University of Southern California Dissertations and Theses

A network analysis of online and offline social influence processes in relation to adolescent smoking and alcohol use

(USC Thesis Other)

A network analysis of online and offline social influence processes in relation to adolescent smoking and alcohol use

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 1

A Network Analysis of Online and Offline Social Influence Processes in relation to
Adolescent Smoking and Alcohol Use

by
Grace C. Huang

A Dissertation Presented to the
ACULT CHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
F Y OF THE USC GRADUATE S
I n Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PREVENTIVE MEDICINE – HEALTH BEHAVIOR)

August, 2013

2
Table of Contents

LIST OF TABLES ................................................................................................................ ....................... 3
DEDICATION .................................................................................................................... ......................... 5
LIST OF FIGURES ............................................................................................................... ...................... 4
ACKNOWLEDGEMENTS .............................................................................................................. ........... 6

ABSTRACT ...................................................................................................................... ........................... 8

CHA .................................................................... 10 PTER 1 — INTRODUCTION .. ......
Specific Aims 10
... ........... ............
Background and Significance 13
Overview of Dissertation Studies 18
Parent Study 21
CHAPTER 2 — PEER INFLUENCES: THE IMPACT OF ONLINE AND OFFLINE
FRIENDSHIP NETWORKS ON ADOLESCENT SMOKING AND ALCOHOL USE ..................... 23
Abstract 23
n Introductio 25
Methods 28
Results 32
Discussion 34
CHA — A COM ARISON OF IN-PERSON AND DIGITAL SOCIAL
INFLUENCES ON ADOLESCENT SMOKING AND ALCOHOL USE ............................................. 44
PTER 3 P
Abstract 44
n Introductio 46
Methods 49
Results 53
Discussion 57
CHAPTER 4 — SMOKING, ALCOHOL USE AND SOCIAL NETWORKING SITES:
IMPACT ON ADOLESCENT FRIENDSHIP SELECTION AND INFLUENCE .............................. 71
Abstract 71
n Introductio 72
Methods 75
Results 79
Discussion 82
CHA ONS ....................................................................... 92 PTER 5 — DISCUSSION AND CONCLUSI
Summary of Overall Findings 92
Implications for Future Research 94
Implications for Future Practice 96
REFERENCES .......................................................................................................................................... 98
3

List of Tables

Table 1.1 Self Reported Sample Characteristics (%, n=1563) .......................................................... 39
Table 1.2 Social Network Site Activity (N=1,536) ................................................................................... 40
Table 1.3 Comparisons of Mutually Exclusive Social Media Use Groups ...................................... 41
Table 1.4 Associations between Online SNS Activity and Risk Behaviors (β) ............................ 42
Table 2.1 Study Sample Characteristics (N=1,707) ............................................................................... 62
63 Table 2.2 Smoking and Alcohol Use Measurement Model Indicators (T1) ..................................
Table 2.3 Pairwise Bivariate Correlations between Peer Influence and Smoking 0utcome
65 Variables ..............................................................................................................................................
Table 2.4 Pairwise Bivariate Correlations between Peer Influence and Drinking Outcome
Variables .............................................................................................................................................. 66
Table 3.1 Tested Effects of Network and Behavior Coevolution Models ...................................... 87
Table 3.2 Descriptive Sample Statistics ....................................................................................................... 88
Table 3.3 Meta Analysis Results (Base model) ......................................................................................... 89
able 3.4 Meta Analysis Results (SNS model) .......................................................................................... 90 T

4

List of Figures

Figure 1 Conceptual Model of Dissertation Studies ............................................................................... 20
Figure 1.1 Interaction Effect between Friend Drinking Status and Risky Online Exposure on
Alcohol Intake ................................................................................................................................... 43
7 Figure 2.1 Hypothesized Model of Digital and In-person Social Influences on Behavior ...... 6
Figure 2.2. Confirmatory Factor Analysis Models of Digital and In-person Social Influence
............. 68 Latent Variables for Adolescent Smoking and Alcohol Use .............................
Figure 2.3 Structural Equation Model (A) with Correlational Paths between Social
.... 69 Influences and Intention ..........................................................................................................
Figure 2.4 Structural Equation Model (B) for Social Influence Mechanisms with Causal
Paths between Social Influences and Intention. ................................................................ 70

5
Dedication

To my mother, Chung-Shiu Han, for always believing in me. Thank you for your
unwavering support, optimism, and prayers, which have encouraged me and spurred me
on in more ways you’ll ever know.

