The Role of Personality Traits, Social Anxiety, Depressive ...

The Role of Personality Traits, Social Anxiety, Depressive Symptoms, and Other

Psychosocial Factors in the Motivation for Social Internet Use

by

C. Kyle Schindler, M.A.

A Dissertation in

Counseling Psychology

Submitted to the Graduate Faculty

of Texas Tech University in

Partial Fulfillment of

the Requirements for

the Degree of

Doctor of Philosophy

Approved

C. Steven Richards, Ph.D.

Chair of Committee

Andrew K. Littlefield, Ph.D.

Co-chair of Committee

Stephen W. Cook, Ph.D.

Robert A. Morgan, Ph.D.

Mark Sheridan

Dean of the Graduate School

August, 2016

Texas Tech University, C. Kyle Schindler, August 2016

ii

Copyright 2016, Charles Kyle Schindler


iii

Acknowledgements

“Silent gratitude isn't much use to anyone.” -G.B. Stern

The dissertation process can be quite difficult at times, and I have been blessed

with support from many amazing people who I should acknowledge. First and foremost,

I would like to sincerely thank my advisor, Dr. Steven Richards, for his support

throughout the last 6 years. Without the aid of your guidance, or the relief of your

patience, experience, and advice, I do not believe I would have been able to make it

through this process. It has truly been a privilege to work with you. I would like to

express my sincerest gratitude to my committee co-chair, Dr. Andrew Littlefield, for his

enthusiasm and his patience as I navigated a sometimes daunting statistical procedure.

Your organization, knowledge, and your time were all highly appreciated. I would like to

thank my committee members, Drs. Bob Morgan and Stephen Cook, for their suggestions

for this dissertation, as well as their valuable insight into graduate school in general.

To my father, my mother, and my younger brothers, Cody and Scott: I love you

more than words can express. You have inspired me to achieve in a way that only a

family could. I simply could not have done this without all of you. To Dr. Jennifer

Vencill, Dr. Andrew Friedman, and Blakely Low: You have served as friends, role

models, and mentors at times when I greatly needed them, motivated me to continue

pushing forward when I felt discouraged, provided a direction when I felt lost, and

offered support and help in innumerable ways. To Adam Cann, Klaudia Pereira, Mike

Crites, Dr. James Cazares, John Schumacher, Dr. Curtis Craig and others too

innumerable to list completely: You have enriched my life in many ways, and have made

my graduate school experience one of the best periods of my life.


iv

Table of Contents

Acknowledgements………………………………………………………………...…….iii

Abstract……..………………………………………………………………..…………...vi

List of Tables……………………………………………………………………………viii

List of Figures……………………………………………………………………….……ix

List of Abbreviations…………………………………………………………………...…x

Chapter One: Introduction …...………………..………………….…………..…………..1

Rationale for the Present Study……………………………….….……………….9

Hypotheses ………………………………………………………….…...………11

Chapter Two: Methods …………………………………………………….……………15

Sample and Procedures..…………………………………………………………15

Instruments …………………………………………………………….…….......16

Chapter Three: Results ...……………………………………………………..................24

Measurement Model………………………………………………………..……28

Structural Model……………………………………………………………..…..33

Chapter Four: Discussion………………..…………………………………………….…36

Limitations…………………………………………………………………….…39

References ………………………………………………………………………...……..46

Appendices ……………………………………………………………………….…..….65

Appendix A: Extended Literature Review…………………………………....….65

Appendix B: Tables and Figures……………………………………………..…..93

Appendix C: Models…………………………………………………….….…....98

Appendix D: Demographic Form……………………………………………....103


v

Appendix E: CES-D……………………………………………………….……105

Appendix F: Big-Five Inventory……………………………………………......106

Appendix G: Perceived Stress Scale – 10 ……………………………………...107

Appendix H: Liebowitz Social Anxiety Scale……………………………….....108

Appendix I: COPE emotional social support items....……………………….....109

Appendix J: Internet Motivation Scale……………………………………..….110

Appendix K: General Problematic Internet Use Scale – ...…………...…….….112

Appendix L: Modified Social Connectedness Scale…………………………....114

Appendix M: Marlow-Crowne Social Desirability Scale………………………116


vi

Abstract

The use of the Internet for emotional and social support has received increasing

attention from researchers. Although early perspectives suggested this activity was

problematic (Caplan, 2002; Davis, 2001; Young, 1998), a growing body of research has

challenged this perspective (Byun et al., 2009), and found that social Internet use may

sometimes serve as an effective coping mechanism (Kraut et al., 2002; Leung, 2007). In

particular, Internet users who are high on certain personality traits (e.g., neuroticism,

introversion), or who are experiencing certain psychosocial factors (e.g. depressive

symptoms, social anxiety), may obtain some benefit from social Internet use, possibly

due to the unique communicative aspects of the Internet (Butt & Phillips, 2008;

Schouten, Valkenburg, & Peter, 2007; Zywica & Danowski, 2008), including facilitating

coping processes (Barak, Boniel-Nissim, & Suler, 2008; Griffiths, Calear, & Banfield,

2009).

Previous research has focused on the negative aspects of this behavior, or if

focusing on the positive aspects, has only assessed the relationships among a few

empirically identified variables. The current study proposed and assessed a more

complex structural equation model, in which the motivation for social Internet use is

predicted by the personality traits (e.g. neuroticism) and psychosocial factors (e.g.

depressive symptoms, social anxiety, perceived stress) which are most strongly

associated with this behavior. This study is the first known to assess the multivariate

relationships among these variables, including preliminary assessment of both negative

(e.g., obsessive thoughts about the Internet) and positive (e.g., increased social

connectedness) outcomes of Internet use in relation to these variables.


vii

College students in a large Southern university (N = 293) were assessed on each

of these factors via self-report measures. Structural equation modeling (SEM) was

utilized to determine the multivariate relationships among these variables and assess the

goodness of fit of the proposed model. Using the saturated model to aid in diagnostic

approaches, multicollinearity was discovered between motives and negative outcomes,

such that it impacted fit of the hypothesized model and prevented interpretability. An

alternative model was proposed which was theoretically sound, was interpretable, and

which was ultimately retained. Although the directionality of the relations among these

variables cannot be fully determined through SEM, there appears to be some comorbidity

between depressive symptoms, social anxiety, and outcomes of SIU. Clinicians should

assess the presence of online relationships when determining social support and

interpersonal functioning for socially anxious and depressed clients, as these relationships

may likely be present and influential. Strengths and limitations of the current study and

methodology are discussed.


viii

List of Tables

1. Factor Loadings and Composite Reliabilities for each Latent Factor………….41

2. Descriptive Statistics for Full Measures………………………………………..102

3. Correlations between Measures………………………………………………..103

4. Fit of Initial Measurement Model and Final Measurement Model with Items

Retained………………………………………………………..………………104

5. Path Loadings and Correlated Disturbances for the Saturated Model………..105

6. Participants’ Reported Internet Use……………………………………………106


ix

List of Figures

1. Measurement Model………………………………………………………..107

2. Hypothesized Structural Model…………………………………………….108

3. Path Loadings for Initial Hypothesized Model…………………………….109

4. Path Loadings for Alternative Model………………………………………110

5. Path Loadings for Saturated Model………………………………………..111


x

List of Abbreviations

BFI: Big Five Inventory

CES-D: Center for Epidemiological Studies – Depression Scale

CFI: Comparative Fit Index

GPIUS-2: General Problematic Internet Use Scale – 2

IMS: Internet Motivation Scale

ISG: Internet Support Group

LSAS: Leibowitz Social Anxiety Scale

MCSDS: Marlowe-Crowne Social Desirability Scale

PANAS: Positive and Negative Affect Schedule

PIU: Problematic Internet Use

PSS-10: Perceived Stress Scale - 10

RMSEA: Root Mean-Square Error of Approximation

SCS-R: Social Connectedness Scale - Revised

SEM: Structural Equation Modeling

SIU: Social Internet Use

ULI: Unit Loading Identification


1

Why do people use the Internet to seek relationships?

Human beings are a highly social species, as they have a natural desire to connect

to others (see Baumeister & Leary, 1995, for a review) and seek close relationships for

comfort and support (Buhrmester, 1996). This human need to connect has been

profoundly modified by the advent of the Internet. Social Internet use (SIU), defined as

use of the Internet to communicate with others, has increased throughout the years. This

includes any electronic media through which a minimum of two individuals communicate

at a distance (e.g. cell phones, computers), and can involve immediate (e.g. Skype) or

delayed communication (e.g. email, messaging). Early research found that up to 14% of

U.S. adolescents reported maintaining an online friendship (Wolak, Mitchell, &

Finkelhor, 2003). More recent surveys have found that 65% of adolescents participate in

social networking sites, 49% read the blogs of others, and 68% use Instant Messaging

software (Jones & Fox, 2009). In a two-year period, Facebook experienced a twofold

increase in membership, going from 500 million users to 1 billion users (Facebook Data

Team, 2010; Vance, 2012) and websites such as Instagram report over 400 million active

monthly users in 2016 (www.instagram.com/press). Internet users have become

increasingly more social in their online behaviors, and researchers have increasingly

questioned the potential consequences, both beneficial and problematic, of SIU. In

particular, SIU is seen as both a valid alternative to traditional face-to-face

communication (Amichai-Hamburger, Wainapel, & Fox, 2002; McKenna, Green, &

Gleason, 2002) and a harmful activity which impacts face-to-face communication (Davis,

2001; Caplan, 2002).


2

Does SIU Impact Mental Health and Social Support?

Early studies on social Internet use tended to focus on the theoretical limitations

of this medium in the forming and maintaining of relationships. Online relationships

were thought to be lacking in both emotional and physical closeness, making them poor

social support buffers when an individual experienced distress (Kraut et al., 1998). It was

also believed that these relationships would take away some time from face-to-face

relationships, leading to further negative consequences and reducing social support.

Indeed, Kraut and colleagues assessed new Internet users over a period of 12-24 months

and found that SIU led to increased self-reported depressive symptoms and feelings of

loneliness.

Those experiencing mood disruptions, anxiety, or stress were also believed to be

more motivated to engage in SIU (Davis, 2001; Caplan, 2002). This was considered

problematic, as these individuals might maintain maladaptive cognitions and beliefs (e.g.

“I am worthless offline, but online I am someone”), which could cause them to use the

Internet excessively and neglect other areas of functioning (e.g. their jobs, schoolwork).

Some evidence supports this assertion. Excessively engaging in SIU has been

found to be comorbid with mood disorders, anxiety disorders, and ADHD in certain

populations (Tokunaga & Rains, 2016; Weinstein & Lejoyeux, 2010). Individuals high

in loneliness have indicated that they prefer SIU over face-to-face interactions, and that

their Internet use has caused disturbances in functioning (de Ayala López., Gutierrez, &

Jiménez, 2015; Morahan-Martin & Schumacher, 2003). Time spent online has been

shown to have a positive association with feelings of loneliness (Matsuba, 2006;

Stepanikova, Nie, & He, 2010; Tokunaga & Rains, 2016), poorer coping strategies


3

(Milani, Osualdella, & Di Blasio, 2009), and decreased social integration (de Ayala

López., Gutierrez, & Jiménez, 2015; Weiser, 2001).

However, several studies have found that there are possibly benefits to SIU as

well. The negative consequences discovered in the original Kraut study (Kraut et al.,

1998) were found to have largely dissipated when the researchers followed up with this

sample (Kraut et al., 2002). Consistent with this, a study with Serbian adolescents found

no relation between depression and level of SIU (Banjanin, Banjanin, Dimitrijevic &

Pantic, 2015). Furthermore, SIU has been associated with decreased depressive

symptoms (Morgan & Cotten, 2003; Shaw & Gant, 2002), as well as decreased

loneliness, improved self-esteem and improved perceptions of social support (Shaw &

Gant, 2002). Moreover, SIU has been associated with improved connection with family,

friends, and individuals with shared interests (Amichai-Hamburger & Hayat, 2011).

Finally, internet users who engage in SIU report specifically perceiving that this kind of

internet use can be psychologically beneficial to them (Campbell, Cumming, & Hughes,

2006).

Internet Support Groups

Studies on the effects of Internet support groups (ISGs; Also known as Online

Support Groups) provide additional support for the potential benefits of this behavior.

ISGs are online message boards or forums where members may receive emotional and

social support, discuss problems, or share information (Barak, Boniel-Nassim, & Suler,

2008). Individuals managing a variety of concerns and significant health conditions have

reported ISG use to be beneficial for them, including hysterectomies (Bunde, Suls,

Martin, & Barnett, 2006), visual impairment (Smedema & McKenzie, 2010), cancer


4

(Beaudoin & Tao, 2007; Han et al., 2008; Seckin, 2013; Shim, Cappella, & Han, 2011),

suicidal ideation (Gilat & Shahar, 2009), HIV/AIDS (Mo & Coulson, 2010), and

Parkinson’s Disease (Attard & Coulson, 2012).

In contrast to perspectives that seeking others out online can be harmful (Caplan,

2002; Davis, 2001), the social and emotional connections formed with others on these

ISGs are frequently found to be the most important predictor of the psychological

benefits (e.g. greater perceived well-being) derived from participating (Han et al., 2008;

Shim et al., 2011). For users in a Parkinson’s ISG, those who did not form these

connections appeared to have a more difficult time using it as a resource for coping and

information (Attard & Coulson, 2012). One study assessed users of diabetes and cancer

ISGs, and found that users’ preference for these ISGs was related to dissatisfaction with

their current face-to-face cancer and diabetes support groups (Chung, 2013), suggesting

that ISG use may be perceived by some as analogous to traditional support groups.

Consistent with the qualitative and empirical research on ISGs, one study (Leung,

2007) has found some support for the contention that specifically using the internet for

mood management and social compensation may be beneficial. Leung assessed 717

children and adolescents for internet use motives, stressful life events, and perceived

social support. It was found that higher levels of stress were related to higher degrees of

mood management and social compensation motives for using the internet. Leung also

found that higher perceived social support, both online and offline, was negatively

correlated with stressful life events. Based on the results found, Leung proposed that SIU

could serve as a buffer against stress by altering mood and providing social support, and

that some Internet users seemed to engage in SIU for this particular purpose. Although


5

cross-sectional in nature, this study, in addition to previously discussed research on ISGs,

suggests that those who might benefit from SIU are those who are highly distressed and

also willing to use the internet for social or emotional support. This research is also

backed by studies with Italian samples which have determined that assessment of

particular motives may predict negative and positive outcomes of SIU (Mazzoni,

Baiocco, Cannata & Dimas, 2016).

Several mental health factors related to distress have been shown to relate to SIU,

specifically use of the internet for social support and emotional coping. However, studies

have found conflicting results about the exact relationships among these factors, likely

due to methodological concerns: In a review of Internet Addiction studies, Byun et al.

(2009) noted that definitions and conceptualizations in this area are inconsistent. For

example, some Internet Addiction perspectives have adopted a framework based off of

gambling addictions (Young, 1998), while other perspectives have based the

conceptualization off of substance abuse disorders (Kaltiala-Heino, Lintonen, & Rimpela,

2004). This inconsistency applies to scale development, as well: Several popular

measures have been developed which do not assess similar hypothesized antecedents of

Internet Addiction (Morahan-Martin & Schumacher, 2000; Young & Rogers, 1998).

Approaches to model building are often hindered by low sample sizes and the

interpretation of poorly-fitting models (Byun et al., 2009). ISG studies are often largely

qualitative in nature (Barak et al., 2008), and there are a relative lack of studies on ISGs

for mental health concerns compared to ISGs for medical conditions (Griffiths et al.,

2008). Prevelance rates also vary highly between Western and Eastern samples,

depending on the measure used (Quinones & Kakabadse, 2015). Thus, this body of


6

research strongly suggests that more comprehensive studies should be conducted on SIU,

with a focus on stronger measures and methodology.

Which variables are most associated with social support seeking online?

Depressive Symptoms. Research on ISGs for individuals with depression have

demonstrated that depressed individuals specifically seek out emotional support when

they are online: A meta-analysis of thirteen studies on ISGs, conducted by Griffiths et al.

(2009), determined that the content of the posts for depression ISGs contained a

significantly greater degree of emotionally supportive content, compared to other kinds of

ISGs (e.g., Anxiety ISGs). Griffiths and colleagues found that many who participated in

ISGs for depression distinctly reported that an attractive feature of the groups was a

tendency to feel emotionally supported and to feel a reduction of their loneliness. Posts

on depression ISGs also frequently contained content related to social companionship

(Muncer, Burrows, Pleace, Loader & Nettleton, 2000). In general, members of

depression ISGs report significant benefit from participating (Houston, Cooper, & Ford,

2002).

Social Anxiety. SIU is argued to be an attractive mode of communication to

socially anxious individuals, particularly due to the relative anonymity afforded by the

Internet, the lack of self-presentational cues, and the ease of controlling the pace and tone

of conversations with others (Caplan, 2007; McKenna & Bargh, 2000). Research has

supported this contention: Studies have shown that individuals high in social anxiety are

rated more positively (e.g. conversation satisfaction, level of anxiety) when conversing

online compared to conversing face-to-face (High & Caplan, 2009). It is also

hypothesized that socially anxious individuals may converse easier online due to an


7

increased inability to react to negative or inhibitory feedback cues from others (Schouten

et al., 2007; Stritzke, Nguyen, & Durkin, 2004). Finally, it is argued that socially anxious

individuals may benefit from SIU as a low-risk approach to practicing social interaction

skills, in order to improve on their subsequent face-to-face interactions (Campbell et al.,

2006). However, one drawback for socially anxious individuals online may be the

possibility of experiencing poorer well-being, as online interactions do not fully

supplement the face-to-face interactions they may still desire (Weidman et al., 2012). In

general, socially anxious individuals appear to be motivated to engage in SIU: Increases

in social anxiety were found to be related to increased desire to use the Internet for

coping purposes (Gordon, Juang, & Syed, 2007).

Perceived Stress. Researchers have found a relationship between the presence of

specific stressful life events and motivation to use the Internet socially (Leung, 2007).

One study to date has assessed the role of general perceived stress specifically on the use

of the Internet to cope emotionally (Deatherage, Servaty-Seib, & Aksoz, 2014).

Deatherage and colleagues assessed 267 college seniors on their degree of perceived

stress, dispositional coping styles, motivations for internet use, and problematic internet

use. Motives involving management of negative affect (i.e. “to cheer up when I am in a

bad mood”) were strongly positively associated with perceived stress, while other

motives (e.g. because it is fun, to celebrate a special occasion with friends) were not

associated. Additionally, problematic internet use was not associated with perceived

stress. Although this study was correlational in nature, and thus did not provide support

for a directional relationship between perceived stress and SIU motivation, this study

provides some tentative support for the possibility of this relationship.


8

Neuroticism. Neuroticism, a trait defined by emotional instability and negative

affect (McCrae & Costa, 1987), is the personality trait most often related to increased

SIU. Studies have determined that a higher level of neuroticism is related to increased

use of social Internet sites like Facebook (Hughes, Rowe, Batey, & Lee, 2012; Seidman,

2013; Wolfradt & Doll, 2001) and to higher SIU in general (Kalmus, Realo, & Siibak,

2011). Individuals high in neuroticism also tend to be more motivated to engage in this

activity for social support and emotional expression: Highly neurotic individuals report

specific motivations to increase companionship and reduce loneliness through their

Internet use (Amiel & Sargent, 2004), to use blogs for self-expression (Guadagno, Okdie,

& Eno, 2008), and to express their “true selves” to others (Amichai-Hamburger,

Wainapel, & Fox, 2002; Tosun & Lajunen, 2010). It has been argued that highly neurotic

individuals prefer the Internet due to the greater control they have over their presentation

and their statements, compared to face-to-face interactions (Butt & Phillips, 2008;

Nadkarni & Hofmann, 2012). This suggests that an online format is specifically

attractive to neurotic individuals who wish to form new relationships.

Extraversion. Both extraversion, a trait defined by higher sociability, liveliness,

and assertiveness, and introversion, the inverse trait (McCrae & Costa, 1987), have been

demonstrated by research to increase motivations for SIU. Researchers argue that the

relations found for both dimensions of this trait are due to the types of social Internet

activities being assessed: Extroverts prefer to use the Internet to build off of existing

relationships, while introverts prefer to use the Internet to make new friends (Amichai-

Hamburger et al., 2002; Bargh, McKenna, & Fitzsimons, 2002; Orchard & Fullwood,

2010; Tosun & Lajunen, 2010). For example, extroverted individuals tend to report more


9

participation on Facebook in general (Jenkins-Guarnieri, Wright, & Hudiburgh, 2012),

and tend to be more involved on other sites that focus primarily on previously-established

friendships (Amichai-Hamburger, Kaplan, & Dorpatcheon, 2008). Introverted

individuals, on the other hand, tend to value more anonymous Internet services such as

ICQ and Instant Messaging (Amichai-Hamburger et al., 2008; Amiel & Sargent, 2004),

and appear to prefer SIU as a means of expressing their “true selves” (Amichai-

Hamburger et al., 2002; Zywica & Danowski, 2008). This research suggests that more

introverted individuals are motivated to seek out and build new relationships online.

