). A multilevel analysis of effort, practice, and performance: Effects
of ability, conscientiousness, and goal orientation. Journal of Applied Psychology, 89,
231-247. doi: http://dx.doi.org/10.1037/0021-9010.89.2.231
Yzer, M C. & Southwell, B. G. (2008). New communication technologies, old questions.
American Behavioral Scientist, 25, 8-20. doi: 10.1177/0002764208321338
Zajonc, R. B. (1998). Emotions. In D. T. Gilbert, S. T. Fiske, & Gardner, L. (Eds), The handbook
of social psychology (4th ed., pp. 591-632). New York, NY: McGraw-Hill.
Zhao, S., Grasmuck, S., & Martin, J. (2008). Identity construction on Facebook: Digital
empowerment in anchored relationships. Computers in Human Behavior, 24, 1816-1836.
doi: 10.1016/j.chb.2008.02.012

119