6
Acknowledgements

I would like to thank Thomas Valente, my mentor and committee chair, for his
guidance and incredible support through this entire process; for inspiring me, believing in
me and challenging me to follow my passion.
Many thanks to my committee members, Jennifer Unger, May Ann Pentz, Chih-Ping
Chou, and Maryalice Jordan-Marsh for their instruction, encouragement and insightful
feedback every step along this journey.
Thanks to Marny Barovich, for being there to offer an ear, provide advice and for
having a solution to every situation imaginable. Thanks to my colleagues Janet Okamoto
for paving the path for me; T. Em Arpawong and Claradina Soto for the camaraderie and
instrumental support; and Hee-sung Shin for all the great social network brainstorm
sessions.
I would like to thank my dear friends and family, aunts and uncles, who have rooted
me on, listened to the nitty-gritties of this process, celebrated with me, watched my kids,
and helped me in so many ways to reach this point in time, without which none of this
would be possible.
Special thanks to my Mom and Dad, who have been my role models and sources of
strength; Irene who has always understood and never doubted; and my two little rays of
sunshine, Karis and Elia, for adding so much color to my life, and for being so irresistibly
ute. c
A million thanks to my husband, Ritchie Huang, who has faithfully journeyed by my
side all along the way, always five steps ahead, to ensure that life has continued its course.
7
I am grateful for the countless weekends of his father-daughter grocery shopping, trips to
the park, the beach, and other places so that I could spend just a little more time at my
desk. His enduring patience, support, and encouragement are the reasons that I have
reached this day.
Thanks be to God for weaving all the pieces together and for knowing this final
utcome before it all got started. o

sitively associated with higher levels of smoking and alcohol use.
As an extension to the findings from the first paper, Study Two is based on a
structural approach to examine the specific pathways of influence using a cognitive-
behavioral framework to test whether social norms and behavioral intentions mediated the
effects between social influences (in-person and digital) on behavior. A comparison of the
relative correlations between in-person and digital risk exposures revealed that having
8
Abstract

Smoking and alcohol use are both prominent risk behaviors among adolescents in
the U.S. On average, about 20% of high school students smoke regularly and as many as
40% of adolescents in the 12th grade report using alcohol in the past 30 days. Peer
influences during adolescence are a strong predictor of risk-taking in part due to more time
spent with peers and the added value they place their relationships. Recent increases in
social media outlets have transformed traditional communication patterns and information
exchange, as well as the dimensions of social influence. Social media include mobile
phones and a multitude of online social networking sites, which are likely to accelerate the
transmission of information and adoption of new ideas or behaviors. This dissertation
study addresses whether adolescents' use of these new technologies affect their risk
behaviors and how these new modes of communication might play a role in shaping
normat . ive perceptions, intentions, and behavioral choices about smoking and alcohol use
Study One takes a multivariate approach to assess the impact of several measures
relating to adolescent use of social network sites (SNS). Results demonstrated that SNSs
differed in function and effects on adolescent risk behaviors. While no specific effects were
attributed to the use of any specific SNS, the collective display of risky content by friends
was po
9
friends who smoke or use alcohol and having friends who post content about these risk-
behaviors online were both positively associated with intentions and longitudinally
associated with behavior.
Study Three takes an in-depth longitudinal approach to examine whether
associations between social influences and risk behaviors were attributed to actual
influences due to behaviors exhibited by friends, or whether the association was attributed
to friendship selection. Stochastic actor-based (SAB) modeling was followed by meta-
analyses of the five school-level networks. Findings indicated that friendship selection
effects were stronger than peer influence effects for both smoking and alcohol use, while
influence effects were comparatively more prominent for alcohol use. In terms of social
media effects, adolescent friendships were formed based on similar use of SNSs Facebook
and Myspace and exposure to friends’ risky online content contributed to increases in
drinking and smoking.
In sum, the three studies, using three different statistical methodologies, may be
considered a “first look” at the interplay between online social networks, traditional in-
person peer friendship networks, and their collective effects on adolescent smoking and
alcohol use.