Rationale for the Present Study

Although there are varying levels of empirical support for each of the factors that

were previously discussed, no overarching model has attempted to show the multivariate

relations between all of these factors. This conclusion holds for these factors’

relationship with healthy, non-excessive degrees of social and emotional coping on the

Internet. Most prominent research on models of Internet use in general focuses primarily

on this behavior as inherently aberrant and harmful, especially when considering the

social components (Caplan, 2002, 2010; Davis, 2001). These models persist, despite a

strong, growing body of literature demonstrating that there are numerous potential

benefits to engaging with others online (Kraut et al., 2002). Moreover, these models are

somewhat limited in scope, are frequently exploratory in nature, and are sometimes

interpreted even when there is poor fit (Byun et al., 2009; Tokunaga & Rains, 2010).

When assessing the possibility of positive outcomes, research tends to be qualitative

(Barak et al., 2008; Griffiths et al., 2009), and studies that have empirically assessed the

relations between factors in a positive manner tend to only look at two or three factors at


10

a time (Leung, 2007). Thus, the relations among factors involved with SIU is still a

nascent research area. There remains a need for a conceptual SIU model which does not

treat this behavior as inherently harmful.

This study will contribute to the literature in this area by being the first of its kind

to create a more comprehensive, empirical model of the personality traits and

psychosocial factors which have been shown to predispose individuals to SIU. This

study will also explore through a confirmatory approach the multivariate relations

between these variables, some of which have not been assessed yet in the literature.

Finally, this study will assess SIU as a possible positive coping mechanism, in contrast to

other major models of Internet use which assert that it is inherently harmful.

The Internet is becoming more social in nature, and individuals are increasingly

starting and maintaining relationships online. This is the first study, to our knowledge, to

assess the roles of, and relations between, multiple empirically-supported factors (i.e.,

depressive symptoms, neuroticism, etc.) in SIU as a positive social and emotional coping

mechanism. In determining the relationship between these factors, this study hopes to

further increase understanding of what might influence the attractiveness of engaging in

SIU for support, and who might find it most appealing and helpful. A strong relationship

between these factors would demonstrate that individuals have multifaceted emotional

and social motivations for using the Internet to meet others. Consistent with some

therapeutic orientations (Palmer-Olsen, Gold, & Woolley, 2011; Rogers, 1961),

researchers have argued that the ability to express one’s “true self”, even over the

Internet, could be helpful for neurotic, introverted, or socially anxious individuals, as

these individuals often have difficulty expressing themselves in face-to-face


11

communication (Amichai-Hamburger et al., 2002; Bargh et al., 2002; Ebeling-Witte,

Frank, & Lester, 2007). Some research has found initial support for this contention,

finding that even the use of blogs for self-expression leads to positive outcomes

therapeutically (Hillan, 2003), possibly due to an increase in perceived social support

(Baker & Moore, 2008).

Although there is not yet a complete consensus on the actual positive or negative

consequences of SIU (Huang, 2010), a greater understanding of the factors behind these

behaviors may still benefit Internet users. SIU could potentially allow many individuals

to seek social and emotional support when it is not available through other means.

Additionally, a more comprehensive understanding of these variables may benefit

therapists and other mental health practitioners. By having a greater understanding of

these psychosocial factors as they relate to SIU, therapists may become better equipped to

work with clients who are seeking and maintaining relationships with others online. This

will become increasingly relevant in psychotherapy, as this population is only expected to

grow in the future. Understanding these behaviors as potential coping strategies and

methods of self-expression for those experiencing stress, depression, social difficulties, or

emotional instability may allow for more accurate conceptualizations and interventions.

Future application of SIU as both a social and emotional support mechanism, for certain

types of clients, is also a possibility.

Hypotheses

The proposed structural model (See Figure 1, pg. 100) will be utilized to assess

the hypothesized relations between these variables at a multivariate level. It is proposed

that the relationship between personality characteristics (neuroticism and introversion)


12

and SIU motivation is mediated by certain psychosocial factors (social anxiety,

depressive symptoms, perceived stress and the tendency to seek out emotional social

support to cope).

Thus, this study hypothesizes:

H1: SIU motives will mediate the relationship between positive and negative outcomes

and psychosocial factors.

H2: SIU motives will have positive relations with both negative and positive outcomes.

H3: Psychosocial factors will have positive relations with SIU motives.

H4: Psychosocial factors will relate positively with neuroticism and negatively with

extraversion, apart from emotional social support coping, which will relate positively to

both.

H5: The relationship between SIU motives and personality characteristics will be

mediated by psychosocial factors.

Research supports the direct relations between latent factors in this model. Meta-

analyses of personality and clinical disorders conclude that neuroticism is frequently one

of the strongest predictors of anxiety and mood disorders (Kotov, Gamez, Schmidt, &

Watson, 2010; Malouff, Thorsteinsson, & Schutte, 2005). Neuroticism is highly

correlated with negative affect in general (Gutierrez, Jimenez, Hernandez, & Puente,

2005), as well as the recurrence of depressive symptoms (Steunenberg, Braam, Beekman,

Deeg, & Kerkhof, 2009). The Vulnerability model, which states that certain personality

factors precede depression, is also one of the more strongly supported models in the

literature (Bagby, Quilty, & Ryder, 2008). Thus, neuroticism can be hypothesized to


13

precede the development of psychosocial difficulties, and the model has included an arc

between neuroticism and social anxiety, and neuroticism and depressive symptoms.

Neurotic individuals also display a tendency to become distressed more easily than their

non-neurotic counterparts (McCrae & Costa, 1987), and to cope in a different manner

than their non-neurotic counterparts (Gunthert, Cohen, & Armeli, 1999). Thus, paths

between neuroticism and perceived stress and neuroticism emotional social support

coping have also been included in the model.

Introversion, a trait defined by less sociability and liveliness in social situations, is

strongly correlated with social phobia, unipolar depression, and dysthymic disorder

(Kotov et al., 2010), and is found to some degree among a wide variety of clinical

symptoms (Malouff et al., 2005). Introversion is also associated with lower levels of life

satisfaction (Lounsbury, Saudargas, Gibson, & Leong, 2005). Consistent with this

research and the Vulnerability model (Bagby et al., 2008), paths between introversion and

depressive symptoms, and introversion and social anxiety, are included in the model.

Although the relationship between introversion and emotional social support coping is

less clear, there does appear to be a relationship (Watson & Hubbard, 1996), thus a path

was also included between these two variables as well.

Coping is inherently a strategy used to manage the demands of situations that are

perceived as stressful, and is typically initiated when experiencing intense negative

emotions or distressing events (Folkman & Moskowitz, 2004). Thus, the model includes

covariances between emotional social support coping and perceived stress, depressive

symptoms, and social anxiety. Although anxiety disorders tend to largely occur before a

comorbid mood disorder (Kessler, Stang, Wittchen, Stein, & Walters, 1999; Mineka,


14

Watson, & Clark, 1998), the causal relationship between sub-clinical social anxiety and

depressive symptoms is less clear. Thus, these factors are hypothesized to covary in the

model. The perception of stress is also hypothesized to covary with depressive symptoms

and social anxiety.

It is predicted that a tendency to use emotional social support coping will

inherently lead to more motivation to engage in this activity online. Some research

supports this relationship: Seepersad (2004) found that online and offline coping

strategies shared a significant degree of overlap. Specifically, it was found that

adolescents who considered communication to be an important aspect of Internet use

were also likely to cope offline with loneliness through emotional expression and social

coping. Thus, a path is included from emotional social support seeking to SIU

motivation.

Prior research discussed has shown strong support for the contention that

particular individuals will be more motivated to use the Internet socially. Internet

Support Group studies demonstrate that depressed individuals maintain specific

motivations to use the Internet for social and emotional support (Griffiths et al., 2009).

Individuals with social anxiety appear to benefit from the conduciveness of the Internet to

self-presentational concerns (Caplan, 2007; McKenna & Bargh, 2000). Additionally,

they appear to specifically desire SIU as a coping mechanism (Gordon et al., 2007).

Stressful life events, or even the perception of stress, may also lead to an increased

motivation to seek others out online (Deatherage et al., 2014; Leung, 2007). In contrast

to this research, Davis (2001) and Caplan (2002, 2010) argue and attempt to demonstrate

that SIU and motivation to engage in it is an inherently harmful activity, which leads to


15

negative outcomes in work and school. Thus, there is no current consensus on the relative

benefits and drawbacks of this behavior. As causality will not be assessed in this study, a

research question will be posed:

R1: Do motivations for SIU have a stronger relationship with positive outcomes

or with negative outcomes?

Methods

Sample

Participants were 293 college students (212 identified as female, 80 identified as

male, 0 identified as non-binary) from introductory psychology classes at a Southwestern

university. Participants ranged in age from 17 to 33 years old (M= 19.4, SD=1.7). To

avoid limiting the accuracy of demographic information, a general “multi-racial”

category was eschewed and participants were allowed to choose any racial identities they

wished to report. Thirty participants identified with more than one race (10.2%).

Including multi-racial individuals, 193 participants identified as White/Caucasian

(65.8%), 31 identified as Black/African-American (10.5%), 19 identified as Asian

American/Asian/Pacific Islander (6.4%), 74 identified as Hispanic/Latino (25.2%), 2

identified as Indian or Middle-Eastern (.6%), and 6 identified as Native American (2%).

The majority of the sample were college freshman (186; 63.4%) or sophomores (62;

21.1%). Forty-six percent of the sample reported currently being in a romantic

relationship.

Procedure

Participants were recruited through SONA, an online study sign-up program.

They received course credit for their participation. Participants were not penalized for


16

failing to complete the study once they have agreed to participate. The current study

received exempt status from the Institutional Review Board (IRB), indicating that

participants did not experience any risks outside of those related to normal life.

Instruments were adapted to online surveys on Qualtrics. Once the participants

signed up for the study on SONA, they were then directed to Qualtrics to complete the

study. They read a webpage containing an informed consent regarding the purpose of the

study (“To gain a greater understanding of why people might seek out new friends and

relationships online”), and were required to indicate their consent to participate via a

checkbox before they were able to continue. Participants first completed a demographics

questionnaire which included various Internet use behaviors they engaged in on a daily

basis (see Appendix C). They then completed the remaining seven measures

(Appendices D - H) in a random order, to account for any order effects. These

questionnaires took approximately 35 to 45 minutes to complete. Upon completion,

participants were directed to a separate webpage which asked them for their contact

information, thanked them for their participation, and provided contact information for

the researcher. Through this method, participants were not matched with their responses

and could remain anonymous.

Instruments

Demographic questionnaire. A short demographic questionnaire was used and

consisted of items about gender, age, ethnicity, and number of years in college. Various

popular online communities and online social behaviors were assessed to provide

qualitative impressions of the degree of SIU in the sample. Examples of behaviors and

communities measured include Facebook use, microblog use (i.e. Tumblr), use of


17

message boards, use of blogging software (either reading or writing on blogs), and use of

messaging software. The amount of time spent per day on each activity, on average, was

also assessed. Assessing amount of use at this level is beneficial, in that it allows for

some general comparison between samples. However, Internet use can vary widely and

the reliability of this estimate in the sample is anticipated to be low.

Center for Epidemiological Studies – Depression Scale (CES-D) (Radloff,

1977). The CES-D is a widely used assessment of depressive symptoms (Nezu, Nezu,

Friedman, & Lee, 2009). The CES-D asks participants 20 questions about the intensity

of depressive symptoms over the previous week (e.g., I thought my life had been a

failure). Items are rated on a 4-point Likert scale (1 = Rarely or none of the time, 4 =

Most or all of the time). Rarely or none of the time responses are scored as a 0 and Most

or all of the time responses are scored as a 3, for total score range between 0 and 60, with

higher scores indicate more severe symptomology.

The CES-D scale scores demonstrated high internal consistency with White

populations (α = .84-.85) and clinical samples (α = .90) in the original study (Radloff,

1977). Of particular interest to the current study, the CES-D has good predictive validity

and scale discriminability in English and French-Canadian college samples (Santor,

Zuroff, Cervantes, & Palacios, 1995; Shean & Baldwin, 2008). Additionally, online

versions of the English CES-D appear to have similar psychometric properties to the

paper-and-pencil version (Ogles, France, Lunnen, Bell, & Goldfarb, 1998). Scale scores

have also been shown to have good factorial validity (Orme, Reis & Herz, 1986), as well

as convergent validity with other popular measures of depression, such as the Beck

Depression Inventory (Santor et al., 1995). The English scale has been determined to


18

demonstrate unidimensionality when the reverse-scored items are removed (Stansbury,

Ried, & Velozo, 2006). Due to its good psychometric properties among many different

populations, the CES-D is recommended for the assessment of depressive symptoms

(Nezu et al., 2009) and was used to measure this construct in the current study.

Internet Motivation Scale (Wolfradt & Doll, 2001). SIU motivations were

assessed with the Internet Motivation Scale (IMS). The Internet Motivation Scale is a

measure intended to assess various motivations participants might have for using the

Internet. This scale includes 20 items which assess 3 different underlying motivations for

using the Internet: information seeking motives (e.g., The Internet updates me on new

trends), entertainment motives (e.g., The Internet stimulates my curiosity), and

interpersonal communication motives (e.g., The Internet is to me a substitute for other

social contacts). Higher scores on this scale signify more motivation to engage in the

Internet in this manner, e.g. a high score on the interpersonal communication motives

subscale indicates greater motivation to use the Internet for this purpose.

Wolfradt and Doll (2001) found high internal consistency for scores of all three

scales (α = .76 - .84) and discriminate validity to personal factors (e.g. self-efficacy) and

social factors (e.g. expectations of others). Other studies have supported the IMS’s high

internal consistency (α = .89) (Gordon et al., 2007) and its three-factor structure in

English-speaking samples (Matsuba, 2006). Items from the interpersonal communication

factor were used in this study, but all items were assessed in order to preserve the

psychometric properties of the scale and provide additional qualitative data.

Leibowitz Social Anxiety Scale (LSAS) (Liebowitz, 1987). The Leibowitz

Social Anxiety Scale is a widely used, clinician-administered assessment of social


19

anxiety. The LSAS contains 24 items which assess a range of situations that individuals

with social anxiety may have feared or avoided in the last week (e.g. Eating in public

places, or Looking at people you don’t know very well in the eyes). These items are

divided into two subscales which address Social interaction situations and Performance

situations, and measures participant’s Avoidance and Fear of each type of situation. Items

are rated on a 4-point Likert-scale (e.g. For Fear, 0 = None and 3 = Severe. For

Avoidance, 0 = Never and 3 = Usually), for a total score range between 0 and 144. An

overall total score can be calculated by summing all items, with higher scores indicating

more severe symptomology.

The English LSAS has been shown to have excellent internal consistency (α =

.96), as well as convergent validity with other measures of social anxiety, and

discriminate validity to measures of general anxiety and depression (Heimberg et al.,

1999). The self-report version of the English LSAS was used in the current study, as the

self-report LSAS has comparable psychometric properties to the original, clinician-

administered version (Fresco et al., 2001).

Perceived Stress Scale (PSS-10) (Cohen, Kamarck, & Mermelstein, 1983).

The Perceived Stress Scale is a widely used assessment for nonspecific perceived stress.

The PSS-10 is a version of the Perceived Stress Scale which contains 10 items that assess

participant’s perceptions of how unpredictable, uncontrollable, and overloading they find

their lives (e.g. In the last month, how often have you found that you could not cope with

all the things that you had to do?). Items are rated on a 5-point Likert-scale (e.g. 0 =

Never, 4 = Very often), for a total score range between 0 and 40.


20

The PSS-10 has been shown to have better psychometric properties than other

versions of the PSS, and is recommended as a result (Lee, 2012). In general, the PSS-10

has good internal consistency (α = .78-.91), test-retest reliability (α =.85), and convergent

validity with other measures of stress in English-speaking populations (Cohen et al.,

1983; Cohen & Williamson, 1988). The reliability of the PSS-10 extends to nationally

representative samples (Cohen & Janicki-Deverts, 2012). Reviews of the PSS-10 have

demonstrated some variation between studies with regard to the strongest factor solution,

alternating between a one-factor and a two-factor solution (Lee, 2012). Thus, both one-

factor and two-factor solutions were assessed in the current study.

The Big Five Inventory (BFI) (John, Donahue & Kentle, 1991) Neuroticism

and introversion were assessed with The Big Five Inventory. The Big Five Inventory is a

widely used assessment of personality traits, based on the Big Five model. The BFI

contains 44 items which assess participants’ perceptions of themselves (e.g. I am

someone who is reserved, or I am someone who can be moody). Items are rated on a 5-

point Likert-type scale (e.g., 1 = Strongly Disagree, 5= Strongly Agree), and items

corresponding with each Big Five trait are averaged into five scale scores.

The English BFI has been shown to have high internal consistency (α = .83-.85),

convergent validity with other, highly validated Big Five measures such as the NEO,

strong discriminant validity, and substantial self-peer agreement (John, Naumann, &

Soto, 2008; John & Srivastava, 1999; Soto & John, 2009). Domain scales are also

reliable (α =.81-.88), and a five-factor structure is supported (John, Naumann, & Soto,

2008; Soto & John, 2009). Although only items for the neuroticism and introversion


21

scales were assessed, the entire scale was given to preserve its psychometric properties

and provide further descriptive data for this sample.

COPE (Carver, Scheier, & Weintraub, 1989). The COPE is a widely used

assessment of coping. The full version of the COPE has 15 distinct scales which assess

various ways one might cope with stressful situations (i.e., Humor, Denial, Behavioral

Disengagement, etc.). The COPE can be modified to assess coping behaviors during a

particular time point, or to assess coping as a more dispositional “trait”. Emotional social

support seeking will be assessed as a dispositional trait through the 4 items on the Use of

emotional social support scale on the COPE (e.g. I discuss my feelings with someone, or I

talk to someone about how I feel). Items are measured on a 4-point Likert scale (e.g., 1 =

I usually don’t do this at all, 4 = I usually do this a lot).

In the dispositional form, the English versions of these items have demonstrated

good internal consistency (α =.85 - .90) (Carver et al., 1989; Cook & Heppner, 1997), and

have consistent support for their unidimensionality (Litman, 2006). When assessed as a

three-factor solution, three of the four items (one was removed due to perceived overlap

with another item) loaded moderately (=.56 - .67) onto one factor (Lyne & Roger, 2000),

and all four items were found to load onto one factor (=.71-.83) in a seven-factor solution

(Eisengart et al., 2006).

Generalized Problematic Internet Use Scale – 2 (GPIUS2) (Caplan, 2010).

Negative outcomes of Internet use were assessed with the Generalized Problematic

Internet Use Scale - 2. The GPIUS2 is an updated version of a widely-used assessment

of characteristics associated with excessive Internet use. The GPIUS2 comprises 5

subscales with 3 items each: Preference for online social interaction (e.g., I prefer online


22

social interaction over face-to-face communication), mood regulation (e.g., I have used

the Internet to make myself feel better when I was down), cognitive preoccupation (e.g., I

would feel lost if I was unable to go online), compulsive Internet use (e.g., I find it

difficult to control my Internet use), and negative outcomes (e.g., My Internet use has

created problems for me in my life). Items are rated on an 8-point Likert scale

(1=Definitely Disagree, 8=Definitely Agree). The GPIUS-2 was developed to update the

previous measure (GPIUS) and incorporate more recent research findings, and can be

used as a set of sub-scales or as a general composite score (Caplan, 2010). The GPIUS-2

has demonstrated reasonable psychometric properties in some previous research: It has

been shown to have high internal consistency in American samples (α = .91), as well as

adequate construct validity (Caplan, 2010). Currently, the psychometric properties of the

full GPIUS-2 have not yet been replicated in American sample, so these properties are

less established. It has been translated into Spanish (α =.90) (Gamez-Guadix, Orue,

Smith, & Calvete, 2013) and Italian (α =.89) (Chittaro & Vianello, 2013). The GPIUS-2

has also been shown to have adequate construct and convergent validity in Mexican

adolescent samples (Gamez-Guadix, Villa-George, & Calvete, 2012).

Social Connectedness Scale - Revised (SCS-R) (Lee, Draper, & Lee, 2001).

Positive life outcomes were assessed through a modified version of the SCS-R. The

scale contains 20 items, rated on a 6-point Likert scale (1 = Strongly Disagree, 6 =

Strongly Agree), which purport to measure cognitions regarding general, enduring

interpersonal closeness in the social world (e.g., I feel distant from people). The SCS-R

has good internal consistency (α =.92), as well as appropriate convergent and

discriminant validity (Lee et al., 2001). Although no known replication of the scale has


23

been done in American college populations, the translated SCS-R has good internal

consistency (α =.91) and high test-retest reliability in Taiwanese college student

populations (Chen & Chung, 2007). The SCS-R will be modified to assess social

connectivity with regard to general SIU (e.g., I feel close to people will be modified to I

feel close to people online). The SCS-R has previously been adapted to assess social

connectivity regarding Facebook use specifically (Greive et al., 2013), and demonstrated

good internal consistency in this particular study (α =.89 - .92). However, this measure

has less established psychometric properties and the modified version will be examined

before use in analyses.