10
CHAPTER 1 — Introduction

Specific Aims
Smoking and alcohol use are both prominent risk behaviors among adolescents in
the U.S. (USDHHS, 2012; Windle, 2003). On average, about 20% of high school students
smoke regularly (Centers for Disease Control & Prevention, 2010) and as many as 40% of
adolescents in the 12
th
grade report using alcohol in the past 30 days (Johnston, O'Malley,
Bachman & Schulenberg, 2012). Peer influences during adolescence are a strong predictor
of risk-taking in part due to more time spent with peers and the added value they place
heir re t lationships.
Recent increases in social media outlets have transformed traditional
communication patterns and information exchange, as well as the dimensions of social
influence. Social media include mobile phones and a multitude of online social networking
sites (SNSs), which are likely to accelerate the transmission of information and adoption of
ew id n eas or behaviors.
This dissertation study will attempt to address whether adolescents’ use of these
new technologies affect their risk behaviors and how these new modes of communication
might play a role in shaping normative perceptions, intentions, and behavioral choices
about smoking and alcohol use. Social network analysis methods will be used to
investigate the social influence and potential social media influence mechanisms that
contribute to smoking and alcohol use among adolescents. The goal of this study is to shed
some light on social media influence mechanisms to inform the design and evaluation of
health promotion interventions.
11
Aim 1: To investigate peer offline and online friendships to determine how online
activities with friends might broker the peer influence processes by either encouraging or
hindering the influence of peer risk behaviors on adolescent health risk behaviors.
Analysis will be based on multilevel logistic regression models of dyadic egocentric
friendship networks.
H1. Adolescent SNS activity will be positively associated with risk behaviors.
H2. Higher levels of online activity will amplify the effects of friends’ risk behaviors on
adolescent risk behaviors.
Aim 2: To compare the pathways by which digital media and in-person influences occur
with regards to adolescent tobacco and alcohol use. Structural equation modeling will be
used within a cognitive-behavioral framework to compare the degrees to which digital
influences (social media risks) and traditional face-to-face influences (friend risk
behaviors) are associated with adolescent smoking and drinking intentions and actual
behaviors.
H1. Exposure to risky in-person influences (IPI) will be positively associated with
subjective norms that favor smoking/alcohol use and intentions towards current
risk behaviors.
H2. Exposure to risky digital influences (DI) will be positively associated with subjective
norms that favor smoking/alcohol use and intentions towards current risk
behaviors.
H3. Smoking/alcohol norms and behavioral intentions will mediate the relationship
between social influences and smoking/alcohol use outcomes.
12
Aim 3: To disentangle the longitudinal effects of social media use on dynamic friendship
network structures and on adolescent tobacco and alcohol use. Stochastic actor-oriented
social network models within the school setting will be used to assess whether social media
use affects the change in friendships based on similar risk behaviors (selection) or the
spread of risk behaviors through the exposure to friends who exhibit these risk behaviors
(influence).
smoking. H1. Selection effects will be stronger than influence effects on adolescent
n adolescent alcohol use. H2. Selection effects will be stronger than influence effects o
Similarity in SNS use will predict friend H3. ship selection.
H4. Exposure to friends’ risky online content will exhibit peer influence effects on
adolescent smoking and alcohol use.

13
Background and Significance
Tobacco and Alcohol Use among Adolescents
Smoking and alcohol use among adolescents are risk behaviors that remain
prominent health problems in the U.S. (USDHHS 2012; Windle, 2003). Despite falling rates
in adolescent smoking over the past decade, 19.5% of them still smoke cigarettes and
almost half (46.5%) report having ever tried smoking (U.S. Census Bureau, 2010).
Furthermore, smoking rates have been shown to increase dramatically during the high
school years (Backinger & Augustson, 2011). Behaviors are adopted during these years
and are often carried over into adulthood, making up 90% of all adult smokers (USDHHS,
2012). Alcohol use among adolescents have declined steadily over the last two decades,
but still remains the most widely used drug by today’s teenagers (Johnston et al., 2012).
National results from the Monitoring the Future (MTF) report indicate that as many as 70%
of students have consumed alcohol by the end of high school and half (51%) report having
been drunk at least once in their life, with boys consistently reporting higher alcohol use
rates than girls (Johnston et al., 2012). In terms of ethnic differences, MTF reports indicate
that Whites report the highest rates of heavy drinking, followed by Hispanics, while African
American youth report substantially lower rates of both alcohol and cigarette smoking
(Johnston et al., 2012). These variations were magnified among students of lower
socioeconomic status (Bachman, O'Malley, Johnston, Schulenberg & Wallace, 2011).
Social Influences and Risk Behaviors
As children enter into adolescence, influences from the family and home setting
gradually decrease as peers gradually become the main source of influence (Giordano,
2003). Social influence processes play a significant role in the adoption of risk behaviors
14
during the adolescent developmental period. Adolescents are especially susceptible to
normative influences as they seek to establish close friendships, explore their identities,
and define their peer group affiliations (Simons-Morton & Farhat, 2010; Steinberg &
Monahan, 2007; Urberg, Degirmencioglu & Pilgrim, 1997). Prior studies have shown that
adolescents’ use of tobacco to be highly associated with their friends’ use (Alexander,
Piazza, Mekos & Valente, 2001; Ennett et al., 2008; Hoffman, Monge, Chou & Valente, 2007;
Kobus, 2003; Sussman & Sun, 2009). This was also true for adolescent alcohol use (Cruz,
Emery & Turkheimer, 2012; Nash, McQueen & Bray, 2005; Trucco, Colder & Wieczorek,
2011; Wills & Cleary, 1999). Parental and sibling use of tobacco and alcohol have also been
shown to be predictive of adolescent behavioral outcomes (Avenevoli & Merikangas, 2003;
Hill, Hawkins, Catalano, Abbott & Guo, 2005; Urberg, Goldstein & Toro, 2005).
Social network analysis methods have greatly contributed to the understanding of
peer influence processes by accounting for social contexts and inherent dependencies in
one’s network structure (Ennett & Bauman, 1994). Furthermore, network analysis enables
the measurement of peers’ self-reported behaviors to control for the tendency of
adolescents to overestimate the prevalence of peer risky behaviors (Reid, Manske &
Leatherdale, 2008; Sussman et al., 1988; Unger & Rohrbach, 2002). School-based social
network studies have shown that an adolescent’s popularity or social status, determined by
the number of peer nominations, increases their likelihood to smoke (Valente, Unger &
Johnson, 2005). Popular students initiate smoking or other risk behaviors as a means to
draw attention and maintain their position in the spotlight. Others who desire the popular
status then perpetuate the norms that favor these risk behaviors. Network level effects
have also been shown to significantly impact youth risk behavior adoption such that
15
adolescents who are more embedded in their networks, or those with a higher proportion
of friends who are also friends with each other, have lower odds of smoking (Ennett et al.,
2006; Ennett et al., 2008). On the other hand, isolates, or those who received little or no
friendship nominations, have been shown to exhibit earlier smoking initiation and higher
rates of problem behaviors (Ennett & Bauman, 1993; Ennett et al., 2008; Pearson, Sweeting,
et al., 2006; Valente et al., 2005). One study found that friendship reciprocity, a measure of
tie strength, was predictive of adolescent smoking compared with other indicators of
friendship quality such as the “feeling of closeness” to their nominated friend and “visits to
friends’ homes” (Ennett et al., 2006).
Threshold models of behavior adoption suggest that individuals engage in certain
behaviors based on the proportion of people in one’s network who are already engaged in
the behavior (Granovetter, 1983). Individuals with low smoking thresholds adopt the
behavior before many others, while individuals with high thresholds will do so only after
the majority of the group has engaged in the behavior (Valente, 1996). Network exposure,
the percentage of people in one’s network who exhibit a certain behavior, provides an
objective measure of social influence (Valente, 2010). For instance, in an egocentric
network, the average behavior rate is calculated to determine the degree of exposure based
on the adolescent’s reports of his/her friends’ smoking status. In a sociometric network
where all respondents within a boundary are surveyed, exposure can be modeled by the
attributes of friends who contribute to incoming our outgoing ties (Hall & Valente, 2007).
Adolescent Social Media Use
Social media sites (SNS) such as Facebook and Myspace have gained immense
popularity among adolescents within the last few years, and have redefined their social
16
network boundaries and spheres of influence. In the U.S., 95% of youth between 12-17
years old access the Internet on a daily basis, and of these, 80% use social network sites
compared to 64% of adults (Lenhart et al., 2011). Mobile phones have gained just as much
traction among teens, where 77% now own a cell phone and report an average of 60 texts
each day (Lenhart, 2012). Furthermore, cell phones are pervasive across socioeconomic
status and have a comparatively wide reach among low income, minority, traditionally
hard-to-reach adolescents (Lenhart, Purcell, Smith & Zickuhr, 2010). Among teens, 63%
indicated that mobile phone texting was their preferred and dominant mode of
communication, superseding traditional phone calling (39%) and face-to-face exchanges
(35%) as the preferred and dominant mode of communication (Lenhart, 2012). Almost five
times as many teens also report choosing to communicate through SNS (29%) instead of
email (6%) for their daily conversations (Lenhart, 2012).
Adolescents benefit primarily from the socialization and communication
opportunities served through social media sites (boyd, 2007; Itō, 2010). Besides serving
the primary function of staying in touch with friends and family, making new friends,
sharing pictures and exchanging ideas, other benefits include providing opportunities for
community engagement, creative expression, formulation of ideas, and information
exchange, and connecting with others with diverse backgrounds and global perspectives
(O'Keeffe & Clarke-Pearson, 2011). Social media platforms are now used in educational
settings for students to collaborate on homework or group projects (boyd, 2008; Selwyn,
2009) and offer unique ways for adolescents to access health information. Investigators
have also found support for the use of mobile texting to increase medical adherence and
17
communication with their health care providers (Krishna, Boren & Balas, 2009; Mitchell &
Ybarra, 2009).
Despite these benefits, much attention has also been devoted to uncovering the risks
associated with social media. Concerns that health professionals have about adolescent
social media use relate mostly to privacy of online disclosures and the access to
inappropriate content that can be easily distributed to other peers or even more broadly to
the general public (Hinduja & Patchin, 2008; Ybarra, Mitchell, Finkelhor & Wolak, 2007).
While the Pew Research findings show that as many as 62% of teens who use a SNS
indicate that they set their profile to be private so that only their friends can see the
content they post, the report also notes that teens with public profiles were at greater risk
of experiencing a negative outcome from a SNS (Lenhart et al., 2011). Social network site
activity also increases the likelihood of exposure to online solicitations (Ybarra & Mitchell,
2008) or potentially harmful advertisements, such as those strategically placed by the
tobacco industry to target adolescents who may be vulnerable (Freeman, 2012). Current
research has focused on adolescent’s indiscriminate display of risky content such as sexual
references (Moreno, Brockman, Rogers & Christakis, 2010) and alcohol use (Moreno,
Briner, et al., 2010). Regardless of whether the displays truly reflect an adolescent’s
behavior, others are likely to perceive them as real and thus promote perceived norms that
increase the spread and adoption of these beliefs and behaviors (Moreno, 2011). There are
psychosocial consequences that are also associated with social network use such as
cyberbullying, which has been linked to depression, anxiety, severe isolation, and suicide
(Patchin & Hinduja, 2010). The intensity of online interactions has also been linked to
depression and the feeling of loneliness (Moreno, 2011), which may further lead
18
adolescents to Websites that promote substance abuse, unsafe sex, or self-destructive
behaviors (O'Keeffe & Clarke-Pearson, 2011).
There are several proposed ways that social media use may facilitate the peer
influence processes. First, adolescents who regularly access social media are exposed to
more content and information, some of which will include inappropriate photographs of
friends, alcohol consumption, smoking or use of other substances (Hinduja & Patchin,
2008; Livingstone & Brake, 2010). Exposure to alcohol references on Facebook and
Myspace have indeed been shown to increase one’s liklihood of problem drinking or an
alcohol related injury (Moreno, Christakis, Egan, Brockman & Becker, 2012). Second, given
the accessibility of social networking sites (SNSs), now increasingly over the mobile phone,
they effectively facilitate the rapid dissemination of information and in effect accelerate the
adoption of new attitudes and behaviors. Third, SNSs also enable the formation of large
networks, which often include upwards of hundreds of ‘friends’ (Backstrom, 2011). This
increases the likelihood of being exposed to biased perceived norms favoring risky
behaviors that are driven by the actions or postings of ‘popular’ individuals who have an
established presence on the sites.
Overview of Dissertation Studies
Despite the pervasive role of social media in adolescents’ lives today and the broad
research interest in the effects of these technologies on their well being, there is little
knowledge on how the use of these media channels affect peer relationships and peer
socialization processes. Given that adolescent risk behaviors are strongly attributed to
friendships that are formed and maintained during this time period, it is crucial to further
understand how these peer processes are affected by the exchanges that occur through
19
these new channels. These channels are now able to extend beyond traditional interactions
that were once limited by geographical proximity. The proposed studies are an attempt to
elucidate the role of social media use in adolescent relationships and the contribution to
peer influence processes with regards to smoking and alcohol use. Figure 1 is a conceptual
iagram d that integrates the three proposed studies.
The overall model illustrates the relationship between the main variables that are
examined in the three studies. “In-person context” represents the degree to which an
adolescent’s friends exhibit smoking and alcohol use intentions or behaviors. “Social-
media context” represents the adolescent’s interaction with (use frequencies of specific
online SNSs) and exposures to risky content from their friends’ online displays. “Subjective
norms” here is measured by students’ perceptions of their friend’s acceptance and
approval of the risk behaviors. “Risk behaviors” are the outcomes of interest that are
investigated in this dissertation, which represent the adolescent’s self-reported smoking
and alcohol use behaviors.