Marlowe-Crowne Social Desirability Scale (MCSDS) (Marlowe & Crowne,

1960). Social desirability was measured with the Marlowe-Crowne Social Desirability

Scale. The MCSDS is a 33-item scale which purports to measure the desire to respond to

items in a manner which is appropriate and acceptable in that individual’s culture. It is

the most widely used instrument in this area, and is often utilized to assess undergraduate

populations (Beretvas, Meyers, & Leite, 2002). Stigma related to certain behaviors, such

as SIU, as well as psychosocial characteristics, such as depression and neuroticism, has

been shown to impact the responses of participants in studies. Thus, this scale was

utilized to determine the impact of this construct on responding, in particular related to

Internet use behaviors. It was also used as an additional validation measure to screen

participants who are possibly under-reporting based on social desirability. The original

study by Marlowe and Crowne (1960) determined good internal consistency (α = .88),

though this was based on a relatively small sample size. The scale has been shown to

have adequate reliability for adult women (α = .797) and adult men (α =.704), based on a


24

meta-analysis of studies which have reported sample-specific internal consistency

statistics (Beretvas et al., 2002). This study also determined a wide range of test-retest

reliabilities for this measure, ranging from poor (α = .38 for a study with a 2-4 week test-

retest interval) to good (α = .86 for a study with a test-retest interval of more than one

month). The full scale is recommended for use, as it has the best psychometric properties

(Barger, 2002).

Some studies have questioned the factor structure and construct validity of this

measure with undergraduate populations (Barger, 2002; Leite & Beretvas, 2005), and

many studies have not reported sample-specific internal consistency estimates (Beretvas

et al., 2002). Although the dubious properties of this measure have emerged in some

samples, these items appear to have adequate internal consistency for the current study (α

= .73). Thus, a high cutoff score ( > 3 SD) will be utilized to avoid some of the concerns

demonstrated in the literature, but to also allow for determination of those cases most

likely to be impacted by response bias.

Results

Descriptive Statistics

Descriptive statistics and other qualitative information related to this sample’s

reported SIU activities can be found in Table 5. Descriptive statistics related to each

instrument used in the current study may be found in Table 2, and correlations between

each instrument may be found in Table 3.

Structural Equation Modeling

Structural equation modeling (SEM), with weighted least squares (WLSMV)

estimation, was utilized to determine the relations between SIU motivation, negative and


25

positive outcomes of this behavior, psychosocial factors (depression, social anxiety,

perceived stress, emotional social support seeking), and personality factors (neuroticism

and introversion). SEM has a number of strengths which make it the most appropriate

technique to assess these relationships. Of primary interest to the current study, SEM

allows for a direct assessment of the hypothesized model and the overall relations

between multiple variables, while linear regression would be relegated to the assessment

of several individual correlations (Tomarken & Waller, 2005). SEM also uses multiple

indicators for each construct, which increases the reliability of factor measurement over

path analysis (Kline, 2011).

SEM with weighted least squares estimation cannot be successfully utilized unless

important assumptions are achieved or considered. The first assumption is that of

multivariate normality, or each indicator being distributed normally at each value of each

other indicator (Kline, 2011). The second assumption is that outliers have not impacted

data and altered model fit as a result. Data screening suggested adequate univariate

normality, but determined the presence of two multivariate outliers, which were removed

from analyses.

The influence of social desirability on participant responses was considered in

data screening. As social desirability may impact the willingness of some subjects to

respond in a forthright manner about subjects such as Internet use, this variable was

assessed to determine potential univariate outliers that would otherwise not be present in

analyses. However, only high scores (Greater than 3 SD) were considered, as this

measure does not appear to have strong psychometric properties (Barger, 2002; Beretvas

et al., 2002; Leite & Beretvas, 2005). One score on this measure was 3 standard


26

deviations above the mean (M = 16.05, SD = 4.86) and this case was removed from

analyses.

SEM analyses were conducted with Mplus (Muthén & Muthén, 1998-2010). Data

analysis was performed in three steps. First, item parcels were created for each latent

factor in the model. Then, the measurement model was examined to determine if item

parcels adequately measured latent factors. Third, the structural model was examined to

analyze the relations among the latent factors.

Measures of fit. Overall fit of the model was assessed partially through Root

Mean Square Error of Approximation (RMSEA) and the Comparative Fit Index (CFI).

These fit indices take into account sample size and complexity of the model, and are thus

recommended for large models (Kline, 2011). However, Kline also notes that fit indices

are only able to provide an overall picture of the model, and may neglect particularly

poor fit in some areas of the model if other areas fit well. Thus, fit indices may report

that a model fits the data adequately, when in reality it does not. Critics of the perceived

overuse of fit indices have demonstrated that perfectly fitting models may actually

account for very little variance (less than 1%) in the endogenous variables (Tomarken &

Waller, 2003). Finally, common cutoffs for these indices are primarily “rules of thumb”,

and are not necessarily accurate. However, these fit indices may still provide valuable

qualitative information about the fit of the model. Thus, these fit incidences are

interpreted and reported with caution, and overall model fit is assessed primarily through

the researcher’s judgment (Kline, 2011). For the structural model, in addition to

researcher judgment, “lower-order” components of the model are taken into greater

account in assessing model fit, which is a recommended approach (Kline, 2011;


27

Tomarken & Waller, 2005). Thus, lower-order model components were assessed, such as

the model chi-square, path coefficients, the effect sizes of the direct and indirect arcs

between factors, and the total variance accounted for by the model.

As some constructs of interest did not have multiple, validated instruments

available for inclusion in this model (e.g. positive and negative outcomes of SIU), each

latent factor was assessed through one instrument, which necessitated the use of item

parcels to provide the necessary number of predictors. Unidimensionality is an important

assumption in the creation of item parcels (Little et al., 2002). Thus, each instrument

was assessed at item-level before items were parceled, to determine undimensionality.

Groups of items with factor loadings all above .5 were considered unidimensional. For

CES-D items, the four reverse-scored items were removed apriori, consistent with

recommendations made by previous research (Stansbury, Ried, & Velozo, 2006) that the

CES-D demonstrates unidimensionality afterward. This was supported by the

determination that all items did not load consistently on a single factor, while the

remaining confirmative items did. Consistent with literature (Lee, 2012), PSS-10 items

were assessed as a 1-factor solution with reverse-scored items removed, a 1 factor-

solution with all items retained, and a 2-factor solution. The 2-factor solution appeared to

fit the data best, and was retained.

The chi-square statistic was significant for this model (<.05). This indicates that

the model does not perfectly fit the data and suggests that the model should be rejected.

However, the chi square test statistic is highly sensitive to minor changes in model fit

with large sample sizes, higher correlations between observed variables, and models with

high degrees of freedom (Kline, 2011). Thus, there is some debate regarding the role of


28

this statistic in interpreting models (Kline, 2011). As this current study has a somewhat

large sample size (n = 293), some high correlations, and a model with high degrees of

freedom (df = 442), this statistic should be interpreted with caution, particularly if lower-

order components of the model seem to support its fit to the data.

The CFI test statistic is recommended to have a cutoff of .90 for “adequate fit”

and a cutoff of .95 for “good fit” (Kline, 2011), and these cutoffs will be utilized for the

current study. The RMSEA test statistic is recommended to have a value of .08 or below

to exhibit adequate fit. This statistic also has a 90% confidence interval which indicates

good generalizability when the absolute value of the intervals is smaller.

Measurement model. Assessment of the hypothesized model in SEM first

requires analysis of the measurement model (Figure 1). This model assesses the

covariances between the indicators and the latent factors. MPlus utilizes the Weighted

Least Squares approach by default for models with both categorical and continuous

variables, and this approach was retained for the purpose of this study (Muthén &

Muthén, 1998-2010). The hypothesized model was theoretically identified. This model

also met the “necessary but not sufficient” conditions (Wang & Wang, 2012, p. 12) for

empirical identification: Each factor had at least three indicators (through item parceling,

discussed below), and the number of free parameters in the model (45) was exceeded by

the number of observations (136). The scale of each latent factor was set using unit

loading identification (ULI; Kline, 2011), such that the scale of one indicator per latent

factor was set to 1. Each latent factor was assumed to be correlated with each other latent

factor, and errors were assumed to be uncorrelated.


29

Item parcels. Given the demonstration of adequate unidimensionality, item

parceling was used to create predictors in the measurement model. Item parceling is a

technique in which scale items are combined and used as predictors, instead of full scale

scores. This is a recommended technique over the use of individual items because it

allows for more continuous predictors, reduces the impact of sampling error, has greater

reliability, and increases parsimony of the measurement model (Little, Cunningham,

Shahar, & Widaman, 2002). Items were removed if intercorrelations between that item

and other items differed noticeably, if the item appeared to have redundancy with another

item, if the item loaded poorly in a 1-factor solution, and if researcher judgment

otherwise determined that the item was inadequate for parceling purposes. After

determining all items that met required criteria for inclusion, these items were placed into

parcels sequentially (e.g. 1,2,3,1,2,3…), unless otherwise noted. These predictors were

then placed into the measurement model sequentially, in order to determine potential

specification errors or problematic variables within the model. Composite reliabilities

were calculated rather than alpha coefficients, as these represent the correlations among

latent variables within the measurement model and more adequately describe the

psychometric properties of the item parcels. See Table 4 in Appendix B for initial fit and

final fit of chosen items, as well as list of items retained.

COPE. As there were only four items from the COPE used for emotional social

support coping, each item was retained and was modeled as a categorical predictor

instead of a continuous predictor. Due to the categorical nature of the predictors, this

measure was assessed first, in order to make sure that specification of the model was

correct. These items demonstrated unidimensionality, consistent with previous research


30

(Litman, 2006). The measurement model demonstrated somewhat poor fit to the data:

Chi-square = 11.615 (df = 2, p=.003), RMSEA = .13 (.064 - .204), CFI = 1.0. The CFI

statistic indicates perfect fit for the initial stages of the model, likely due to a combination

of low degrees of freedom and the steps taken in item parceling to ensure that only high

validity items are included in the model.

CES-D. Consistent with previous research (Stansbury, Ried, & Velozo, 2006),

negatively worded items were removed. Then, redundant items were removed, using

differences in intercorrelations and face validity as criteria (e.g. “I felt sad” and “I felt

depressed”). Three parcels were created out of remaining items. This latent factor and

its predictors were then added it in to the model with COPE items, leading to a

measurement model with excellent fit, Chi-square = 18.84 (df = 13, p=.13), RMSEA =

.039 (.00 - .08), CFI = 1.0.

IMS. For IMS items, none of the 7 items were removed, as these items appeared

to demonstrate unidimensionality. One parcel with 3 items was created, and 2 parcels

with 2 items were created. The IMS latent factor was added into the model, leading to a

model with excellent fit: Chi-square = 30.55 (df = 32, p=.54), RMSEA = .000 (.000 -

.040), CFI = 1.0. RMSEA being set to 0 is a product of calculating this statistic, as this is

the only step in the model building at which the degrees of freedom are greater than the

Chi-square statistic.

BFI. For BFI neuroticism items, 2 items were also removed, both of which were

reverse-scored. The remaining reverse-scored item, and five positively-scored items,

were determined to have consistent intercorrelations with each other, were retained, and

were paired to create three parcels. This latent factor was added into the model, leading


31

to a model with adequate fit: Chi-square = 96.29 (df = 59, p=.00), RMSEA = .046 (.029 -

.063), CFI = .99.

For BFI extraversion items, 2 items were removed, one of which was reverse-

scored and one of which was positively-scored. The remaining six items were split into

three parcels, with one reverse-scored item paired with one positively-scored item. This

latent factor was then added into the model, leading to a model with adequate fit: Chi-

square = 186.01 (df = 94, p=.00), RMSEA = .058 (.046 - .070), CFI = .98.

PSS-10. Five item parcels were created by pairing each of the 4 reverse-scored

items with a positively-worded item, then pairing up the two remaining positively-

worded items. This latent factor was then added into the model, leading to a model with

adequate fit: Chi-square = 346.61 (df = 174, p=.00), RMSEA = .058 (.049 - .067), CFI =

.96.

LSAS. For LSAS items, the subscale which assessed Fear related to Social

Interaction was utilized for item parcel creation, as it was the most subscale most

conceptually similar to socially anxious individuals’ theorized motivation to engage in

SIU. Nine items were retained, and were distributed into three parcels. This latent factor

was then added into the model, leading to a model with adequate fit: Chi-square = 426.75

(df = 231, p=.00), RMSEA = .054 (.046 - .062), CFI = .95.

SCS-R. For modified SCS-R items, six items with adequate factor loadings and

similar intercorrelations were retained out of the original 20 items, and were distributed

into 3 parcels. This latent factor was then added in to the model, leading to continued

adequate fit: Chi-square = 517.68 (df = 296, p=.00), RMSEA = .051 (.043 - .058), CFI =

.95.


32

GPIUS. For GPIUS items, each subscale’s items were combined into a

subsequent parcel, for five parcels total. This final latent factor was added in to the

model, leading to a final measurement model: Chi-square = 703.33 (df = 428, p=.00),

RMSEA = .047 (.041 - .053), CFI = .93.

Measurement model fit. As demonstrated above, this model demonstrated

adequate fit to the data. The model chi-square was significant, indicating that the model

did not perfectly fit the data. However, factor loadings were uniformly high and in the

predicted direction: All factor loadings were above .6 apart from one loading on

Extraversion, with no cross-loadings detected (See Table 1 for factor loadings and

composite reliabilities of parcels).

Table 1: Factor Loadings and Composite Reliabilities for each Latent Factor

Latent Factor Factor Loadings Composite Reliability

Depression .90, .84, .79 .88

SIU Motivation .89, .80, .64 .82

Emotion Social Support Coping .92, .89, .80, .95 .94

Social Anxiety .86, .79, .80 .86

Perceived Stress .65, .70, .60, .80, .71 .78

Neuroticism .85, .68, .69 .78

Extraversion .75, .59, .72 .73

Positive Outcomes .69, .83, .73 .80

Negative Outcomes .72, .75, .78, .79, .80 .85

Contrary to expectations, Emotional Social Support Coping was not significantly

correlated with Depression (.00), Perceived Stress (.04), or Social Anxiety (.01). Model

results were retained without further specification, due to strong and expected path

loadings, no cross-loadings, no correlation of errors, and adequate overall fit of the

model.


33

Structural model

As the measurement model was retained, the final step of analysis involved

assessment of the hypothesized structural paths between latent factors in the

measurement model, including disturbance terms. The hypothesized structural model

(Figure 2) described the relationship between Positive and Negative Outcomes and

psychosocial factors (Depression, Social Anxiety, Emotional Social Support Coping, and

Perceived Stress) as mediated by SIU Motivation, and the relationship between SIU

Motivation and personality factors (Neuroticism and Extraversion) as mediated by these

psychosocial factors.

The initial hypothesized model (Figure 3) converged upon a solution and fit

reasonably well to the data: Chi-square = 801. 067 (df=442, p=.000), CFI = .908,

RMSEA = .053 (.047 - .058). Relations between Extraversion and Depression (-.103)

and Extraversion and Perceived Stress (.075) were not significant, contrary to

expectations. Relations between SIU Motives and Perceived Stress (.078) and Emotional

Social Support Coping (.056) were also not significant, contrary to expectations. Finally,

depression was correlated with Perceived Stress, but no other psychosocial factors were

correlated with each other, contrary to expectations. All other paths were in the expected

direction and level of significance. Negative Outcomes and Positive Outcomes were not

correlated (-.112). The correlation between Negative Outcomes and Motives approached

a value of 1 (.97), indicating possible multicollinearity.

The saturated model was assessed to determine potential relations unaccounted

for within the hypothesized model and determine possible reasons for the

multicollinearity which occurred during model building. Paths between social anxiety


34

and motives (.098), and depression and motives (.149) lost significance in the saturated

model, compared to the hypothesized model. Significant paths between social anxiety

and negative outcomes (.201), and depression and negative outcomes (.17) emerged in

the saturated model. Trimming insignificant paths from the saturated model led to a final

model with these two paths retained, with the path between social anxiety and negative

outcomes (.38) and depression and negative outcomes (.30) remaining significant. A

significant path also emerged between positive outcomes and perceived stress (-.29), and

motives and depression (.26). Paths between motives and positive (.87) and negative

outcomes (.81) remained high.

The fluctuation of path loadings when non-significant paths were removed

indicated the presence of multicollinearity. Latent variables involved in the

multicollinearity were sequentially removed from the model, and alternative models were

assessed, to determine the possible source. A Haywood case occurred between motives

and negative outcomes (-1.703) when positive outcomes was removed from the model,

suggesting that these remaining variables were collinear. Possible reasons for

multicollinearity between these variables was investigated further, specifically the

possibility that the SIU Motivation items are redundant with the GPIUS-2 items.

Thematically, some items from the GPIUS-2 measure preference for SIU (e.g. “I prefer

online social interaction over face-to-face interaction”) and perception of social support

online (e.g. “I have used the Internet to make myself feel better when I was down”), in

addition to measuring negative outcomes. These items may overlap with SIU Motivation

items (e.g. “The Internet is to me a substitute for other social contacts”, “The Internet


35

helps me coping with personal problems”), to the extent that endorsement of one item

necessarily leads to similar endorsement of the other item.

The hypothesized model was thus not retained, due to the multicollinearity between these

variables disrupting path loadings and decreasing interpretation of the model. As the

Negative Outcomes factor appeared to have stronger and more theoretically consistent

relations with other factors in the model, and SIU Motivation items had weak or non-

significant relations with psychosocial factors, it appears that Motivation items are

redundant to Outcomes items and should not be retained.

An alternative model was specified (See Figure 4), in which the Motivation

variable was removed from the model and each Outcomes factors was directly predicted

by each of the psychosocial factors. Model fit appeared to remain similar to the original

hypothesized model: Chi-square = 596.203 (df=353, p=.000), CFI = .938, RMSEA =

.048 (.042 - .055). While the Chi-square statistic decreased, it remained significant and

did not appear to approach non-significance. The CFI statistic increased somewhat and

the RMSEA statistic decreased slightly, though both continued to indicate that the model

remained in the “adequate” fit range. Positive Outcomes and Depression had a

significant positive relationship (.23), while Negative Outcomes had a significant positive

relationship with both Depression (.26) and with Social Anxiety (.24). Positive and

Negative Outcomes had a now-significant positive relationship (.54) in the modified

model, suggesting that Motives may have suppressed this relationship through its

redundancy in the original hypothesized model. All other paths remained consistent and

in the same direction as previously hypothesized. As the revised model appears to have


36

similar fit to the hypothesized model, while also being more parsimonious, this model

will be retained.

Discussion

Previous research has determined relations between SIU, including Internet use

outcomes, and variables such as depressive symptoms, social anxiety, neuroticism,

introversion, and stress. This is the first study to date to assess the multivariate relations

between these particular variables. This study is also the first known to evaluate relations

among the positive and negative outcomes of Internet use and various factors which may

increase or decrease motivation to engage in the behavior. The current study sought to

understand these relations better in order to inform research on Internet use as a coping

mechanism, and to provide an initial framework for more consistent future research on

the various motives and outcomes of this behavior.

Participants appeared to engage in SIU to an extent that was similar to that in

other studies. Microblog usage was most frequent, with 28% of participants endorsing

daily use of Pinterest and 52% of participants endorsing daily use of Instagram. Around

9% of participants endorsed daily use of dating websites, including popular dating apps

such as Tinder (6%). Participants also endorsed popular online activities which usually

involve some anonymous interactions, such as 55% endorsing daily Twitter use, 6%

endorsing Reddit, and 6% endorsing engagement in MMORPGs. Many used Facebook

(75%) and Facetime/Skype (30%), suggesting that a good deal of SIU involves

interactions with previously-established friends and family.

As SIU motives were not retained for the final model due to difficulties with

multicollinearity impeding interpretation, the majority of original hypotheses and


37

questions (H1, H2, H3, and H5) could not be adequately confirmed or disconfirmed for

either model. The research question posed could not be addressed adequately either. The

fourth hypothesis (H4), that psychosocial factors will relate positively to neuroticism and

negatively with extraversion (except coping, which relates positively to both), was mostly

confirmed in the respecified model. All relations occurred in the expected direction,

apart from a positive, non-significant relationship between extraversion and perceived

stress.

Multicollinearity between SIU Motives and Negative Outcomes suggests some

complications which could be addressed in future studies. SIU Motives appears to assess

very general social motivations and beliefs related to using the Internet, likely

contributing to the overlap these items had with Preference items on the GPIUS-2.

Assessment of particular motives may be beneficial, consistent with recent research on

this area (Mazzoni, Baiocco, Cannata & Dimas, 2016), and may prevent

multicollinearity. Combination of SIU Motives with coping items may effectively

address this area, and could include emotional social support motives (e.g., “My online

friends help me feel better when I’m upset”), general social support motives (e.g.,

“People I meet online are generally caring and willing to listen to me”), and distraction

motives (e.g., “It is enjoyable to play games or chat with people online to pass the time”).

SIU Motives may also be more accurately assessed through a variable measuring

actual or perceived face-to-face social support. That is, motivation to engage in SIU

could be thought of as the inverse of perceived or actual face-to-face social support, such

that individuals with very little social support are more motivated to engage in SIU. A


38

variable of this nature would be theoretically consistent with the GPIUS-2 and would

likely not have similar items, leading to less possibility of multicollinearity.