F

igure 1 Conceptual Model of Dissertation Studies

In Study One, I first take a descriptive look at the degree of overlap between online
and offline friendships. In the subsequent portion, I take a look at the contribution of
online friendships and exposure to risky online content to risk behavior adoption above
and beyond previously established predictors such as peer smoking, peer alcohol use,
parent behaviors, gender, and other demographic variables. Interaction effects between
online friendships and peer risk behaviors are tested to determine whether online
friendships moderate the association between peer risk behavior and adolescent risk
20
behavior.
In Study Two, I take a step further to compare social influence and potential social
media influence pathways on adolescent smoking and alcohol use behaviors based on the
general framework posited by the Theory of Reasoned Action (Fishbein & Ajzen, 1975;
Montano & Kasprzyk, 2008). Lastly, Study Three is a longitudinal analysis of potential

21
diversity, consisting of 69%
social media effects on network co-evolution (Snijders, Van de Bunt & Steglich, 2010).
Specifically, I examine whether the association between peer risk behaviors and adolescent
risk behaviors are due to the selection of friends based on similarity in social media use
and risky behaviors, or whether behaviors are adopted due to peer influences and
characteristics of their social media use.
Parent Study
The dissertation study is based on data collected from the Social Network Study, a
longitudinal study of high school adolescents, which was designed to answer
methodological questions about different data collection practices in social network studies
as well as theoretical questions about the effects of different types of peer relationships on
individual network characteristics and on smoking and alcohol use influences. Network
data were collected based on varying boundary specifications (community, grade, class)
and different types of peer networks (friendship, most admired, likely to succeed, romantic,
most popular) (Valente, Fujimoto, Unger, Soto & Meeker, 2013).
The Social Network Study took place at five comprehensive high schools in the El
Monte Union High School District (EMUHSD)
1
, about 12 miles east of the Los Angeles
metropolitan area. The study and data collection process was fully supported by the school
district superintendent, as well as the principals and teachers from each school due to the
strong collaborative relationship garnered from previous studies.
El Monte, the ninth largest city of Los Angeles County (out of 88), has a population of
approximately 113,000 (U.S. Census Bureau, 2010). The city is characterized by ethnic
Hispanic, a recent increase in the Asian population to 25%,

1
None of these high schools are considered charter or magnet schools.
22
about 5% White, and .4% Black or African American. The average household size is 4.04,
and median age is 31.6%. El Monte land use is about 58% residential, 11% retail, 10%
industrial and is known to attract commercial and retail businesses, as well as international
corporations through its Foreign Trade Zone and a pro-active Chamber of Commerce. The
EMUHSD was established in 1901, and has since expanded from one high school to five
over the last six decades (City of El Monte 2012). It now serves 200,000 people
representing mostly industrial, factory and retail workers and families in the middle or
ower income economic groups (EMUHSD 2012). l

aphically distinct user characteristics and had differential effects on risk behaviors.
Exposure to risky online content had a direct impact on adolescents’ risk behaviors
and significantly interacted with risk behaviors of their friends. These results provide
evidence that friends’ online behaviors should be considered a viable source of peer
influence. Implications for future studies and interventions are discussed.
23
CHAPTER 2 — Peer Influences: The Impact of Online and Offline
Friendship Networks on Adolescent Smoking and Alcohol Use

Abstract
Online social networking sites (SNSs) have become a dominant mode of
communication between adolescents. However, little is known about the effects of online
activity with friends on health behaviors. The use of SNSs between friends is examined, as
well as the degree to which SNS activities relate to face-to-face peer influences and
adolescent risk behaviors.
Longitudinal egocentric friendship network data as well as adolescent social media
use and risk behaviors were collected from 1,563 tenth grade students across five Southern
California high schools. Fixed effects models were used to assess the effects of SNS use, SNS
friendships, and risky online exposures on adolescent smoking and alcohol use.
The frequency of SNS use and number of SNS friends were not significantly associated with
adolescent risk behaviors. However, exposure to friends’ online pictures of partying or
drinking was significantly associated with both smoking (β=.09, p<.001) and alcohol use
(β=.06, p<.05). While adolescents with drinking friends had higher risk levels for drinking,
adolescents without drinking friends were more likely to be affected by increasing
exposure to risky online pictures (β=-.13, p<.05). Myspace and Facebook had
demogr
24
This study provides further evidence that adolescents who are exposed to friends’
risky online displays are more likely to smoke and use alcohol. The effects are magnified
for adolescents without face-to-face drinking friends. Continued research to examine
online peer influence mechanisms are needed to effectively educate adolescents about
these risks.