The respecified model appears to fit the data well, especially given its relative

complexity, and has been retained for the purposes of this study. Of particular interest is

a positive relationship between depressive symptoms and positive outcomes. This

finding appears to support the results found in a meta-analysis conducted by Griffiths et

al. (2009) and Campbell et al. (2006), as highly depressed individuals may be more

willing to seek out social support online and report experiencing a particular benefit to

these relationships. However, this may also support perspectives which argue that those

who prefer online social support may experience increased depressed mood from their

relatively greater interactions online. The relationship between social anxiety and

positive outcomes is also noteworthy, given the lack of significance. Contrary to the

findings of this study, previous research has determined that certain elements of online

social interaction are particularly useful for socially anxious individuals, such as the

ability to control self-presentational concerns and timing of conversations (Schouten et

al., 2007; Stritzke, Nguyen, & Durkin, 2004). The non-significant relationship between

emotional social support coping and both outcomes suggests that this form of coping may

not play a large role in participants’ experiences online. This provides some tentative

support for the cognitive-behavioral model put forth by Davis (2001) and Caplan (2002),

which states that use of the Internet for social purposes inherently leads to difficulties

once face-to-face interactions are replaced.

Positive relations between negative outcomes and depressive symptoms (.26) and

social anxiety (.24) are also consistent with the cognitive-behavioral model. Though


39

items assessing SIU motives were no longer in the model, the GPIUS-2’s scale measuring

SIU preferences appears to also measure these motivations to an extent. Thus, the

relationship between these variables appear to provide some tentative support for this

hypothesis.

Perceived stress did not relate to either outcome in the respecified model, though

it related to positive outcomes in the saturated model and has a significant, positive

relationship with depressive symptoms. Due to these relations, depressive symptoms

may account for a significant portion of variance between perceived stress and each

outcome, with social anxiety also accounting for a significant portion of variance in

relation to negative outcomes. The relation between these two variables and each

outcome may warrant further exploration.

Though correlational in nature, the results of the study appear to provide support

for each outcome, suggesting that other, unknown components of SIU might determine

the relative positive or negative impact of this behavior on the individual more

accurately. The cognitive-behavioral model posed by Caplan (2002) and Davis (2001)

appears to have somewhat greater support, suggesting that the possibility of negative

outcomes may be more likely than positive outcomes. However, this may due primarily

to the use of a relatively more established measure. Given some of the limitations of the

current model and study, interpretation of this model is tentative and primarily to indicate

areas of future research and consideration.

Limitations

Although many instruments used in this study have shown strong psychometric

properties (e.g. CES-D, BFI), the use of some instruments without established


40

psychometrics likely reduced interpretability of the model. SIU motives items appeared

to share significant variance with GPIUS-2 items, to the extent that they appeared

multicollinear. This was confirmed through assessment of the face validity of these items

and assessment of the saturated model. Though removal of items which assess motives

allowed for a more parsimonious model without reduced fit, measures such as the

GPIUS-2 are not widely established yet. Other measures, such as the modified SCS-R,

were created specifically for the current study and do not have established psychometric

qualities outside of this study, though internal consistency was adequate (α = .82).

Though these instruments appeared to have acceptable psychometric properties, the

possibility of poor construct validity or decreased power may still hinder interpretation of

these results.

In particular, the modified SCS-R was utilized as a measure of positive outcomes

of social Internet use, as it appeared to approximate the perception of good social support

online. Individuals who feel connected to others online should ideally have experienced

positive interactions beforehand. However, there are alternative ways in which this

construct could be assessed which might relate more closely to positive outcomes,

conceptually. Specifically, the modified SCS-R primarily appears to assess satisfaction

with online relationships. Thus, it may not adequately assess general positive

consequences experienced through these relationships, as a relationship between this

satisfaction and actual benefits experienced is only implicitly assumed. Additionally, this

measure does not differentiate between maladaptive and adaptive SIU, making it difficult

to draw conclusions. Use of a measure which focuses on general outcomes from

behaviors, such as positive affect or well-being scales, should be considered in future


41

studies, e.g. the Positive and Negative Affect Schedule (PANAS). In general, further

development and validation of the measures for positive and negative outcomes, and SIU

motives in particular, will improve on some of the limitations in this area.

A further limitation of this study is a non-diverse sample, which consisted entirely

of college-going students. These students were also largely White (66% of sample) and

were college freshman or sophomores (84% of sample). Thus, the results of this study

may not generalize to other populations or even other groups of college-going students

(e.g. non-traditional students), and should be carefully interpreted.

Finally, interpretation of the results was hindered by the measurement model.

Though the measurement model had high path loadings and adequate enough fit to assess

some of the hypotheses in the current study, the fit was not ideal. In particular, the chi-

square statistic remained significant and the CFI was lower than expected, indicating that

the model is not an exact fit. Replication of the model will be necessary to make more

definite conclusions about the relations among each of the variables.

Further limitations of the current study are related to the use of structural equation

modeling. Although it allows directionality to be specified, SEM is not able to establish

causality among variables. Thus, there remains the possibility that the relations among

these variables have different directionality than hypothesized. The possibility of bi-

directionality or curvilinear relation should also be considered. Additionally, despite the

assessment of a plausible alternative model assessing a different causal pathway, there

remains the possibility that there are several other equivalent and non-equivalent models

not assessed in this study which fit the data equally or better. Thus, the results found

should be interpreted with caution, and the model considered only a plausible explanation


42

(Tomarken & Waller, 2003). In particular, more accurate and interpretable assessment of

the role of increased or decreased motivation to engage in SIU, and subsequent negative

or positive outcomes, would be better determined through a longitudinal SEM approach

such as a panel study.

Finally, a common criticism of SEM analysis is that there is always a possibility

that some variables which are not assessed in this study may still impact the relations

among latent factors in the model. For this study in particular, certain demographic

characteristics, such as gender and cultural factors, and other psychosocial factors, such

as loneliness, have some support in regards to their relationship with SIU motivation as a

coping mechanism. As the model has demonstrated adequate but not perfect fit to the

data, determination of other empirically supported variables and inclusion of these

variables into the model may improve interpretability.

This study was the first to assess the relationship among SIU, positive and

negative outcomes of this behavior, and all empirically associated variables. Though

difficulties with multicollinearity required significant respecification of the model and

limited confirmation or disconfirmation of apriori hypotheses, the model provided some

insight into multivariate relations between these variables that had not been determined

previously in the literature. Additional research will be needed to provide more definite

conclusions regarding the nature of these relations, but this study contributes unique

initial findings which may benefit this research through determination of particular

directions this research might take. Continued research in this area will be necessary, as

communication in all forms is becoming increasingly more reliant on electronic

mediums. Determining the benefits and pitfalls of this communication medium on


43

interpersonal relationships will be important, especially with populations that could

utilize this technology most (e.g. individuals in rural areas, individuals with medical

concerns that prevent mobility). Though there continues to be a relative lack of clarity

regarding the directional relationship between various mental health outcomes and SIU,

this comorbidity is demonstrably present in this study and the literature reviewed within.

Clinicians should potentially consider the presence of online communication

when assessing interpersonal relationships and social activities, as these behaviors may

serve as an indicator of possible mental health difficulties, particularly social anxiety or

depressive symptoms. It appears that individuals who engage in heavy Internet use,

when there are face-to-face relationships available to them, are potentially at risk for

mental health difficulties, based on some results of this study. Frequency and severity of

use may be beneficial areas to explore further with these clients, particularly in relation to

their satisfaction with face-to-face relationships. In contrast to problematic forms of SIU,

it is possible that individuals who supplement their face-to-face relationships with online

relationships, rather than replacing them, will experience particular benefits that may be

worthwhile. In conclusion, though the potential benefits and consequences of SIU

remain to be seen, this study provides some support for the possibility that both can occur

under particular circumstances and in relation to particular factors. Further determination

of these areas will be beneficial at providing a new potential area of social and emotional

support that individuals around the world can utilize.


44

References

Amichai-Hamburger, Y., & Hayat, Z. (2011). The impact of the Internet on the social

lives of users: A representative sample from 13 countries. Computers in Human

Behavior, 27(1), 585-589. doi: 10.1016/j.chb.2010.10.009

Amichai-Hamburger, Y., Kaplan, H., & Dorpatcheon, N. (2008). Click to the past: The

impact of extroversion by users of nostalgic websites on the use of Internet social

services. Computers in Human Behavior, 24(5), 1907-1912. doi:

10.1016/j.chb.2008.02.005

Amichai-Hamburger, Y., Wainapel, G., & Fox, S. (2002). "On the Internet no one knows

I'm an introvert": Extroversion, neuroticism, and Internet interaction.

Cyberpsychology & Behavior, 5(2), 125-128. doi: 10.1089/109493102753770507

Amiel, T., & Sargent, S. L. (2004). Individual differences in Internet usage motives.

Computers in Human Behavior, 20(6), 711-726. doi: 10.1016/j.chb.2004.09.002

Attard, A., & Coulson, N. S. (2012). A thematic analysis of patient communication in

Parkinson's disease online support group discussion forums. Computers in Human


Bagby, R. M., Quilty, L. C., & Ryder, A. C. (2008). Personality and depression.

Canadian Journal of Psychiatry-Revue Canadienne De Psychiatrie, 53(1), 14-25.

Baker, J. R., & Moore, S. M. (2008). Blogging as a Social Tool: A Psychosocial

Examination of the Effects of Blogging. Cyberpsychology & Behavior, 11(6),

747-749. doi: 10.1089/cpb.2008.0053

Banjanin, N., Banjanin, N., Dimitrijevic, I., & Pantic, I. (2015). Relationship between

internet use and depression: Focus on physiological mood oscillations, social


45

networking and online addictive behavior. Computers In Human

Behavior, 43308-312. doi:10.1016/j.chb.2014.11.013

Barak, A., Boniel-Nissim, M., & Suler, J. (2008). Fostering empowerment in online

support groups. Computers in Human Behavior, 24(5), 1867-1883. doi:

10.1016/j.chb.2008.02.004

Barger, S. D. (2002). The Marlowe-Crowne Affair: Short Forms, Psychometric Structure,

and Social Desirability. Journal Of Personality Assessment, 79(2), 286-305.

Bargh, J. A., McKenna, K. Y. A., & Fitzsimons, G. M. (2002). Can you see the real me?

Activation and expression of the "true self" on the Internet. Journal of Social

Issues, 58(1), 33-48. doi: 10.1111/1540-4560.00247

Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal

attachments as a fundamental human motivation. Psychological Bulletin, 117,

497-529. Doi:10.1037//0033-2909.117.3.497

Beaudoin, C. E., & Tao, C. C. (2007). Benefiting from social capital in Online support

groups: An empirical study of cancer patients. Cyberpsychology & Behavior,

10(4), 587-590. doi: 10.1089/cpb.2007.9986

Beretvas, S. N., Meyers, J. L., & Leite, W. L. (2002). A reliability generalization study of

the Marlowe-Crowne Social Desirability Scale. Educational And Psychological

Measurement, 62(4), 570-589. doi:10.1177/0013164402062004003

Brandtzaeg, P. B. (2012). Social Networking Sites: Their Users and Social Implications -

A Longitudinal Study. Journal of Computer-Mediated Communication, 17(4),

467-488. doi: 10.1111/j.1083-6101.2012.01580.x


46

Buhrmester, D. (1996). Need fulfillment, interpersonal competence, and the

developmental contexts of early adolescent friendship. In W. M. Bukowski, A. F.

Newcomb, & W. W. Hartup (Eds.), The company they keep: Friendship in

childhood and adolescence (pp. 158–185). Cambridge University Press:

Cambridge.

Bunde, M., Suls, J., Martin, R., & Barnett, K. (2006). Hystersisters online: Social support

and social comparison among hysterectomy patients on the Internet. Annals of

Behavioral Medicine, 31(3), 271-278. doi: 10.1207/s15324796abm3103_9

Butt, S., & Phillips, J. G. (2008). Personality and self reported mobile phone use.


Byun, S., Ruffini, C., Mills, J. E., Douglas, A. C., Niang, M., Stepchenkova, S., . . .

Blanton, M. (2009). Internet Addiction: Metasynthesis of 1996-2006 quantitative

research. Cyberpsychology & Behavior, 12(2), 203-207. doi:

10.1089/cpb.2008.0102

Campbell, A. J., Cumming, S. R., & Hughes, I. (2006). Internet use by the socially

fearful: Addiction or therapy? Cyberpsychology & Behavior, 9(1), 69-81. doi:

10.1089/cpb.2006.9.69

Caplan, S. E. (2002). Problematic Internet use and psychosocial well-being: development

of a theory-based cognitive-behavioral measurement instrument. Computers in

Human Behavior, 18(5), 553-575. doi: 10.1016/s0747-5632(02)00004-3

Caplan, S. E. (2003). Preference for online social interaction - A theory of problematic

Internet use and psychosocial well-being. Communication Research, 30(6), 625-

648. doi: 10.1177/0093650203257842


47

Caplan, S. E. (2005). A social skill account of problematic internet use. Journal of

Communication, 55(4), 721-736. doi: 10.1111/j.1460-2466.2005.tb03019.x

Caplan, S. E. (2007). Relations among loneliness, social anxiety, and problematic

Internet use. Cyberpsychology & Behavior, 10(2), 234-242. doi:

10.1089/cpb.2006.9963

Caplan, S. E. (2010). Theory and measurement of generalized problematic Internet use: A

two-step approach. Computers in Human Behavior, 26(5), 1089-1097. doi:

10.1016/j.chb.2010.03.012

Carver, C. S., Scheier, M. F., & Weintraub, J. K. (1989). Assessing coping strategies – A

theoretically based approach. Journal of Personality and Social Psychology,

56(2), 267-283. doi: 10.1037//0022-3514.56.2.267

Chen, L. J., & Chung, S. K. (2007). Loneliness, social connectedness, and family income

among undergraduate females and males in Taiwan. Social Behavior and

Personality, 35(10), 1353-1364. doi: 10.2224/sbp.2007.35.10.1353

Chittaro, L., & Vianello, A. (2013). Time perspective as a predictor of problematic

Internet use: A study of Facebook users. Personality and Individual Differences,

55(8), 989-993. doi: 10.1016/j.paid.2013.08.007

Chung, J. E. (2013). Social interaction in online support groups: Preference for online

social interaction over offline social interaction. Computers in Human Behavior,

29(4), 1408-1414. doi: 10.1016/j.chb.2013.01.019

Cohen, S., & Janicki-Deverts, D. (2012). Who's Stressed? Distributions of Psychological

Stress in the United States in Probability Samples from 1983, 2006, and 2009.


48

Journal of Applied Social Psychology, 42(6), 1320-1334. doi: 10.1111/j.1559-

1816.2012.00900.x

Cohen, S., Kamarck, T., Mermelstein, R. (1983). A global measure of perceived

stress. Journal of Health and Social Behavior, 24, 385-396.

Cohen, S., & Williamson, G. (1988). Perceived stress in a probability sample of the

United States. In S. Spacapam & S. Oskamp (Eds.), The social psychology of

health: Claremont Symposium on applied social psychology (pp. 31-

67). Newbury Park: Sage

Cook, S. W., & Heppner, P. P. (1997). A psychometric study of three coping measures.

Educational and Psychological Measurement, 57(6), 906-923. doi:

10.1177/0013164497057006002

Davis, R. A. (2001). A cognitive-behavioral model of pathological Internet use.

Computers in Human Behavior, 17(2), 187-195. doi: 10.1016/s0747-

5632(00)00041-8

de Ayala López, M. L., Gutierrez, J. S., & Jiménez, A. G. (2015). Problematic Internet

use among Spanish adolescents: The predictive role of Internet preference and

family relationships. European Journal Of Communication, 30(4), 470-485.

doi:10.1177/0267323115586725

Deatherage, S., Servaty-Seib, H. L., & Aksoz, I. (2014). Stress, Coping, and Internet Use

of College Students. Journal of American College Health, 62(1), 40-46. doi:

10.1080/07448481.2013.843536


49

Ebeling-Witte, S., Frank, M. L., & Lester, D. (2007). Shyness, Internet use, and

personality. Cyberpsychology & Behavior, 10(5), 713-716. doi:

10.1089/cpb.2007.9964

Eisengart, S. P., Singer, L. T., Kirchner, H. L., Min, M. O., Fulton, S., Short, E. J., &

Minnes, S. (2006). Factor structure of coping: Two studies of mothers with high

levels of life stress. Psychological Assessment, 18(3), 278-288. doi:

10.1037/1040-3590.18.3.278

Facebook data team. (2010). Retrieved from http://www.facebook.com/data

Folkman, S., & Moskowitz, J. T. (2004). Coping: Pitfalls and promise. Annual Review of

Psychology, 55, 745-774. doi: 10.1146/annurev.psych.55.090902.141456

Fresco, D. M., Coles, M. E., Heimberg, R. G., Liebowitz, M. R., Hami, S., Stein, M. B.,

& Goetz, D. (2001). The Liebowitz Social Anxiety Scale: a comparison of the

psychometric properties of self-report and clinician-administered formats.

Psychological Medicine, 31(6), 1025-1035.

Galegher, J., Sproull, L., & Kiesler, S. (1998). Legitimacy, authority, and community in

electronic support groups. Written Communication, 15(4), 493-530. doi:

10.1177/0741088398015004003

Gamez-Guadix, M., Orue, I., Smith, P. K., & Calvete, E. (2013). Longitudinal and

Reciprocal Relations of Cyberbullying With Depression, Substance Use, and

Problematic Internet Use Among Adolescents. Journal of Adolescent Health,

53(4), 446-452. doi: 10.1016/j.jadohealth.2013.03.030

Gamez-Guadix, M., Villa-George, F. I., & Calvete, E. (2012). Measurement and analysis

of the cognitive-behavioral model of generalized problematic Internet use among


50

Mexican adolescents. Journal of Adolescence, 35(6), 1581-1591. doi:

10.1016/j.adolescence.2012.06.005

Gilat, I., & Shahar, G. (2009). Suicide Prevention by Online Support Groups: An Action

Theory-Based Model of Emotional First Aid. Archives of Suicide Research, 13(1),

52-63. doi: 10.1080/13811110802572148

Gordon, C. F., Juang, L. P., & Syed, M. (2007). Internet use and well-being among

college students: Beyond frequency of use. Journal of College Student

Development, 48(6), 674-688. doi: 10.1353/csd.2007.0065

Grieve, R., Indian, M., Witteveen, K., Tolan, G. A., & Marrington, J. (2013). Face-to-

face or Facebook: Can social connectedness be derived online? Computers in

Human Behavior, 29(3), 604-609. doi: 10.1016/j.chb.2012.11.017

Griffiths, K. M., Calear, A. L., & Banfield, M. (2009). Systematic Review on Internet

Support Groups (ISGs) and Depression: Do ISGs Reduce Depressive Symptoms?

Journal of Medical Internet Research, 11(3). doi: 10.2196/jmir.1270

Guadagno, R. E., Okdie, B. M., & Eno, C. A. (2008). Who blogs? Personality predictors

of blogging. Computers in Human Behavior, 24(5), 1993-2004. doi:

10.1016/j.chb.2007.09.001

Gunthert, K. C., Cohen, L. H., & Armeli, S. (1999). The role of neuroticism in daily

stress and coping. Journal of Personality and Social Psychology, 77(5), 1087-

1100. doi: 10.1037/0022-3514.77.5.1087

Gutierrez, J. L. G., Jimenez, B. M., Hernandez, E. G., & Puente, C. P. (2005). Personality

and subjective well-being: big five correlates and demographic variables.


51

Personality and Individual Differences, 38(7), 1561-1569. doi:

10.1016/j.paid.2004.09.015

Han, J. Y., Shaw, B. R., Hawkins, R. P., Pingree, S., McTavish, F., & Gustafson, D. H.

(2008). Expressing Positive Emotions within Online Support Groups by Women

with Breast Cancer. Journal of Health Psychology, 13(8), 1002-1007. doi:

10.1177/1359105308097963

Heimberg, R. G., Horner, K. J., Juster, H. R., Safren, S. A., Brown, E. J., Schneier, F. R.,

& Liebowitz, M. R. (1999). Psychometric properties of the Liebowitz Social

Anxiety Scale. Psychological Medicine, 29(1), 199-212. doi:

10.1017/s0033291798007879

High, A. C., & Caplan, S. E. (2009). Social anxiety and computer-mediated

communication during initial interactions: Implications for the hyperpersonal

perspective. Computers in Human Behavior, 25(2), 475-482. doi:

10.1016/j.chb.2008.10.011

Houston, T. K., Cooper, L. A., & Ford, D. E. (2002). Internet support groups for

depression: A 1-year prospective cohort study. American Journal of Psychiatry,

159(12), 2062-2068. doi: 10.1176/appi.ajp.159.12.2062

Hoybye, M. T., Johansen, C., & Tjornhoj-Thomsen, T. (2005). Online interaction. effects

of storytelling in an Internet breast cancer support group. Psycho-Oncology,

14(3), 211-220. doi: 10.1002/pon.837

Huang, C. (2010). Internet Use and Psychological Well-being: A Meta-Analysis.

Cyberpsychology Behavior and Social Networking, 13(3), 241-249. doi:

10.1089/cyber.2009.0217


52

Hughes, D. J., Rowe, M., Batey, M., & Lee, A. (2012). A tale of two sites: Twitter vs.