25
Introduction
Smoking and alcohol use among adolescents are still prominent risk behaviors in
the U.S. (USDHHS, 2012). Despite falling rates in adolescent smoking over the past decade,
15.8% report smoking cigarettes in the past month and almost half (46.3%) have ever tried
smoking (CDC, 2012). Over 80% of adult smokers begin smoking during adolescence
(USDHHS, 2012). Alcohol use has declined steadily over the last two decades, but still
remains the drug most widely used by today’s adolescents (Johnston et al., 2012). The
national Monitoring the Future (MTF) study indicates that 70% of students have consumed
alcohol and half (51%) have been drunk at least once in their life by end of high school
(Johnston et al., 2012).
Adolescent friendships and Risk Behaviors
Peer influences play a significant role during adolescence, a time when new
identities, friendships, and peer group affiliations are solidified and parental influences
gradually diminish (Kobus, 2003; Simons-Morton & Farhat, 2010). Peers have a profound
effect on each other and may encourage experimentation of risky behaviors when there is
normative pressure to do so (Valente et al., 2005). There is also substantial evidence that
adolescents’ use of tobacco and alcohol are highly associated with their friends’ use (Cruz
et al., 2012; Hoffman et al., 2007; Trucco et al., 2011; Valente, Fujimoto, Soto, Ritt-Olson &
Unger, 2013).
Adolescents and Social Media Use
Recent increases in social media outlets have transformed traditional
communication and information exchange mechanisms, as well as the dimensions of social
influence. Online social networking sites (SNSs) such as Facebook, Twitter and Myspace
26
have gained immense popularity among adolescents within the last few years and have
redefined their network boundaries and spheres of influence. In the U.S., 95% of youth
between 12-17 years old access the Internet on a daily basis and of these, 80% use social
network sites (Lenhart et al., 2011). Almost five times as many adolescents use SNSs (29%)
instead of email (6%) for daily communication (Lenhart, 2012).
With increased accessibility through mobile devices, SNSs provide a mechanism for
adolescents to connect with friends instantaneously (Itō, 2010). Studies indicate that
adolescents benefit from the socialization opportunities such as staying in touch, sharing
pictures, exchanging ideas (boyd, 2007; Itō, 2010). SNSs have also been used to foster
community engagement, creative expression, and diversity (O'Keeffe & Clarke-Pearson,
2011).
Recent attention, however, has been directed toward uncovering the risks
associated with SNS use including adolescents’ creation and display of inappropriate
content such as sexual references and substance use (Livingstone, 2008; Moreno,
Brockman, et al., 2010; Patchin & Hinduja, 2010). Exposure to risky content posted by
friends can cultivate unfavorable norms that are then rapidly spread through the online
networks and contribute to the adoption of risky beliefs and behaviors (Moreno, Egan &
Brockman, 2011). Other risks include higher exposures to sexual solicitations, bullying (M.
Ybarra, Mitchell & Espelage, 2012), tobacco advertisements (Freeman, 2012), and
psychosocial consequences such as depression, anxiety, and loneliness (Moreno et al.,
2011; Ybarra, Alexander & Mitchell, 2005), which may pave the way toward higher
likelihoods of substance abuse, unsafe sex, or other self-destructive behaviors (O'Keeffe &
Clarke-Pearson, 2011).
27
Social Media Contexts
The current prevalence of adolescent engagement in SNSs suggests that their online
networks reflect their offline ones in that most online connections extend from existing
face-to-face relationships (Mesch & Talmud, 2007; Reich, Subrahmanyam & Espinoza,
2012). One study of adolescents found a moderate overlap between their closest online
and offline friends and suggested that adolescent online contexts are generally used to
strengthen existing relationships (Reich et al., 2012). Evidence also suggests that these
sites are distinct in demographic distribution. In an ethnographic survey of both Myspace
and Facebook users, boyd (2011) described how race and class influenced adolescents’
choice of SNS. Myspace was described as a place for creative expression, a portal for
discovering new musical artists and tastes. Users were also more likely to be younger,
Hispanic, and from lower socio-economic status backgrounds (Ahn, 2012; Hargittai, 2007).
In contrast, the clean, predictable, and functional format of Facebook appealed to others.
Adolescents who aspired to connect with friends in college viewed Facebook as a marker of
status and maturity. Migrating from Myspace to Facebook was a “growing up” process as
“adult” relationships through Facebook superseded the need for more introspective
features on Myspace (Robards, 2012).
Social Media Use and Health among Adolescents
Little is known about the effects of social media use on adolescent health behaviors.
One study of 400 adolescent Myspace profiles found that 56% of them contained alcohol
references and among these, 49% talked about explicit alcohol use (Moreno, Briner, et al.,
2010). Studies of homeless youth indicate that online friendships were associated with
both risky behaviors such as increased exchange sex and protective behaviors such as
28
HIV/STI testing—depending on the type of relationships that were fostered and topics
discussed through these networks (Rice, Monro, Barman-Adhikari & Young, 2010; Young &
Rice, 2011). Further understanding about the nature of online ‘friendships’ is necessary to
mitigate these harmful effects on adolescents.
Online communication portals have the ability to simultaneously transmit new
attitudes and behaviors to countless people beyond geographic boundaries (Bandura,
2002). Facebook and Myspace are two examples of online portals used by teens that are
the focus of this study. Content displayed by peers can be a powerful source of influence
through peer modeling which are likely to promote biased normative perceptions,
especially for adolescents who have many friends on SNSs and for those who frequently
visit them. The goal of this study is to investigate peer offline and online friendships to
determine how online activities with friends might broker the peer influence processes by
either encouraging or hindering the influence of peer risk behaviors on adolescent health
risk behaviors. The questions examined are: (RQ1) whether there are positive associations
between adolescent SNS activity and risk behaviors, and (RQ2) whether higher levels of
online activity might amplify the effects of friends’ risk behaviors on adolescent risk
behaviors.
Methods
Data were drawn from the Social Network Study, a longitudinal study of high school
adolescents designed to answer methodological and theoretical questions about data
collection practices and effects of different peer relationships on risk outcomes (Valente,
2010). The sample consisted of 10
th
grade students at five comprehensive high schools in
29
the El Monte Union High School District
2
. At the time this study, El Monte was the ninth
largest city of Los Angeles County with a population of approximately 113,500 and an
ethnic distribution of 69.0% Hispanic, 24.9% Asian, 4.9% White, and .4% Black/African
American (U.S. Census Bureau, 2010).
Study Design and Data Collection
The first two waves of the Social Network Study were collected in October 2010 and
April 2011. Paper and pencil surveys were administered during class on a regular school
day. Of the total 2,290 enrolled 10
th
graders, 2,016 returned valid parental consent forms
(88.0%) with 1,823 agreeing to participate in the study. Some 28 of these students did not
provide student assent, reducing the eligible pool to 1,795. A total of 1,719 students
completed surveys at the first wave (T1) of data collection, 1,620 students completed the
survey at the second wave (T2) and 1,563 students completed the surveys at both time
points.
Measures
Egocentric Network Characteristics. Each student (ego) was asked “name seven best
friends regardless of where they live or go to school” and basic demographic and risk
behavior information for each of their friends (alters). These nominations were then used
to construct egocentric (personal) networks for each individual. Friends’ risk behaviors
were assessed by asking students to respond whether their friends “ever smoked a
cigarette” and “drink alcohol at least once a month” (1=yes, 2=no).
3
Students were asked if