Facebook and the personality predictors of social media usage. Computers in

Human Behavior, 28(2), 561-569. doi: 10.1016/j.chb.2011.11.001

Jenkins-Guarnieri, M. A., Wright, S. L., & Hudiburgh, L. M. (2012). The relationships

among attachment style, personality traits, interpersonal competency, and

Facebook use. Journal of Applied Developmental Psychology, 33(6), 294-301.

doi: 10.1016/j.appdev.2012.08.001

John, O. P., Donahue, E. M., & Kentle, R. L. (1991). The Big Five Inventory – Versions

4a and 54. Berkeley, CA: University of California, Berkeley, Institute of

Personality and Social Research.

John, O. P., Naumann, L. P., & Soto, C. J. (2008). Paradigm shift to the integrative Big

Five trait taxonomy: History, measurement, and conceptual issues. In O. P. John,

R. W. Robins & L. A. Pervin (Eds.), Handbook of personality: Theory and

research (3rd ed.) (pp. 114-158). New York: Guilford Press.

John, O.P., & Srivastava, S. (1999). The Big-Five Trait Taxonomy: History,

Measurement, and Theoretical Perspectives. In L. A. Pervin & O. P. John

(Eds.), Handbook of personality: Theory and research (2nd ed.) (pp. 102-139),

New York: Guilford Press.

Joinson, A. N. (2001). Self-disclosure in computer-mediated communication: The role of

self-awareness and visual anonymity. European Journal of Social Psychology,

31(2), 177-192. doi: 10.1002/ejsp.36

Jones, S., & Fox, S. (2009). Generations online in 2009. Pew Internet and American Life


53

Project. Retrieved from http://www.pewInternet.org/Reports/2009/Generations-Online-

in-

2009.aspx

Kalmus, V., Realo, A., & Siibak, A. (2011). Motivves for Internet use and their

relationships with personality traits and socio-demographic factors. Trames-

Journal of the Humanities and Social Sciences, 15(4), 385-403. doi:

10.3176/tr.2011.4.04

Kaltiala-Heino, R., Lintonen, T., & Rimpela, A. (2004). Internet addiction? Potentially

problematic use of the Internet in a population of 12-18 year-old adolescents.

Addiction Research & Theory, 12(1), 89-96. doi: 10.1080/1606635031000098796

Kessler, R. C., Stang, P., Wittchen, H. U., Stein, M., & Walters, E. E. (1999). Lifetime

co-morbidities between social phobia and mood disorders in the US National

Comorbidity Survey. Psychological Medicine, 29(3), 555-567. doi:

10.1017/s0033291799008375

Kim, J., LaRose, R., & Peng, W. (2009). Loneliness as the Cause and the Effect of

Problematic Internet Use: The Relationship between Internet Use and

Psychological Well-Being. Cyberpsychology & Behavior, 12(4), 451-455. doi:

10.1089/cpb.2008.0327

Kimbrough, A. M., Guadagno, R. E., Muscanell, N. L., & Dill, J. (2013). Gender

differences in mediated communication: Women connect more than do men.


Kline, R. B. (2011). Principles and practice of structural equation modeling (3rd ed.).

New York, NY, US: Guilford Press.


54

Kotov, R., Gamez, W., Schmidt, F., & Watson, D. (2010). Linking "Big" Personality

Traits to Anxiety, Depressive, and Substance Use Disorders: A Meta-Analysis.

Psychological Bulletin, 136(5), 768-821. doi: 10.1037/a0020327

Kraut, R., Kiesler, S., Boneva, B., Cummings, J., Helgeson, V., & Crawford, A. (2002).

Internet paradox revisited. Journal of Social Issues, 58(1), 49-74. doi:

10.1111/1540-4560.00248

Kraut, R., Patterson, M., Lundmark, V., Kiesler, S., Mukopadhyay, T., & Scherlis, W.

(1998). Internet paradox - A social technology that reduces social involvement

and psychological well-being? American Psychologist, 53(9), 1017-1031. doi:

10.1037//0003-066x.53.9.1017

LaRose, R., Connolly, R., Lee, H., Li, K., & Hales, K. D. (2014). Connection Overload?

A Cross Cultural Study of the Consequences of Social Media Connection.

Information Systems Management, 31(1), 59-73. doi:

10.1080/10580530.2014.854097

LaRose, R., Lin, C. A., & Eastin, M. S. (2003). Unregulated Internet usage: Addiction,

habit, or deficient self-regulation? Media Psychology, 5(3), 225-253. doi:

10.1207/s1532785xmep0503_01

Lee, E. H. (2012). Review of the Psychometric Evidence of the Perceived Stress Scale.

Asian Nursing Research, 6(4), 121-127. doi: 10.1016/j.anr.2012.08.004

Lee, R. M., Draper, M., & Lee, S. (2001). Social connectedness, dysfunctional

interpersonal behaviors, and psychological distress: Testing a mediator model.

Journal of Counseling Psychology, 48(3), 310-318.


55

Leite, W., & Beretvas, S. (2005). Validation of scores on the Marlowe-Crowne Social

Desirability Scale and the Balanced Inventory of Desirable Responding.

Educational & Psychological Measurement, 65(1), 140-154.

Leung, L. (2007). Stressful life events, motives for Internet use, and social support among

digital kids. Cyberpsychology & Behavior, 10(2), 204-214. doi:

10.1089/cpb.2006.9967

Liebowitz, M. R. (1987). Social phobia. Modern Problems of Pharmacopsychiatry, 22,

141-173.

Litman, J. A. (2006). The COPE inventory: Dimensionality and relationships with

approach- and avoidance-motives and positive and negative traits. Personality

and Individual Differences, 41(2), 273-284. doi: 10.1016/j.paid.2005.11.032

Little, T. D., Cunningham, W. A., Shahar, G., & Widaman, K. F. (2002). To parcel or not

to parcel: Exploring the question, weighing the merits. Structural Equation

Modeling, 9(2), 151-173. doi: 10.1207/s15328007sem0902_1

Lounsbury, J. W., Saudargas, R. A., Gibson, L. W., & Leong, F. T. (2005). An

investigation of broad and narrow personality traits in relation to general and

domain-specific life satisfaction of college students. Research in Higher

Education, 46(6), 707-729. doi: 10.1007/11162-004-4140-6

Lyne, K., & Roger, D. (2000). A psychometric re-assessment of the COPE questionnaire.

Personality and Individual Differences, 29(2), 321-335. doi: 10.1016/s0191-

8869(99)00196-8


56

MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and

determination of sample size for covariance structure modeling. Psychological

Methods, 1(2), 130-149. doi: 10.1037//1082-989x.1.2.130

Malouff, J. M., Thorsteinsson, E. B., & Schutte, N. S. (2005). The relationship between

the five-factor model of personality and symptoms of clinical disorders: A meta-

analysis. Journal of Psychopathology and Behavioral Assessment, 27(2), 101-114.

doi: 10.1007/s10862-005-5384-y

Matsuba, M. K. (2006). Searching for self and relationships online. Cyberpsychology &

Behavior, 9(3), 275-284. doi: 10.1089/cpb.2006.9.275

Mazzoni, E., Baiocco, L., Cannata, D., & Dimas, I. (2016). Is internet the cherry on top

or a crutch? Offline social support as moderator of the outcomes of online social

support on Problematic Internet Use. Computers In Human Behavior, 56, 369-

374. doi:10.1016/j.chb.2015.11.032

McCrae, R. R., & Costa, P. T. (1987). Validation of the 5-factor model of personality

across instruments and observers. Journal of Personality and Social Psychology,

52(1), 81-90. doi: 10.1037/0022-3514.52.1.81

McKenna, K. Y. A., & Bargh, J. A. (2000). Plan 9 from cyberspace: The implications of

the internet for personality and social psychology. Personality and Social

Psychology Review, 4(1), 57-75. doi: 10.1207/s15327957pspr0401_6

McKenna, K. Y. A., Green, A. S., & Gleason, M. E. J. (2002). Relationship formation on

the Internet: What’s the big attraction? Journal of Social Issues, 58(1), 9-31.

Doi: 10.1111/1540-4560.00246


57

Milani, L., Osualdella, D., & Di Blasio, P. (2009). Quality of Interpersonal Relationships

and Problematic Internet Use in Adolescence. Cyberpsychology & Behavior,

12(6), 681-684. doi: 10.1089/cpb.2009.0071

Mineka, S., Watson, D., & Clark, L. A. (1998). Comorbidity of anxiety and unipolar

mood disorders. Annual Review of Psychology, 49, 377-412. doi:

10.1146/annurev.psych.49.1.377

Mo, P. K. H., & Coulson, N. S. (2010). Living with HIV/AIDS and Use of Online

Support Groups. Journal of Health Psychology, 15(3), 339-350. doi:

10.1177/1359105309348808

Morahan-Martin, J., & Schumacher, P. (2000). Incidence and correlates of pathological

Internet use among college students. Computers in Human Behavior, 16(1), 13-

29. doi: 10.1016/s0747-5632(99)00049-7

Morahan-Martin, J., & Schumacher, P. (2003). Loneliness and social uses of the Internet.

Computers in Human Behavior, 19(6), 659-671. doi: 10.1016/s0747-

5632(03)00040-2

Morgan, C., & Cotten, S. R. (2003). The relationship between Internet activities and

depressive symptoms in a sample of college freshmen. Cyberpsychology &

Behavior, 6(2), 133-142. doi: 10.1089/109493103321640329

Muscanell, N. L., & Guadagno, R. E. (2012). Make new friends or keep the old: Gender

and personality differences in social networking use. Computers in Human


Muthén, L.K., & Muthén, B.O. (1998-2010). Mplus user’s guide (6th ed.). Los Angeles:

Muthén & Muthén.


58

Nadkarni, A., & Hofmann, S. G. (2012). Why do people use Facebook? Personality and

Individual Differences, 52(3), 243-249. doi: 10.1016/j.paid.2011.11.007

Nezu, A. M., Nezu, C. M., Friedman, J., & Lee, M. (2009). Assessment of depression. In

I. H. Gotlib & C. L. Hammen (Eds.), Handbook of depression (2nd ed.) (pp. 44-

68). New York, NY, US: Guilford Press.

Ogles, B. M., France, C. R., Lunnen, K. M., Bell, M. T., & Goldfarb, M. (1998).

Computerized depression screening and awareness. Community Mental Health

Journal, 34(1), 27-38. doi: 10.1023/a:1018760128239

Orchard, L. J., & Fullwood, C. (2010). Current Perspectives on Personality and Internet

Use. Social Science Computer Review, 28(2), 155-169. doi:

10.1177/0894439309335115

Orme, J. G., Reis, J., & Herz, E. J. (1986). Factorial and discriminate validity of the

Center for Epidemiological Studies Depression (CES-D) Scale. Journal of

Clinical Psychology, 42(1), 28-33. Doi:10.1002/1097-

4679(198601)42:1<28::AID-JCLP2270420104>3.0.CO;2-T

Palmer-Olsen, L., Gold, L. L., & Woolley, S. R. (2011). Supervising emotionally focused

therapists: A systematic research-based model. Journal of Marital and Family

Therapy, 37(4), 411-426. doi: 10.1111/j.1752-0606.2011.00253.x

Peter, J., Valkenburg, P. M., & Schouten, A. P. (2005). Developing a model of adolescent

friendship formation on the Internet. Cyberpsychology & Behavior, 8(5), 423-430.

doi: 10.1089/cpb.2005.8.423

Pew Research Center (2006). Online dating. Retrieved from

http://www.pewInternet.org/Reports/2006/Online-Dating.aspx


59

Quinones, C., & Kakabadse, N. K. (2015). Self-concept clarity, social support, and

compulsive Internet use: A study of the US and the UAE. Computers In Human

Behavior, 44, 347-356. doi:10.1016/j.chb.2014.11.019

Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the

general population. Applied Psychological Measurement, 1, 385-401.

Doi:10.1177/014662167700100306

Rogers, C. R. (1961). On becoming a person. Oxford England: Houghton Mifflin.

Santor, D. A., Zuroff, D. C., Cervantes, P., & Palacios, J. (1995). Examining scale

discriminability in the BDI and CES-D as a function of depressive severity.

Psychological Assessment, 7(2), 131-139. doi: 10.1037/1040-3590.7.2.131

Satorra, A., & Bentler, P. M. (1994). Corrections to test statistics and standard errors in

covariance structure analysis. In A. von Eye & C. C. Clogg (Eds.), Latent

variables analysis: Applications for developmental research. (pp. 399-419).

Thousand Oaks, CA, US: Sage Publications, Inc.

Schouten, A. P., Valkenburg, P. M., & Peter, J. (2007). Precursors and underlying

processes of adolescents' online self-disclosure: Developing and testing an

"Internet-attribute-perception" model. Media Psychology, 10(2), 292-314. doi:

10.1080/15213260701375686

Seckin, G. (2013). Satisfaction with health status among cyber patients: testing a

mediation model of electronic coping support. Behaviour & Information

Technology, 32(1), 91-101. doi: 10.1080/0144929x.2011.603359

Seepersad, S. (2004). Coping with loneliness: Adolescent online and offline behavior.

Cyberpsychology & Behavior, 7(1), 35-39. doi: 10.1089/109493104322820093


60

Seidman, G. (2013). Self-presentation and belonging on Facebook: How personality

influences social media use and motivations. Personality and Individual

Differences, 54(3), 402-407. doi: 10.1016/j.paid.2012.10.009

Shaw, L. H., & Gant, L. M. (2002). In Defense of the internet: The relationship between

Internet communication and depression, loneliness, self-esteem, and perceived

social support. Cyberpsychology & Behavior, 5(2), 157-171. doi:

10.1089/109493102753770552

Shean, G., & Baldwin, G. (2008). Sensitivity and specificity of depression questionnaires

in a college-age sample. Journal of Genetic Psychology, 169(3), 281-288. doi:

10.3200/gntp.169.3.281-292

Shim, M., Cappella, J. N., & Han, J. Y. (2011). How Does Insightful and Emotional

Disclosure Bring Potential Health Benefits? Study Based on Online Support

Groups for Women With Breast Cancer. Journal of Communication, 61(3). doi:

10.1111/j.1460-2466.2011.01555.x

Shotton, M. A. (1991). The costs and benefits of computer addiction. Behaviour &

Information Technology, 10(3), 219-230.

Skues, J. L., Williams, B., & Wise, L. (2012). The effects of personality traits, self-

esteem, loneliness, and narcissism on Facebook use among university students.

Computers in Human Behavior, 28(6), 2414-2419. doi:

10.1016/j.chb.2012.07.012

Smedema, S. M., & McKenzie, A. R. (2010). The relationship among frequency and type

of internet use, perceived social support, and sense of well-being in individuals


61

with visual impairments. Disability and Rehabilitation, 32(4), 317-325. doi:

10.3109/09638280903095908

Soto, C. J., & John, O. P. (2009). Ten facet scales for the Big Five Inventory:

Convergence with NEO PI-R facets, self-peer agreement, and discriminant

validity. Journal of Research in Personality, 43(1), 84-90. doi:

10.1016/j.jrp.2008.10.002

Sproull, L., & Kiesler, S. B. (1991). Connections: New ways of working in the networked

organization. Cambridge, MA US: The MIT Press.

Stansbury, J. P., Ried, L. D., & Velozo, C. A. (2006). Unidimensionality and bandwidth

in the center for epidemiologic studies depression (CES-D) scale. Journal of

Personality Assessment, 86(1), 10-22. doi: 10.1207/s15327752jpa8601_03

Stepanikova, I., Nie, N. H., & He, X. B. (2010). Time on the Internet at home, loneliness,

and life satisfaction: Evidence from panel time-diary data. Computers in Human


Steunenberg, B., Braam, A. W., Beekman, A. T. F., Deeg, D. J. H., & Kerkhof, A.

(2009). Evidence for an association of the big five personality factors with

recurrence of depressive symptoms in later life. International Journal of Geriatric

Psychiatry, 24(12), 1470-1477. doi: 10.1002/gps.2291

Stritzke, W. G. K., Nguyen, A., & Durkin, K. (2004). Shyness and computer-mediated

communication: A self-presentational theory perspective. Media Psychology,

6(1), 1-22. doi: 10.1207/s1532785xmep0601_1

Suler, J. (2004). The online disinhibition effect. Cyberpsychology & Behavior, 7(3), 321-

326. doi: 10.1089/1094931041291295


62

Tanis, M. (2007). Online social support groups: The appeal of support via the Internet.

Psychologie & Gezondheid, 35(5), 234-242.

Tian, Q. (2013). Social Anxiety, Motivation, Self-Disclosure, and Computer-Mediated

Friendship: A Path Analysis of the Social Interaction in the Blogosphere.

Communication Research, 40(2), 237-260. doi: 10.1177/0093650211420137

Tokunaga, R. S., & Rains, S. A. (2010). An Evaluation of Two Characterizations of the

Relationships Between Problematic Internet Use, Time Spent Using the Internet,

and Psychosocial Problems. Human Communication Research, 36(4), 512-545.

doi: 10.1111/j.1468-2958.2010.01386.x

Tokunaga, R. S., & Rains, S. A. (2016). A review and meta‐analysis examining

conceptual and operational definitions of problematic internet use. Human

Communication Research, doi:10.1111/hcre.12075

Tomarken, A. J., & Waller, N. G. (2003). Potential problems with "well fitting" models.

Journal of Abnormal Psychology, 112(4), 578-598. doi: 10.1037/0021-

843x.112.4.578

Tomarken, A. J., & Waller, N. G. (2005). Structural equation modeling: Strengths,

limitations, and misconceptions. Annual Review of Clinical Psychology, 1, 31-65.

doi: 10.1146/annurev.clinpsy.1.102803.144239

Tosun, L. P., & Lajunen, T. (2010). Does Internet use reflect your personality?

Relationship between Eysenck's personality dimensions and Internet use.



63

Valkenburg, P. M., & Peter, J. (2008). Adolescents' identity experiments on the Internet -

Consequences for social competence and self-concept unity. Communication

Research, 35(2), 208-231. doi: 10.1177/0093650207313164

Vance, A. (2012). Facebook: The making of 1 billion users. Bloomberg Businessweek.

Retrieved from http://www.businessweek.com/articles/2012-10-04/facebook-the-

making-of-1-billion-users

Wang, J. & Wang, X. (2012). Structural Equation Modeling: Applications Using Mplus.

New York: John Wiley.

Watson, D., & Hubbard, B. (1996). Adaptational style and dispositional structure: Coping

in the context of the five-factor model. Journal of Personality, 64(4), 737-774.

doi: 10.1111/j.1467-6494.1996.tb00943.x

Weidman, A. C., Fernandez, K. C., Levinson, C. A., Augustine, A. A., Larsen, R. J., &

Rodebaugh, T. L. (2012). Compensatory internet use among individuals higher in

social anxiety and its implications for well-being. Personality and Individual

Differences, 53(3), 191-195. doi: 10.1016/j.paid.2012.03.003

Weinstein, A., & Lejoyeux, M. (2010). Internet Addiction or Excessive Internet Use.

American Journal of Drug and Alcohol Abuse, 36(5), 277-283. doi:

10.3109/00952990.2010.491880

Weiser, E. B. (2001). The functions of Internet use and their social and psychological

consequences. Cyberpsychology & Behavior, 4(6), 723-743. doi:

10.1089/109493101753376678


64

Wellman, B., Quan-Haase, A., Witte, J., & Hampton, K. (2001). Does the Internet

increase, decrease, or supplement social capital? Social networks, participation,

and community commitment. American Behavioral Scientist, 45(3), 436-455. doi:

10.1177/00027640121957286

Wolfradt, U., & Doll, J. (2001). Motives of adolescents to use the Internet as a function

of personality traits, personal and social factors. Journal of Educational

Computing Research, 24(1), 13-27.

Wolak, J., Mitchell, K. J. , & Finkelhor, D. (2003). Escaping or connecting?

Characteristics of youth who form close online relationships. Journal of

Adolescence, 26,105-119. doi:10.1016/S0140-1971(02)00114-8

Yao, M. Z., & Zhong, Z. J. (2014). Loneliness, social contacts and Internet addiction: A

cross-lagged panel study. Computers in Human Behavior, 30, 164-170. doi:

10.1016/j.chb.2013.08.007

Young, K. S. (1998). Internet addiction: The emergence of a new clinical disorder.

Cyberpsychology and Behavior, 1(3), 237-244.

Young, K. S., & Rogers, R. C. (1998). The relationship between depression and Internet

addiction. Cyberpsychology & Behavior, 1(1), 25-28. doi: 10.1089/cpb.1998.1.25

Zywica, J., & Danowski, J. (2008). The Faces of Facebookers: Investigating Social

Enhancement and Social Compensation Hypotheses; Predicting Facebook and

Offline Popularity from Sociability and Self-Esteem, and Mapping the Meanings

of Popularity with Semantic Networks. Journal of Computer-Mediated

Communication, 14(1), 1-+. doi: 10.1111/j.1083-6101.2008.01429.x


65

Appendix A

Overview

The current study observes the relationships between SIU motivation, depressive

symptoms, social anxiety, perceived stress, emotional social support seeking,

neuroticism, and introversion. The hypothesized relationship between these variables

will be assessed through the use of structural equation modeling. The two major

perspectives on the benefits of SIU will be examined through the assessment of alternate

models, and the inclusion of negative and positive outcomes of internet use in the model.

The purpose of this review is to define the constructs in this study, and review previous

literature which has elucidated the relationships between these constructs. This review

will provide a brief examination of the increasing sociality of the internet, and define key

terms. The two major perspectives on the potential benefits and consequences of SIU

will be critically examined. Finally, the constructs most frequently associated with SIU

motivation will be reviewed. The significance of this research area, specifically in the

clinical realm, will be highlighted throughout this review.

Definitions

Problematic Internet Use (PIU): Problematic Internet Use is defined by overuse

of the Internet, such that the user experiences negative occupational, interpersonal, and

academic consequences. Early research on this construct did not have unanimously

agreed upon terminology, and many early studies used disparate terms such as Internet

addiction (Young, 1996) and Pathological Internet Use (Davis, 2001) to describe the

same general phenomenon defined above. Other terms which have been used include


66

cyberspace addiction, Internet addiction disorder, online addiction, and high Internet

dependency (Byun et al., 2009). These terms will be used in this review as they appear in

the relevant studies, in order to maintain continuity with the literature. Problematic

Internet Use appears to be the most preferred and most frequently used term in the

literature currently (Tokunaga & Rains, 2010), as it does not pathologize the behavior to

an excessive degree (LaRose et al., 2003), and as such, will be used throughout the

remainder of the review to describe this construct.

SIU: SIU is generally defined as use of the Internet to communicate with other

people. This construct encompasses use of Instant Messaging clients, discussion on

public or private Internet forums, social networking sites, online multiplayer games,

microblogs, or any other online activity which includes direct communication with other

persons. These online behaviors may be done to maintain existing relationships or to

initiate novel relationships. Other terms for this construct include computer-mediated

communication (CMC). SIU was chosen for this study due to its greater focus on the

social motivations for this behavior, rather than a focus on the behavior itself.

Sociality of the Internet

Human beings are a highly social species, as they have a natural desire to connect

to others (Baumeister & Leary, 1995) and seek close relationships for comfort and

support (Buhrmester, 1996). It is thus not surprising that the advent and rapid expansion

of the Internet throughout the population in the 1990s coincided with a great deal of

debate among scholars and researchers as to the social implications of this burgeoning

technology (Shotton, 1991; Sproull & Kiesler, 1991; Young, 1998). At the turn of the


67

century, over half of all Americans had Internet access (Wellman, Quan-Haase, Witte, &

Hampton, 2001). Although early Internet technology was considerably less sophisticated

than current platforms, early Internet users frequently engaged in SIU; Instant messaging

clients, such as IRC and ICQ, and online communities, such as self-help groups

(Galegher, Sproull, & Kiesler, 1998), were found to be popular. Many early Internet

users reported that communication with others on the Internet could be a viable

alternative to traditional face-to-face communication (McKenna, Green, & Gleason,

2002).

SIU has increased dramatically with the continued advancement of the Internet.

Early research found that up to 14% of U.S. adolescents reported maintaining an online

friendship (Wolak, Mitchell, & Finkelhor, 2003). More recent surveys have found that

65% of adolescents participate in social networking sites, 49% read the blogs of others,

and 68% use Instant Messaging software (Jones & Fox, 2009). One of the largest social

networking sites, Facebook, claimed half of a billion users in 2010 (Facebook Data

Team, 2010), claimed around one billion users total in 2012 (Vance, 2012), and as of

early 2013, claimed 1.3 billion active monthly users

(http://www.statisticbrain.com/facebook-statistics). Many popular online dating sites

claim hundreds of thousands of members (PEW Research Center, 2006). Internet forums

and message boards, where users express themselves and discuss topics and hobbies of

interests, have also increased in membership, with some of the largest message boards

numbering hundreds of thousands of users and billions of posts

(www.thebiggestboards.com). As the Internet has become more social in nature, and


68

Internet users have become more social in their online interactions, researchers have

increasingly questioned the potential benefits and consequences of SIU.

Is SIU Helpful?

Young (1996) conducted one of the first empirical studies to assess the possible

consequences of Internet use. Based on reports of Internet overuse and negative

consequences associated with it, Young hypothesized that Internet use could become an

addictive behavior, and sought to determine possible criteria which could define Internet

addiction. Young determined that criteria for pathological gambling most adequately fit

the nature of Internet addiction, and defined it as “an impulse-control disorder that does

not involve an intoxicant”. Young developed a questionnaire which modified items for

pathological gambling into questions about pathological internet use. She gave this

measure to 496 individuals who had answered flyers or advertisements for the study, or

who had typed in “Internet addiction” into search engines, or who were members of

Internet addiction support groups.

Young reported considerable differences between those who were determined to

be Dependent, i.e. those who responded Yes to five or more criteria, and Non-Dependent,

i.e. those who responded Yes to less than five criteria. Young found that a majority of

Dependents (83%) had been on the Internet for less than a year at the time of the study,

and suggested that Internet addiction could happen relatively quickly. Controlling for

occupational uses and other necessary uses of the Internet, Dependents were found to

report using the internet 33.6 hours more, on average, than Non-Dependents. Dependents


69

spent the majority of their time engaging in SIU: They used chat rooms significantly

more often than Non-Dependents (35% and 7%, respectively), as well as MUDs (Multi-

User Dungeons), an online multi-user fantasy roleplaying game which was popular

during the early period of the Internet (28% and 5%, respectively).

Dependents reported that they experiencing a wide variety of impairment in their

functioning due to this higher degree of internet use: Academic impairment (40%

indicated Moderate, 58% indicated Severe), Relationship impairment (45% Moderate,

53% Severe), Financial impairment (38% Moderate, 52% Severe), and Occupational

impairment (34% Moderate, 51% Severe) were all reported by a majority of the

Dependents. Young argued that the results of this study provided promising initial

support for the inclusion of Internet addiction as a diagnostic category and a distinct

clinical disorder.

Although Young found significant support for internet use as an addictive

behavior, the results found have several notable flaws. Despite being acknowledged in

the original study, Young’s use of a largely self-selected population is problematic, as

many of these individuals apparently suspected or believed that they were addicted to the

Internet already. Confirmation bias in this sample could have possibly inflated some of

the differences found between Dependents and Non-Dependents. This may have also

inflated the Dependent’s self-reported impairments that they believed were due to their

internet use. Additionally, the relative inexperience with the Internet found in the sample

may have inflated the results, as has been found in one longitudinal study conducted later

(Kraut et al., 2002). Finally, LaRose et al. (2003) have pointed out that Young may have

a potential conflict-of-interest in establishing Internet Addiction as a psychological


70

disorder, as she currently maintains a website which purports to offer therapy for Internet

Addiction (www.netaddiction.com).

In an early, seminal study on SIU, Kraut et al. (1998) argued that use of the

Internet to form interpersonal relationships would have a negative impact on Internet

users, as these bonds would be largely superficial in nature. They argued that online

communication lacks the physical closeness necessary for sufficient social interaction, as

the other person typically cannot be seen while interacting. Related to this, emotional

intimacy would also be difficult to achieve, as the expression and perception of emotions

would be difficult through a chat medium. Due to their lack of emotional and physical

closeness, online relationships would not be able to provide a social support buffer

against life stressors, and the individuals who engaged in these relationships over

traditional face-to-face relationships would experience negative consequences during

stressful times as a result. Kraut and colleagues performed one of the few longitudinal

studies to-date to provide support for their hypothesis, and assessed the psychological

well-being of 231 new Internet users over a period of 12-24 months. It was found that,

controlling for initial differences in well-being, engaging in SIU led to greater depression

and feelings of loneliness in new Internet users. Kraut and colleagues coined the term

Internet Paradox to describe this effect, and concluded that the Internet, despite being a

largely social technology, has ironic and counterproductive effects on users’ social

involvement and subsequent psychological well-being.

Consistent with the argument put forth by Kraut et al., Davis (2001) proposed a

cognitive-behavioral model of Pathological Internet Use, or internet use that leads to

negative consequences in user’s careers or academic functioning and negative impacts in


71

psychosocial well-being. Davis argued that Pathological Internet Use may take on a

generalized form, i.e. a pathological attitude towards Internet use in general, not a

reliance on any particular activity or website. Davis argued that individuals who use the

Internet broadly and excessively are doing so because of preexisting mental health

problems and social stress, which lead to a preference for SIU as a coping mechanism.

Davis argued that maladaptive cognitions about Internet use (e.g.,“I am worthless offline,

but online I am someone’’), which can often co-occur with disorders such as depression

and social anxiety, cause susceptible individuals to use the Internet excessively and

subsequently experience negative life outcomes.

Caplan (2002) built off of Davis’s theory by creating and operationalizing a

measure for generalized Problematic Internet Use (PIU). Caplan found a seven-factor

model of generalized PIU: 1.) Mood alteration, or use of the Internet to alter negative

mood states, 2.) Social Benefits, or the level of perceived social benefits of Internet use,

3.) Negative Outcomes, 4.) Compulsivity, 5.) Excessive Time, 6.) Withdrawal, and 7.)

Interpersonal Control, or the degree of perception that there is increased social control

when interacting with others on the Internet. Caplan determined that these findings were

consistent with the cognitive-behavioral model developed by Davis (2001), as the factors

found related to problematic cognitions (e.g. Social Benefits), subsequent problematic

behaviors (e.g. Excessive Time), or negative outcomes experienced afterward. In

particular, Caplan found that perceived benefits had the strongest correlation with

measures of psychosocial health, such as depression, loneliness, and self-esteem. Based

on these results, Caplan argued that those who were motivated to use the Internet in a


72

social manner were inherently more predisposed to PIU, as the underlying cognitions for

this behavior were maladaptive.

Caplan (2003) built off of Caplan (2002) and found further support for this

contention: Scores on loneliness and depression measures significantly predicted

participant’s preference for SIU, compared to face-to-face interaction. It was also found

that preference for SIU predicted a significant amount of variance in symptoms of PIU,

and preference also mediated the relationship between psychosocial health factors and

negative outcomes of SIU. Caplan (2005) expanded upon these results and found further

support for the idea that a preference for SIU leads to PIU and greater negative outcomes.

Caplan found that a perceived lack of self-presentational skill led to greater PIU,

providing support for the hypothesis that the unique format of the Internet causes some

individuals to prefer it, and also leads to negative outcomes.

In contrast to the previous perspectives, LaRose et al. (2003) expressed

disagreement with Young (1996) and what they viewed as an overly-pathologized

conceptualization of internet use as an addiction, and attempted to expand upon the

underlying behavioral mechanisms of PIU. LaRose and colleagues believed that

deficient self-regulation in particular, defined as the inability to self-monitor an activity

appropriately, was an important factor in the development of PIU. In line with classical

conditioning approaches, they hypothesized that individuals who use the Internet to cope

with depression or loneliness become conditioned to engage in this behavior,

experiencing increasing incentives to behave in this manner. The dysphoric mood states

that lead to internet use also prevent the ability to self-regulate the degree of internet use

that is undergone, thus leading to increasingly deficient self-regulation and PIU. Thus,


73

this perspective argues that psychosocial problems hinder behavior, rather than precede

behavior, in the development of PIU. LaRose and colleagues conducted a study on 465

college students, and found that there was a significant positive relationship between

deficient self-regulation and Internet use in general. This relationship held when only

those who met criteria for Internet addiction were included, based on proposed criteria

from Kimberly S. Young and Rogers (1998). Finally, depressive symptoms were found

to significantly relate to deficient self-regulation of internet use. LaRose and colleagues

concluded that deficient self-regulation played a large role in the development of PIU.

Caplan (2010) attempted to update the cognitive-behavioral model presented in

Caplan (2002) by integrating the model with findings in LaRose et al. (2003) that using

the internet to alleviate depressed mood led to deficient self-regulation. Caplan proposed

a model wherein preference for SIU predicts desire to use the internet for mood

regulation, and both of these factors then led to deficient self regulation. This inability to

self-monitor then leads to negative outcomes associated with PIU. Caplan found strong

evidence for the model and argued that this study provided further support for the link

between using the internet for mood regulation and negative outcomes.

Consistent with the results found in Caplan’s studies (Caplan, 2002, 2003, 2005,

2010), several studies have found support for the idea that seeking others out online may

be harmful for mental health. In a review of internet addiction studies, Weinstein and

Lejoyeux (2010) found that excessive use of the internet is often highly comorbid with

mood disorders, anxiety disorders, and ADHD in certain populations. Morahan-Martin

and Schumacher (2003) found that lonely individuals preferred SIU more than their non-

lonely counterparts, tended to make more friends online, reported heightented satisfaction


74

with their online friends, used the internet more often for emotional support, and believed

that their Internet use caused disturbances in their functioning. Consistent with these

findings, time spent online has been shown to have a positive association with feelings of

loneliness (Matsuba, 2006; Stepanikova et al., 2010), and has been shown to potentially

increase feelings of loneliness (Kim et al., 2009). Building off of these studies, Yao and

Zhong (2014) performed a longitudinal study in order to assess the possible casual

relationships between Internet Addiction, loneliness, and depressive symptoms in Hong

Kong university students, and found that Internet Addiction did lead to increased feelings

of loneliness. Although this does not support the cognitive-behavioral perspective put

forth by Davis (2001), which argues that loneliness precedes excessive Internet use, it

does provide support for the relationship between excessive Internet use and negative

psychosocial outcomes. Increased time spent online has been found to associate with

poorer coping strategies and interpersonal relationships in general (Milani et al., 2009),

and increased time spent online with social motivations has been shown to relate to

decreased social integration (Weiser, 2001).

Limitations of PIU Perspectives. However, some researchers have questioned

the validity of results found in studies of PIU. Byun et al. (2009) reviewed 39

quantitative studies on Internet Addiction which were published between 1996 to 2006,

in order to provide suggestions for the direction of future research in this area. Byun and

colleagues determined that there were several significant methodological concerns which

impacted the interpretability of results found in these studies. First, it was argued that the

definition for this phenomenon is highly inconsistent between researchers (see Definition

section). In addition to difficulties with consistency in definitions, Byun and colleagues


75

noted that the conceptualization of Internet Addiction also tended to change considerably

between researchers. Some Internet Addiction perspectives adopted a framework based

off of gambling addictions (Young, 1998) while other perspectives based their

conceptualization off of substance abuse disorders (Kaltiala-Heino, Lintonen, & Rimpela,

2004). Byun and colleagues also stated that researchers frequently use inconsistent

criteria for measurement of the disorder as well, noting that several popular measures had

been developed which did not assess similar hypothesized antecedents of Internet

Addiction (Morahan-Martin & Schumacher, 2000; Young & Rogers, 1998). Based on

the inconsistent conceptualizations and criteria used in Internet Addiction research, it was

argued that it is meaningless to try to compare results across these studies (Kaltiala-Heino

et al., 2004). Byun and colleagues also criticized Internet Addiction researchers for

frequently using convenience samples of adolescents and college students, which may

inflate reported rates of Internet Addiction in these studies, as well as frequently having

small sample sizes. Finally, Byun and colleagues noted that Internet Addiction

researchers frequently use exploratory, rather than confirmatory, approaches in studies.

Tokunaga and Rains (2010) performed a meta-analysis of 100 studies on PIU

conducted between the years of 2000 and 2009, in order to assess the evidence for and

against the cognitive-behavioral model put forth by Caplan (2002) and Davis (2001), and

the deficit self-regulation model proposed by LaRose et al. (2003). Tokunaga and Rains

assessed the overall relationships between PIU, time spent using the Internet, social

anxiety, loneliness, and depression. They conducted a path analyses and determined that

the deficient self-regulation model (LaRose et al., 2003) fit the data well: Psychosocial

factors (loneliness, depression, and social anxiety) lead to PIU (i.e. deficit self-


76

regulation), which then leads to increased time spent online. However, the pathology

model (Caplan, 2002; Davis, 2001), describing a path from psychosocial factors to time

spent online, which then leads to PIU, did not fit well to the data. Tokunaga and Rains

(2010) also described several limitations of research in this area: PIU studies have almost

largely been cross-sectional in nature, are frequently limited in scope, and are only able to

provide a ‘snapshot’ of a potentially cyclical process between psychosocial factors and

internet use behaviors. Thus, there remains the possibility that psychosocial difficulties

and SIU share a more complex relationship than those hypothesized by PIU perspectives.

For example, curvilinear relationships between variables such as the number of friends on

social networks and psychological well-being have been found (LaRose et al., 2014):

Well-being has been shown to increase with number of friends up to a certain point, at

which point well-being begins to drop off.

In contrast to research on PIU, several studies have found that there are possibly

more benefits to SIU than negative consequences. After describing the original Internet

Paradox (Kraut et al., 1998), Kraut et al. (2002) revisited the sample of their previous

study three years later in order to determine the long-term effects of internet use. Kraut

and colleagues found that the effects found in the previous study had largely diminished:

The association between depression and internet use had significantly declined within a

three year period, the association between loneliness and internet use was now non-

significant. In addition, participants in the study reported that there was no longer a

negative impact of internet use on their level of familial communication and the size and

strength of their interpersonal circles. However, increased internet use was associated

with an increase in daily life stress over this three year period. Kraut and colleagues


77

performed a second longitudinal study with a different sample to assess the

generalizability of the results found in the first study and in Kraut et al. (1998). The

results of Study 2 closely resembled those of Study 1: Heavier internet users found

significant increases in their social circles and face-to-face communication with friends

and family, their involvement in community activities increased, and most psychological

well-being factors were positive. However, internet use was again associated with

increased daily life stress, as well as less commitment to the local area.

Kraut and colleagues hypothesized that the diminishment of negative outcomes

between Kraut et al. (1998) and Kraut et al. (2002) could have been due to several

possible reasons. One possibility put forth was that the sample in Kraut et al. (1998) was

comprised primarily of new Internet users, and that the novelty of the Internet led to

frequent unrewarding use at first. As participants matured in their Internet use, the

Internet itself matured, more people began using the Internet in general, and avenues for

social interaction became more plentiful as a result, participants may have been able to

focus more often on personally rewarding uses and eschew the unnecessary uses.

SIU has been associated with decreased depression (Morgan & Cotten, 2003;

Shaw & Gant, 2002), decreased loneliness, improved self-esteem and perceived social

support (Shaw & Gant, 2002), and improved connection with family, friends, and

individuals with shared interests (Amichai-Hamburger & Hayat, 2011). In addition,

internet users who engage in SIU report specifically perceiving that this kind of internet

use can be psychologically beneficial to them, although they also report believing that the

Internet can be addictive (Campbell et al., 2006). Facebook use in particular has been


78

shown to relate with decreased depression and anxiety scores, as well as an increase in

perceived well-being (Grieve, Indian, Witteveen, Tolan, & Marrington, 2013).

Internet Support Groups

Although research on the benefits and consequences of SIU is conflicted, studies

on the effects of internet support groups (ISGs), or online support groups, show strong

support for the potential benefits of this behavior. ISGs are online message boards,

forums, or instant messaging clients where members may talk to each other and discuss

problems or share information. Use of ISGs is reported to be beneficial for individuals

who are managing a variety of concerns and significant health conditions, including

hysterectomies (Bunde et al., 2006), visual impairment (Smedema & McKenzie, 2010),

cancer (Beaudoin & Tao, 2007; Han et al., 2008; Seckin, 2013; Shim et al., 2011),

suicidal ideation (Gilat & Shahar, 2009), HIV/AIDS (Mo & Coulson, 2010), and

Parkinson’s Disease (Attard & Coulson, 2012).

Barak et al. (2008) reviewed quantitative literature on ISGs and concluded that

these groups provided multifaceted components which increased participants’ self-

empowerment and led to improved well-being. Barak and colleagues noted that

communication on the Internet allowed for decreased inhibition and a subsequently

increased willingness to share information with others, dubbed the disinhibition effect

(Suler, 2004). Consistent with Joinson (2001) and Suler (2004), Barak and colleagues

determined that the Internet provides a sense of anonymity which allows for less feelings

of vulnerability and an increased comfort in sharing with others. Consistent with findings

in Schouten et al. (2007) and Stritzke et al. (2004), it was found that a perception of

invisibility is also beneficial to participants, in that their self-presentational concerns are


79

reduced and others’ negative reactions are not visible. Finally, Barak and colleagues

determined that the tendency of delayed communication in online settings (i.e.

communication with several minutes or hours in between) provides a benefit, in that it

allows participants the opportunity to process their thoughts and feelings longer at a

deeper level. Finally, the invisibility and anonymity provided by the Internet allow for a

secondary effect of neutralizing the statuses of participants in the group. Barak and

colleagues argued that the inability to assess factors related to power and status, namely

physical appearance, job status, etc., allowed for a decreased inhibition and a greater

willingness to speak up without fear of authority. Despite the disinhibition effect often

leading to beneficial outcomes that are typically considered crucial in successful support

groups, such as an increased willingness to share deeper emotions and be honest, it was

noted that disinhibition may also lead to negative outcomes, in the form of rude language,

harsh criticism, anger, or hatred (Tanis, 2007).

Barak and colleagues also reviewed studies which showed that the act of

participating in an online support group provided benefits above and beyond those which

led to disinhibition. One primary benefit to online support groups is the catharasis that

can be gained from writing about one’s experiences. In one study reported by Barak et

al. (2008) in this area, Scandinavian participants who told stories about their experiences

with breast cancer experienced an increased sense of self-empowerment and control

(Hoybye, Johansen, & Tjornhoj-Thomsen, 2005). Hoybye and colleagues reported that

the shift from being acted upon (i.e. breast cancer) to being active in the storytelling

process led to this shift. Barak and colleagues also reviewed studies which show that the

benefits derived from face-to-face support groups translated to online support groups as


80

well; Online support groups fostered an increased connection to emotions, the ability to

gain information and knowledge about one’s presenting concern, the opportunity to gain

social support and increase social networks, and the ability to reexamine one’s decision-

making.

In addition to reviewing quantitative studies on the benefits of online support

groups, Barak et al. (2008) assessed the qualitative experiences of participants who

participate in these support groups, and found that there were some cons to this

experience which may be maladaptive or harmful. Participants reported that one aspect

of online support groups they considered problematic was the sheer number of groups

available. Choosing a “good” group was often difficult, due to the wide variety of groups

to investigate. Some groups were found to provide incomplete or inaccurate descriptions

in listings, and group size could range from overwhelmingly large to small and inactive.

Information provided by these groups could sometimes be outdated, biased, or outright

wrong, which could be particularly harmful with more serious presenting concerns.

Additionally, new participants in groups oftentimes had to quietly observe the support

group for a period of time to get a sense of the temperament and norms in the group.

However, participants reported several positive aspects of their experiences, as well.

Participants reported that the therapeutic factor of these groups were surprising, and

indicated particular surprise at how strong their emotional reactions were to discussions.

In addition, many participants reported making strong one-on-one relationships with

other group members, with these members often providing support and coaching on how

to share more effectively in the group.


81

Consistent with findings in Barak et al. (2008), and in contrast to perspectives on

PIU which argue that seeking out others online can be harmful (Caplan, 2002; Davis,

2001), the social and emotional connections formed with others on these ISGs are

frequently found to be the most important predictor of the psychological benefits derived

from participating (Han et al., 2008; Shim et al., 2011), and users who do not form these

connections appear to have their online experience compromised significantly (Attard &

Coulson, 2012). Consistent with studies reviewed by Barak et al. (2008), preference for

ISGs has also been shown to increase when individuals are dissatisfied with current face-

to-face support groups (Chung, 2013), suggesting that ISG use may be perceived by some

as analogous to traditional support groups.

Consistent with research on ISGs, one study (Leung, 2007) has found strong

support for the contention that specifically using the internet for mood management and

social compensation may be beneficial. Leung assessed 717 children and adolescents for

internet use motives, stressful life events, and perceived social support. Lueng found that

higher levels of stress were related to higher degrees of mood management and social

compensation motives for the internet. Leung also found that higher social support,

whether online or offline, provided a buffer against many stressful live events. Although

cross-sectional in nature, this study, in addition to previously discussed research on ISGs,

suggests that those who might benefit from SIU are those who are highly distressed and

also willing to use the internet for social or emotional support. Several mental health

factors related to distress have been shown to relate to SIU, specifically use of the

internet for social support seeking and emotional coping.


82

Who might benefit from support on the Internet?

Studies on Problematic Internet Use are rife with methodological concerns which

jeopardize the validity and generalizability of the results (Byun et al., 2009), and

ultimately hinder the strength behind the argument that use of the Internet for social

reasons is inherently harmful. In addition, the most empirically supported approach in

the realm of PIU argues that excessive Internet use is not addictive but largely a result of

poor self-regulation skills, a behavior which can consciously be reversed by the

individual if desired (LaRose et al., 2003; Tokunaga & Rains, 2010). In contrast to PIU

research, qualitative and quantitative reviews of studies on ISGs (Barak et al., 2008) and

research on childhood and adolescent relationships (Leung, 2007) demonstrate that there

are considerable benefits to engaging in emotional support and social support seeking on

the Internet. However, data in this area is frequently qualitative and cross-sectional in

nature. This body of research strongly suggests that it is not a matter of whether or not

the Internet can be beneficial for support, but for whom and when will it be most

effective.

Depressive Symptoms. Research specifically on depression ISGs has

demonstrated that depressed individuals specifically seek out emotional support when

they are online: A meta-analysis of thirteen studies on depression ISGs, conducted by

Griffiths et al. (2009), determined that the content of the posts for depression ISGs

contained a significantly greater degree of emotionally supportive content, compared to

other kinds of ISGs. Griffiths and colleagues found that many who participated in ISGs

for depression distinctly reported that an attractive feature of the groups was a tendency

to feel emotionally supported and to feel a reduction of their loneliness. Posts on


83

depression ISGs also frequently contained content related to social companionship

(Muncer, Burrows, Pleace, Loader & Nettleton, 2000).

Houston et al. (2002) followed 103 users of a depression ISG for one year to

determine the impact of the support group on their interpersonal relationships, well-

being, and their level of depressive symptoms. A large majority of participants had been

diagnosed with a depressive disorder by a health-care professional (N=101) and were

currently in treatment for depression at baseline (N=96). Around one quarter of the

participants reported little to no social support (18%-25%), and perceptions of social

support did not significantly deviate throughout the study. However, participants at the

1-year mark reported receiving an average of 26-50% of their social support through the

Internet. Around one-third of participants (33.8%) reported a decrease in depressive

symptoms, with this decrease happening more often in participants who frequently

participated in the ISG than those who participated less frequently. Additionally, the

majority of participants reported perceiving that these support groups were helpful for

managing their symptoms.

Loneliness. Loneliness has been a widely assessed construct in the motivations

for SIU (Caplan, 2003), as well as the possible negative outcomes of SIU (Kim et al.,

2009). This construct may be assessed often due to the oft held belief in the population

that heavy Internet users are frequently lonely (Campbell et al., 2006). Studies have

found that higher SIU is associated with increased loneliness (Brandtzaeg, 2012; Weiser,

2001), and longitudinal studies have found that loneliness both precedes excessive

internet use (Yao & Zhong, 2014) and is a consequence of it (Kim et al., 2009). Studies

have shown that individuals who report high loneliness and avoidant coping styles will


84

not attempt to cope differently when they are online, suggesting that they will not use the

Internet adaptively (Seepersad, 2004). In contrast, some studies have found that lonely

individuals do use the Internet adaptively, and report positive outcomes as a result.

Valkenburg and Peter (2008) surveyed 1,158 Dutch adolescents and found that

participants who reported more loneliness were more likely to engage in identity

experimentation, and led to a greater sense of social competence. Valkenburg and

colleagues argued that the Internet could serve as a platform for lonely individuals to

practice social skills due to the relative anonymity it affords, which is consistent with

previous research on the effects of perceived anonymity in online communication

(Joinson, 2001). Other studies have found that individuals who report more loneliness

actually have more Facebook friends, suggesting that they are trying to compensate for

face-to-face interactions through online interactions (Skues, Williams, & Wise, 2012).

The exact role of loneliness in SIU has not been consistently found. Some

research has suggested that the inconsistency found is possibly due to loneliness being

more appropriate as an indirect predictor, or that other conceptually-similar factors are

possibly better predictors than loneliness. Caplan (2003) found that loneliness and

depression together predicted 19% of the variance in preference for online interaction.

However, loneliness by itself only accounted for 1% of the variance, and was also

mediated by several other factors in predicting negative outcomes of internet use. Caplan

(2007) argued that the inconsistent findings for loneliness as a predictor is due to the

assumption of previous research that all lonely individuals would be drawn to the

Internet. Caplan argued that there is frequently a conflagration that individuals

experiencing situational loneliness (e.g., lack of time for social activities, recent move to


85

a new city) will be similar to those with dispositional loneliness, and that many lonely

individuals would not necessarily be drawn to the Internet to interact socially. Caplan

argued that social anxiety would serve as a more conceptually accurate predictor, as

socially anxious individuals would have a specific motivation to take advantage of the

anonymity afforded by the Internet in their social interactions (Joinson, 2001). Caplan

assessed 343 undergraduate students on preference for online social interaction, social

anxiety, loneliness, and negative outcomes of internet use. Caplan found that social

anxiety explained a significant, unique amount of variance in the preference for online

social interaction. Caplan also found that when loneliness was added to the model after

social anxiety was entered, it did not produce a significant increase in variance explained.

Caplan concluded that social anxiety appeared to be a good predictor of preference for

online social interaction, and that the relationship between loneliness and preference for

online social interaction is spurious. Consistent with the arguments made by Caplan

(2007), Lee (2013) acknowledged the inconsistent findings of loneliness and sought to

determine more clearly the relationship between loneliness and negative outcomes. Lee

assessed 265 South Korean students on their loneliness, their degree of self-disclosure on

social networks, their perception of their social support, and their well-being. Lee found

that social support mediated the relationship between self-disclosure and well-being for

lonely individuals. The results of this study suggest that the presence of strong ties online

may have a more direct role in the benefits and consequences of SIU than loneliness

itself.

Social Anxiety. SIU is argued to be an attractive mode of communication to

socially anxious individuals, particularly due to the relative anonymity afforded by the


86

Internet, the lack of self-presentational cues, and the ease of controlling the pace and tone

of conversations with others (Caplan, 2007; Joinson, 2001; McKenna & Bargh, 2000).

Research has supported this contention: Studies have shown that individuals with social

anxiety specifically benefit from an increase in interpersonal potential while online (High

& Caplan, 2009), and possibly benefit from an increased inability to react to negative or

inhibitory feedback cues from others (Schouten et al., 2007; Stritzke et al., 2004). It is

argued that socially anxious individuals may additionally benefit from SIU as a low-risk

approach to practicing social interaction skills, in order to improve on subsequent face-to-

face interactions (Campbell et al., 2006). Consistent with this research, increases in

social anxiety were found to be related to increased desire to use the Internet for coping

purposes (Gordon et al., 2007). As social anxiety increases, individuals have been shown

to display greater motivation to make new friends via blogs and disclose more on blogs,

which in turn is associated with higher quality friendships made (Tian, 2013). However,

while socially anxious individuals do report greater comfort and self-disclosure when

online, they may also experience poorer well-being due to the conduciveness of SIU in

avoiding face-to-face interactions (Weidman et al., 2012).

Neuroticism. Neuroticism, a trait defined by emotional instability and negative

affect (McCrae & Costa, 1987), is the personality trait most often related to increased

SIU. Studies have determined that a higher level of neuroticism is related to increased

use of social Internet sites like Facebook (Hughes et al., 2012; Seidman, 2013; Wolfradt

& Doll, 2001) and to higher SIU in general (Kalmus et al., 2011). The specific social and

emotional motivations of neurotic individuals are supported: Highly neurotic individuals

report specific motivations to increase companionship and reduce loneliness through their


87

Internet use (Amiel & Sargent, 2004), to use blogs for self-expression (Guadagno et al.,

2008), and to express their “true selves” to others (Amichai-Hamburger et al., 2002;

Tosun & Lajunen, 2010). It has been argued that highly neurotic individuals prefer the

Internet due to the greater control they have over their presentation and their statements

(Butt & Phillips, 2008; Nadkarni & Hofmann, 2012), suggesting that an online format is

specifically attractive to neurotic individuals who wish to form new relationships.

Extraversion. Both extraversion, a trait defined by higher sociability, liveliness,

and assertiveness, and introversion, the inverse trait (McCrae & Costa, 1987), have been

demonstrated to increase motivations for SIU in the literature. Researchers argue that the

significance found for both dimensions of this trait is due to the types of social Internet

activities being assessed: Extroverts prefer to use the Internet to build off of existing

relationships, while introverts prefer to use the Internet to make new friends (Amichai-

Hamburger et al., 2002; Bargh et al., 2002; Orchard & Fullwood, 2010; Tosun &

Lajunen, 2010). For example, extroverted individuals tend to report more participation

on Facebook in general (Jenkins-Guarnieri et al., 2012), and tend to be more involved on

other sites that focus primarily on previously-established friendships (Amichai-

Hamburger et al., 2008). Introverted individuals, on the other hand, tend to value more

anonymous Internet services such as ICQ and Instant Messaging (Amichai-Hamburger et

al., 2008; Amiel & Sargent, 2004), and appear to prefer SIU as a means of expressing

their true selves (Amichai-Hamburger et al., 2002; Zywica & Danowski, 2008),

suggesting that there are more social and emotional motives present for these individuals.

However, some studies have found the opposite pattern: Introverted individuals have


88

also been shown to communicate less and self-disclose less, leading to less friendship

formation online (Peter, Valkenburg, & Schouten, 2005).

Perceived Stress. A relationship between the presence of specific stressful life

events and motivation to use the Internet socially has been found (Leung, 2007). One

study to date has assessed the role of general perceived stress specifically on the use of

the Internet to cope emotionally (Deatherage et al., 2014). Deatherage and colleagues

assessed 267 college seniors on their degree of perceived stress, dispositional coping

styles, motivations for internet use, and problematic internet use. Emotion-focused

coping tendencies were found to be strongly positively associated with perceived stress,

and more specifically, coping-related motives to use the Internet (i.e. “to cheer up when I

am in a bad mood”) were also strongly positively associated with perceived stress.

Problematic internet use was not associated with perceived stress. Although this study

was correlational in nature, and thus did not provide support for a directional relationship

between perceived stress and SIU motivation, this study provides some tentative support

for the possibility of this relationship.

Cross-cultural Factors. More recently, research has begun to specifically focus

on the role cultural factors play in the impact of SIU on psychological well-being.

LaRose et al. (2014) assessed undergraduates in three countries (Ireland, Korea, and the

United States) on their levels of deficit self regulation, psychosocial outcomes, and their

perceived connection demands (i.e. feelings of social obligation to interact with friends

on social media),. LaRose and colleagues found that a strong sense of collectivism

appeared to buffer the inverse relationship between connection demands and negative

affect. They hypothesized that highly collectivist participants might see connection


89

demands as a natural extension of their face-to-face social activities, lessening possible

negative affective reactions.

Gender. Some research has specifically assessed the impact of gender on SIU

motivations. Kimbrough, Guadagno, Muscanell, and Dill (2013) assessed 381

undergraduate students to determine gender differences in general SIU motivations. It

was found that women were more likely than men to have a preference for SIU, including

the use of text messaging, social networking in general, and video chatting. Women were

found to also video chat for a significantly greater amount of time than men. The authors

conclude that this provides support for the influence of social roles, with women

behaving more communally online and engaging in more SIU. However, this study was

limited in several respects: In terms of its methodology, this study used a 26-item survey

to assess all constructs of interest, and the psychometric properties of these items were

not reported, nor established in the previous literature. Additionally, the majority of the

sample was White (85%), limiting generalizability considerably.

Some studies have specifically assessed gender differences in social network use.

Muscanell and Guadagno (2012) assessed 238 undergraduate students who reported

being a member of one social networking site, and found that gender and personality

differentially impacted preferences for SIU on social networks. Muscanell and Guadagno

found that men were actually more likely to engage in some aspects of SIU: Namely,

men reported greater use of social network sites to find dates, network for careers, and to

make new friends. Women reported posting more public messages on these sites and

sending more private messages to others. Personality also interacted with gender: Men

who were low in agreeableness tended to make more blog posts than men who were high


90

in agreeableness, and women who were low in agreeableness tended to Instant Message

others more often than women who were high in agreeableness.

Although gender has been shown to impact motivation for SIU to an extent, the

data is not entirely consistent. Research seems to show that men and women are not

necessarily different in their online sociability, but that they are different in which kinds

of social online activities they prefer (Kimbrough et al., 2013; Muscanell & Guadagno,

2012). Research on gender differences in SIU as a coping mechanism appear to support

this: Online support groups for depression which report gender makeup have shown an

even split between groups which are predominately female and groups which are

predominately male (Griffiths et al., 2009).

Gaps in the Literature

Although there is varying levels of empirical support for each of these factors, no

overarching model has attempted to show the multivariate relationships between all of

these factors in regards to their involvement with healthy, non-excessive degrees of social

and emotional coping on the Internet. Most prominent research on models of Internet use

in general focus primarily on this behavior as inherently aberrant and harmful, especially

when considering the social components (Caplan, 2002, 2010; Davis, 2001; Young,

1998). These models persist, despite a strong, growing body of literature demonstrating

that there are numerous potential benefits to engaging with others online (Kraut et al.,

2002), as well as considerable methodological and conceptual concerns which threaten

the validity of these models and their results (Byun et al., 2009; Tokunaga & Rains,

2010) and possible conflicts of interest for some researchers (LaRose et al., 2003). Thus,

the possible positive impact of SIU is still a nascent research area. The majority of


91

research in this area is focused on the role of Internet Support Groups (Barak et al.,

2008), and is often qualitative in nature rather than empirical. For example, a

considerable portion of research on the relationship between depression and SIU is

comprised of qualitative ISG studies (Griffiths et al., 2009). Studies that have

empirically assessed the relationships between factors in a positive manner tend to only

look at two or three factors at a time (Leung, 2007).

This study will fill a hole in the literature by being the first of its kind to create a

more comprehensive, empirical model of the various psychosocial characteristics which

have been shown to predispose individuals to SIU. This study will also explore the

multivariate relationships between these variables, some of which have not been assessed

yet in the literature. Finally, this study will assess SIU as a possible positive coping

mechanism, in contrast to other major models of Internet use which assert that it is

inherently harmful.

Conclusion

The Internet is becoming more social in nature, and individuals are increasingly

starting and maintaining relationships online. This is the first study known to assess the

roles of, and relationships between, multiple empirically-supported factors in SIU as a

positive social and emotional coping mechanism. In determining the relationship

between these factors, this study hopes to further increase understanding of what might

influence the attractiveness of engaging in SIU for support, and who might find it most

appealing. A strong relationship between these factors would demonstrate that

individuals have multifaceted emotional and social motivations for using the Internet to

meet others. Consistent with some therapeutic orientations (Palmer-Olsen et al., 2011;


92

Rogers, 1961), researchers have argued that the ability to express one’s “true self”, even

over the Internet, could be helpful for neurotic, introverted, or socially anxious

individuals, as these individuals often have difficulty expressing themselves in face-to-

face communication (Amichai-Hamburger et al., 2002; Bargh et al., 2002; Ebeling-Witte

et al., 2007). Some research has found initial support for this contention, finding that

even the use of blogs for self-expression leads to positive outcomes therapeutically

(Hillan, 2003), possibly due to an increase in perceived social support (Baker & Moore,

2008). Research on Internet Support Groups strongly supports this contention as well,

finding that users receive several psychosocial benefits from participation (Barak et al.,

2008).

Although there is not yet a complete consensus on the actual positive or negative

consequences of SIU (Huang, 2010), a greater understanding of the factors behind these

behaviors may still benefit therapists and other mental health practitioners. By having a

greater understanding of these psychosocial factors as they relate to SIU, therapists may

become better equipped to work with clients who are seeking and maintaining

relationships with others online. This will become increasingly relevant in

psychotherapy, as the population of Internet users is only expected to grow in the future.

Understanding these behaviors as potential coping strategies and methods of self-

expression for those experiencing stress, depression, social difficulties, or emotional

instability may allow for more accurate conceptualizations and interventions. Future

application of SIU as both a social and emotional support mechanism, for certain types of

clients, is also a possibility.


93

Appendix B

Table 2: Descriptive Statistics for Full Measures

α Minimum Maximum Mean Std. Deviation Variance Skewness Kurtosis

SDS .73 4 31 16.05 4.86 23.64 .08 .07

CES-D .90 0 48 17.06 10.48 109.74 .61 -.27

LSAS .96 0 110 44.54 25.24 637.12 .23 -.55

BFI (Neuroticism) .73*

8 38 24.00 5.83 34.02 -.23 -.06

BFI (Extroversion) 13 40 26.61 5.74 32.99 .13 -.53

COPE (Emotional Support) .91 4 16 9.99 3.51 12.31 .07 -.90

IMS (Social) .93* 4 35 19.88 5.68 32.26 -.17 -.01

SCS-R .82 27 113 72.56 12.56 157.88 .18 1.25

GPIUS-2 .94 8 111 44.29 21.35 455.99 .44 -.50

PSS-10 .81 1 34 18.23 6.03 36.32 -.12 .17

Note: *Cronbach’s alpha coefficients reported for full measure. SDS: Social Desirability Scale; CES-D: Center for Epidemiological Studies – Depression Scale; LSAS: Liebowitz Social Anxiety Scale; BFI: Big Five Inventory; SCS-R: Social Connectedness Scale – Revised; GPIUS – 2: Generalized Problematic Internet Use Scale – 2; PSS-10: Perceived Stress Scale – 10.


94

Table 3: Correlations between measures

SDS CES-D LSAS BFI-N BFI-E COPE IMS SCS-R GPIUS-2 PSS-10

SDS 1 CES-D -.243** 1 LSAS -.258** .263** 1 BFI-N -.428** .456** .334** 1 BFI-E .064 -.245** -.382** -.154** 1 COPE .093 -.076 .022 .121* -.189** 1 IMS -.158** .202** .173** .163** -.049 .034 1

SCS-R -.021 -.058 -.045 -.044 .113 .122* .515** 1 GPIUS-2 -.182** .346** .343** .243** -.144* .043 .608** .272** 1 PSS-10 -.344** .593** .304** .596** -.084 .031 .207** -.05 .308** 1

*Correlation is significant at .05 level (2-tailed) **Correlation is significant at .01 level (2-tailed)

Note: SDS: Social Desirability Scale; CES-D: Center for Epidemiological Studies – Depression Scale; LSAS: Liebowitz Social Anxiety Scale; BFI-N: Big Five Inventory, Neuroticism subscale; BFI-E: Big Five Inventory, Extraversion subscale; SCS-R: Social Connectedness Scale – Revised; GPIUS – 2: Generalized

Problematic Internet Use Scale – 2; PSS-10: Perceived Stress Scale – 10.


95

Table 4: Fit of initial measure and final measure with items retained

Fit of Initial Model Fit of Final Model Items Retained

CES-D χ ² = 409.78 (df = 164, p=.000)

RMSEA = .07 (.063 - .08)

CFI = .89

χ ² = 140.95 (df = 77, p=.000)

RMSEA = .053 (.039 - .067)

CFI = .97

1, 2, 5, 6, 7, 9, 10, 11, 13,

14, 15, 17, 19, 20

IMS χ ² = 18.94 (df = 9, p=.026)

RMSEA = .061 (.021 - .100)

CFI = .98

--- all

COPE χ ² = 11.61 (df = 2, p=.003)

RMSEA = .128 (.064 - .204)

CFI = 1.0

--- all

LSAS χ ² = 748.41 (df = 252, p=.000)

RMSEA = .082 (.076 - .089)

CFI = .93

χ ² = 72.93 (df = 27, p=.000)

RMSEA = .077 (.056 - .098)

CFI = .98

5, 7, 10, 11, 15, 18, 19,

23, 24

PSS-10 χ ² = 8.26 (df = 5, p=.142)

RMSEA = .047 (.000 - .102)

CFI = .99

--- all

BFI-N χ ² = 170.58 (df = 20, p=.000)

RMSEA = .160 (.139 - .183)

CFI = .89

χ ² = 23.37 (df = 9, p=.005)

RMSEA = .074 (.038 - .111)

CFI = .98

4, 9, 14, 19, 29, 39

BFI-E χ ² = 293.25 (df = 20, p=.000)

RMSEA = .216 (.194 - .238)

CFI = .88

χ ² = 51.44 (df = 9, p=.000)

RMSEA = .127 (.094 - .162)

CFI = .97

1, 6, 11, 21, 26, 31

SCS-R χ ² = 3017.65 (df = 170, p=.000)

RMSEA = .240 (.232 - .247)

CFI = .41

χ ² = 36.64 (df = 9, p=.000)

RMSEA = .099 (.065 - .135)

CFI = .98

1, 5, 8, 10, 14, 16

GPIUS χ ² = 18.20 (df = 5, p=.003)

RMSEA = .095 (.051 - .144)

CFI = .98

--- all


96

Table 5: Path loadings and correlated disturbances for saturated model

Neuroticism Extraversion Depression Social Anxiety ESS Coping Perceived Stress

Motives Positive Outcomes

Negative Outcomes

Neuroticism 1

Extraversion -.22** 1

Depression .49** -.10 1

Social Anxiety .32** -.33** .01 1

ESS Coping .18** .23** -.06 .04 1

Perceived Stress .76** .08 .45** .10 -.12 1

Motives -.02 .01 .15 .10 .05 .14 1

Positive Outcomes -.04 .01 .13 -.09 .07 -.21 .76** 1

Negative Outcomes -.11 -.01 .17* .20** .01 .12 .59** .12 1

Note: *Significant at .01 level (2-tailed), **Significant at .05 level


97

Table 6: Participants’ Reported Internet Use

Website or Activity Percentage endorsed use (daily) Range (daily) M (SD)

Facebook (accessing) 75% 0 – 370 min. 33.8 (46.5)

Facebook (posting) 28% 0 – 4 posts .38 (.66)

Texting 99% 0 – 1000 texts 127.7 (165.3)

Facetime/Skype/etc. 30% 0 – 300 min. 15.5 (38.2)

Reddit 6% 0 – 180 min. 3.6 (19.8)

MMORPGs 6% 0 – 400 min. 10.0 (45.0)

Twitter 55% 0 – 250 min. 30.0 (43.6)

Microblogs

Tumblr 14% 0 – 300 min. 58 (57.7)

Pinterest 28% 0 – 1000 min. 49.1 (110.0)

Instagram 52% 0 – 500 min. 62.9 (67.8)

Message Boards/Forums

1 message board/forum 5%

2+ message boards/forums 1%

Dating Websites

Any 9%

Tinder 6%


98

Appendix C

Figure 1: Measurement Model. Predictors for the latent variables Depression, Introversion, Neuroticism, Social Anxiety, Perceived Stress, and SIU

Motivations are shown as parcels. Curved lines indicate estimate correlations of latent variables.

ESS Coping Depression Positive Outcomes

Social Anxiety

Internet Motivations

Perceived Stress

Neuroticism Negative Outcomes

Extroversion

CESD 1

CESD 2

CESD 3

BFI 1

BFI 2

BFI 3

BFI 4

BFI 5

BFI 6

COPE 1

COPE 2

COPE 3

COPE 4

LSAS 1

LSAS 2

LSAS 3

LSAS 4

PSS 1

PSS 2

PSS 3

IMS 1

IMS 2

IMS 3

SCS 1

SCS 2

SCS 3

GPIUS 1

GPIUS 2

GPIUS 3

E1 E2 E3 E4 E5 E6 E7 E8 E9 E10 E11 E12 E13 E14 E15 E16 E17 E18 E19 E20 E21 E22 E23 E24 E25 E26 E27 E28 E29

Depressive Symptoms

ESS Coping


99

Figure 2: Hypothesized Structural Model

Note: (+) relation hypothesized to be positive, (-) relation hypothesized to be negative


100

Figure 3: Path Loadings for Initial Hypothesized Model

Note: IC = inconsistent with hypotheses; C = consistent with hypotheses


101

Figure 4: Path Loadings for Alternative Model


102

Figure 5: Path loadings for Saturated Model

Note: Non-significant and non-hypothesized paths are not shown. See Table 5 for full path loadings for saturated model.


103

Appendix D

1. What best describes your gender? Please choose one.

Male Female Transgender

2. How old are you? __________

3. What is your race? Please circle one or more as appropriate.

Caucasian/White

African American/Black

Hispanic/Latino

Asian American/Asian/Pacific Islander

Indian/Middle Eastern

Native American

My race is not listed here (please specify): ______________________

4. What is your current class year at Texas Tech? Please circle one.

First year

Second year

Third year

Fourth year

Fifth year

Other (please specify): _______________________

5. Are you currently in a romantic relationship? Y N

If yes, circle the type of relationship which most applies.

Married/Engaged Dating Open Other (please specify): _________


104

This survey measures the amount of time per day, on average, that you spend on

various online activities. If you do not use a particular service, please write “n/a” into one of the blanks for that service. If you access more than 3 blogs, forums/message boards, or dating websites on a given day, write in the 3 ones you use most.

*Facebook: _____ minutes per day, ____ # posts per day, _______# messages sent per day

*Reddit: _____ minutes per day, ______# posts per day

*Twitter: _____ minutes per day, ______# of tweets per day

*Email: _____ minutes per day, ____# logins per day, ______# of messages sent per day

*Skype, Facetime, other video chat services: _____ minutes per day

*Text Messages: _____ # of texts sent per day

*Instant Messaging (MSN, AIM, etc.): _____ minutes per day

*MMORPGs (World of Warcraft, League of Legends, etc.) : _____ minutes per day

*Last.fm, Spotify, Soundcloud, other social music sites: _____ minutes per day

*Message boards/forums:

1. Name: ____________________, minutes per day: ______, posts per day: _______



*Dating websites (OkCupid, PlentyofFish, Tindr, Grindr, etc.):

1. Name: ________________________, minutes per day: ________



*Blogs (Yours or others, Tumblr, Pinterest, Instagram, etc.):




105


Appendix E

Using the scale below, indicate the number which best describes

How often you felt or behaved in this way during the past week:

0 Rarely or none of the time (less than 1 day)

1 Some or a little of the time (1-2 days)

2 Occasionally or a moderate amount of time (3-4 days)

3 Most or all of the time (5-7 days)

1. I was bothered by things that usually don’t bother me. 0 1 2 3

2. I did not feel like eating, my appetite was poor. 0 1 2 3

3. I felt that I could not shake the blues, even with help from

my family and friends. 0 1 2 3

4. I felt that I was just as good as other people. 0 1 2 3

5. I had trouble keeping my mind on what I was doing. 0 1 2 3

6. I felt depressed. 0 1 2 3

7. I felt that everything I did was an effort. 0 1 2 3

8. I felt hopeful about the future. 0 1 2 3

9. I thought my life had been a failure. 0 1 2 3

10. I felt fearful. 0 1 2 3

11. My sleep was restless. 0 1 2 3

12. I was happy. 0 1 2 3

13. I talked less than usual. 0 1 2 3

14. I felt lonely. 0 1 2 3

15. People were unfriendly. 0 1 2 3

16. I enjoyed life. 0 1 2 3

17. I had crying spells. 0 1 2 3

18. I felt sad. 0 1 2 3


106

19. I felt that people disliked me. 0 1 2 3

20. I could not get “going.” 0 1 2 3

Appendix F

Here are a number of characteristics that may or may not apply to you. For example, do you agree

that you are someone who likes to spend time with others? Please write a number next to each

statement to indicate the extent to which you agree or disagree with that statement.

1

Disagree

Strongly

2

Disagree

a little

3

Neither agree

nor disagree

4 Agree

a little

5

Agree

strongly

I am someone who… 1. _____ Is talkative

2. _____ Tends to find fault with others

3. _____ Does a thorough job

4. _____ Is depressed, blue

5. _____ Is original, comes up with new ideas

6. _____ Is reserved

7. _____ Is helpful and unselfish with others

8. _____ Can be somewhat careless

9. _____ Is relaxed, handles stress well.

10. _____ Is curious about many different things

11. _____ Is full of energy

12. _____ Starts quarrels with others

13. _____ Is a reliable worker

14. _____ Can be tense

15. _____ Is ingenious, a deep thinker

16. _____ Generates a lot of enthusiasm

17. _____ Has a forgiving nature

18. _____ Tends to be disorganized

19. _____ Worries a lot

20. _____ Has an active imagination

21. _____ Tends to be quiet

22. _____ Is generally trusting

23. _____ Tends to be lazy

24. _____ Is emotionally stable, not easily upset

25. _____ Is inventive

26. _____ Has an assertive personality

27. _____ Can be cold and aloof

28. _____ Perseveres until the task is finished

29. _____ Can be moody

30. _____ Values artistic, aesthetic experiences

31. _____ Is sometimes shy, inhibited

32. _____ Is considerate and kind to almost everyone

33. _____ Does things efficiently

34. _____ Remains calm in tense situations

35. _____ Prefers work that is routine

36. _____ Is outgoing, sociable

37. _____ Is sometimes rude to others

38. _____ Makes plans and follows through with

them


114

39. _____ Gets nervous easily

40. _____ Likes to reflect, play with ideas

41. _____ Has few artistic interests

42. _____ Likes to cooperate with others

43. _____ Is easily distracted

44. _____ Is sophisticated in art, music, or literature


107

Appendix G

Instructions: The questions in this scale ask you about your feelings and thoughts during the last month. In

each case, please indicate with a check how often you felt or thought a certain way.

1. In the last month, how often have you been upset because of something that happened unexpectedly?

0 = never 1 = almost never 2 = sometimes 3 = fairly often 4 = very often

2. In the last month, how often have you felt that you were unable to control the important things in

your life?


3. In the last month, how often have you felt nervous and "stressed"?


4. In the last month, how often have you felt confident about your ability to handle your personal

problems?


5. In the last month, how often have you felt that things were going your way?


6. In the last month, how often have you found that you could not cope with all the things that you had

to do?


7. In the last month, how often have you been able to control irritations in your life?


8. In the last month, how often have you felt that you were on top of things?


9. In the last month, how often have you been angered because of things that were outside of your

control?


10. In the last month, how often have you felt difficulties were piling up so high that you could not

overcome them?



108

Appendix H

This measure assesses the way that social phobia plays a role in your life across a variety

of situations. Read each situation carefully and answer two questions about that

situation. The first question asks how anxious or fearful you feel in the situation. The

second question asks how often you avoid the situation. If you come across a situation

that you ordinarily do not experience, please imagine that you were faced with that

situation, and then rate the degree to which you would fear this hypothetical situation and

how often you would tend to avoid it. Please base your ratings on the way that the

situations have affected you in the last week.

Fear or Anxiety Avoidance

0 = None

1 = Mild

2 = Moderate

3 = Severe

0 = Never (0%)

1 = Occasionally (1-33%)

2 = Often (33- 67%)

3 = Usually (67 – 100%)

1. Telephoning in public

2. Participating in small groups

3. Eating in public places

4. Drinking with others in public places

5. Talking to people in authority

6. Acting, performing, or giving a talk in

front of an audience

7. Going to a party

8. Working while being observed

9. Writing while being observed

10. Calling someone you don’t know very

well

11. Talking with people you don’t know

very well

12. Meeting strangers

13. Urinating in a public bathroom

14. Entering a room when others are

already seated

15. Being the center of attention

16. Speaking up at a meeting

17. Taking a test

18. Expressing a disagreement or

disapproval to people you don’t know very

well

19. Looking at people you don’t know very

well in the eyes

20. Giving a report to a group

21. Trying to pick up someone

22. Returning goods to a store

23. Giving a party

24. Resisting a high pressure salesman


109

Appendix I

Then respond to each of the following items by blackening one number on your answer sheet for

each, using the response choices listed just below. Please try to respond to each item separately

in your mind from each other item. Choose your answers thoughtfully, and make your answers

as true FOR YOU as you can. Please answer every item. There are no "right" or "wrong"

answers, so choose the most accurate answer for YOU--not what you think "most people" would

say or do. Indicate what YOU usually do when YOU experience a stressful event.

1 = I usually don't do this at all

2 = I usually do this a little bit

3 = I usually do this a medium amount

4 = I usually do this a lot

------------------------------------------------------------------------

1. I discuss my feelings with someone. 1 2 3 4

2. I try to get emotional support from friends or relatives. 1 2 3 4

3. I get sympathy and understanding from someone. 1 2 3 4

4. I talk to someone about how I feel. 1 2 3 4


110

Appendix J

The Internet Motivation Scale The following statements describe some motivations for using the internet. For each statement, please indicate whether you agree with the statement, disagree with the statement, or are neutral about the statement. 1 = Completely Disagree 2 =Disagree 3 =Neutral 4 =Agree 5 =Completely Agree 1. The Internet is to me a substitute for other social contacts.

1 2 3 4 5

2. I receive real news through the Internet.

1 2 3 4 5

3. The Internet helps me in passing my time.

1 2 3 4 5

4. The Internet helps me coping with personal problems.

1 2 3 4 5

5. The Internet has a lot to offer: I can talk with friends and acquaintances.

1 2 3 4 5

6. I use the Internet to express myself.

1 2 3 4 5

7. The Internet offers more variation than other media do.

1 2 3 4 5

8. The Internet provides me with many things of interest that I can’t access anywhere else.

1 2 3 4 5

9. I use the Internet to form my own opinion.

1 2 3 4 5


111

10. I distract myself from school stress by using the Internet.

1 2 3 4 5

11. I use the Internet because of its current information.

1 2 3 4 5

12. The Internet promotes my way of life.

1 2 3 4 5

13. The Internet helps me to solve practical problems.

1 2 3 4 5

14. The Internet makes me feel like I am close to others.

1 2 3 4 5

15. I consider the Internet as an additional mass medium.

1 2 3 4 5

16. The Internet stimulates my curiosity.

1 2 3 4 5

17. I have found new friends and acquaintances through the Internet.

1 2 3 4 5

18. The Internet updates me on new trends.

1 2 3 4 5

19. Ever since I went on-line, I make less use of other media.

1 2 3 4 5

20. The Internet forces me to make choices between its many offers.

1 2 3 4 5


112

Appendix K

Negative Outcomes:

Below are some statements about Internet use. Please rate each statement based on how strongly you agree or disagree with it.

Definitely Disagree

Strongly Disagree

Somewhat Disagree

Slightly Disagree

Slightly Agree

Somewhat Agree

Strongly Agree

Definitely Agree

1. My internet use has made it difficult for me to manage my life.

1

2

3

4

5

6 7 8

2. I have missed social engagements or activities because of my Internet use.

1

2

3

4

5

6 7 8

3. My Internet use has created problems for me in my life.

1

2

3

4

5

6 7 8

4. I prefer online social interaction over face-to-face interaction.

1

2

3

4

5

6 7 8

5. Online social interaction is more comfortable for me than face-to-face interaction.

1

2

3

4

5

6 7 8

6. I have used the Internet to talk with others when I was feeling isolated.

1

2

3

4

5

6 7 8

7. I have used the Internet to make myself feel better when I was down.

1

2

3

4

5

6 7 8

8. I have used the Internet to make myself feel better when I’ve felt upset.

1

2

3

4

5

6 7 8

9. When I haven’t been online for some time, I become preoccupied with the thought of going online.

1

2

3

4

5

6 7 8


113

10. I would feel lost if I was unable to go online.

1

2

3

4

5 6 7 8

11. I think obsessively about going online when I am offline.

1

2

3

4

5

6 7 8

12. I prefer communicating with people online rather than face-to-face.

1

2

3

4

5

6 7 8

13. I have difficulty controlling the amount of time I spend online.

1

2

3

4

5

6 7 8

14. I find it difficult to control my Internet use.

1

2

3

4

5 6 7 8

15. When offline, I have a hard time trying to resist the urge to go online.

1

2

3

4

5

6 7 8


114

Appendix L Positive Outcomes:

The following are a number of statements that reflect various ways in which we view ourselves. Rate the degree to which you agree or disagree with each statement using the following scale (1 = Strongly Disagree and 6 = Strongly Agree). There are no right or wrong answers. Do not spend too much time with any one statement and do not leave any unanswered. 1 = Strongly Disagree 2 = Somewhat Disagree 3 =Slightly Disagree 4 =Slightly Agree 5 =Somewhat Agree 6 =Strongly Agree

1. I am comfortable in the presence of strangers when I’m online.

1 2 3 4 5 6

2. I am in tune with the online world.

1 2 3 4 5 6

3. Even among my friends online, there is no sense of brother/sisterhood.

1 2 3 4 5 6

4. Online, I fit in well in new situations.

1 2 3 4 5 6

5. I feel close to people online.

1 2 3 4 5 6

6. Online, I feel disconnected from the world around me.

1 2 3 4 5 6

7. Even on websites involving people I know, I don’t feel that I really belong.

1 2 3 4 5 6

8. I see online friends as friendly and approachable.

1 2 3 4 5 6

9. I feel like an outsider when I’m online.

1 2 3 4 5 6


115

10. I feel understood by the people I know when I’m online.

1 2 3 4 5 6

11. Online, I feel distant from people.

1 2 3 4 5 6

12. I am able to relate to my online friends.

1 2 3 4 5 6

13. I have little sense of togetherness with my online friends.

1 2 3 4 5 6

14. I find myself actively involved in online friend’s lives.

1 2 3 4 5 6

15. Online, I catch myself losing a sense of connectedness with society.

1 2 3 4 5 6

16. I am able to connect with other people online.

1 2 3 4 5 6

17. I see myself as a loner when I am online.

1 2 3 4 5 6

18. I don’t feel related to most people online.

1 2 3 4 5 6

19. My online friends feel like family.

1 2 3 4 5 6

20. Online, I don’t feel I participate with anyone or any group.

1 2 3 4 5 6


116

Appendix M

Social Desirability

Listed below are a number of statements concerning personal attitudes and traits. Read each item and select “T” if the statement is True for you, or select “F” if the statement is False for you.

True False

1. Before voting I thoroughly investigated the qualifications of all the candidates. T F

2. I never hesitate to go out of my way to help someone in trouble. T F

3. It is sometimes hard for me to go on with my work if I am not encouraged. T F

4. I have never intensely disliked anyone. T F

5. On occasion I have had doubts about my ability to succeed in life. T F

6. I sometimes feel resentful when I don’t get my way. T F

7. I am always careful about my manner of dress. T F

8. My table manners at home are as good as when I eat out in a restaurant. T F

9. If I could get into a movie without paying and be sure I was not seen, I probably

would do it. T F

10. On a few occasions, I have given up doing something because I thought too little

of my ability. T F

11. I like to gossip at times. T F

12. There have been times when I felt like rebelling against people in authority

even though I knew they were right. T F

13. No matter who I’m talk to, I’m always a good listener. T F

14. I can remember “playing sick” to get out of something. T F

15. There have been occasions when I took advantage of someone. T F

16. I’m always willing to admit it when I make a mistake. T F

17. I always try to practice what I preach. T F

18. I don’t find it particularly difficult to get along with loud-mouthed, obnoxious people. T F


117

19. I sometimes try to get even rather than forgive and forget. T F

20. When I don’t know something I don’t at all mind admitting it. T F

21. I am always courteous, even to people who are disagreeable. T F

22. At times I have really insisted on having things my own way. T F

23. There have been occasions when I felt like smashing things. T F

24. I would never think of letting someone else be punished for my wrongdoings. T F

25. I never resent being asked to return a favor. T F

26. I have never been irked when people expressed ideas very different from my own. T F

27. I never make a long trip without checking the safety of my car. T F

28. There have been times when I was quite jealous of the good fortunes of others. T F

29. I have almost never felt the urge to tell someone off. T F

30. I am sometimes irritated by people who ask favors of me. T F

31. I have never felt that I was punished without cause. T F

32. I sometimes think when people have a misfortune they only got what they deserved. T F

33. I have never deliberately said something that hurt someone’s feelings. T F

The Role of Personality Traits, Social Anxiety, Depressive ...

Documents