2
charter or These five high schools comprised the entire school district. None of these schools are considered
magnet schools.
3
Lifetime smoking and past-month alcohol use were selected because these indicators were more
comparable in their prevalence rates. Furthermore, use of a past-month smoking indicator would not
provide sufficient power for analyses conducted for this paper.
30
they used Facebook and Myspace and if so, then asked whether their alters were also their
friends through these SNSs (1=yes, 2=no). Online risky behaviors of friends were assessed
by asking whether alters ever “posted pictures of themselves partying or drinking alcohol
online” and “talk about partying online.” The measures at T1 were used to characterize
students’ egocentric networks and used as predictive indicators in the analyses for this
study.
Social Media Use. In addition to the two social media indicators described, students
were also asked to indicate how frequently they visited the social network sites Facebook
and Myspace in the past month as an indicator for students’ SNS use. Response options
were coded as 1=never, 2=rarely (about once a month or less), 3=occasionally (about once a
week or less), 4=frequently (about once every 2-3 days) and 5=very frequently (about once a
day or more).
Tobacco and Alcohol Use. Smoking and alcohol use were assessed by students’ self-
reports at both time points. Data from T2 were used as the outcome indicators for this
study. Smoking was coded into a composite score composed of responses to five smoking
frequency and intention questions. The items that were used to create the 5-point score
included: 12-month smoking intention (coded as 1=non-susceptible; 2=susceptible), age
smoked first whole cigarette (3=ever smoked), number of days smoked cigarettes during
the past 30 days (4=past month smoker) and daily smoking for 30 days (5=daily smoker).
Due to a skewed distribution of 70.2% never-smokers, the smoking variable was also
dichotomized by recoding students who reported smoking at any intensity (smoking score
of 3 or higher) as “ever-smokers.” The 5-point smoking status and ever-smoke indicators
were both tested and compared.

31
Similarly, alcohol use was also coded into a 5-point composite score and a
dichotomized ever-drink indicator. The items used to create the composite score included:
12-month drinking intention (1=non-susceptible; 2=susceptible), age at first drink of alcohol
except for religion purposes (3=ever drinker), number of days having at least one drink of
alcohol during the past 30 days (4=past month drinker) and number of days having five or
more drinks of alcohol in a row during the past 30 days (5=past month binge drinker).
Data Analysis
Descriptive analyses of students’ demographic characteristics for both T1 and T2
were conducted. Those lost to follow-up had significantly lower academic achievement,
lower SES, and higher levels of risk behaviors than those who were surveyed at both time
points. Myspace and Facebook users were compared to determine whether these two
indicators should be considered separately or collapsed as one SNS use variable. Intraclass
correlations (ICC) for smoking and alcohol use outcomes were calculated to determine
whether there were significant within-school similarities of students’ smoking and alcohol
use levels. Linear regression models with school-level fixed effects were fitted to test the
effects of online activity with friends on smoking and alcohol use outcomes at T2 (RQ1)
while controlling for these behaviors and other factors at T1
4
.
Indicators for online activity with friends were created using the sum of alters who
were ego’s friends on the social network sites (Facebook and Myspace) and the sum of
alters who displayed risky online behaviors. Multiple imputation using chained equations
(StataCorp, 2011) was performed for missing data on four control variables (health status,

4
School 4 was selected as the reference school based on descriptive analyses comparing the outcome
indicators of all 5 schools across both data time points. School 4 had consistently lowest scores in all risk
categories.
32
SES, grades, and parent smoking/drinking status), approximately 10% for each variable.
Alter smoking and drinking indicators were dichotomized (0=no friends smoke/drink; 1=
one or more friends smoke/drink) and then used to construct the two-way interaction
terms.
For RQ2, the interaction terms were added to the model to determine whether SNS
activity with friends either amplified or tempered the association between alters’ risk
behaviors and the ego’s risk behaviors. The above analyses were repeated using “ever-
smoke” and “ever-drink” as outcome variables to test for any differences in associations.
All analyses were conducted in STATA 12.0 (StataCorp, 2011).
Results
Table 1.1 presents the self-reported demographic characteristics of our sample
across the two waves of data. Students who responded to the survey were evenly
distributed across gender and on average 15 years old. About two-thirds were
Hispanic/Latino and nearly one-fourth were Asian, closely reflecting the ethnic distribution
of El Monte City. Half of the student sample (50%) reported speaking English and another
language equally at home and about one-third of students (33%) reported speaking more
English than another language. Socioeconomic status was represented by the ratio of
rooms to the number of people who live at home. Students reported having on average 3.2
rooms and 3.4 people living at home, indicating slight overcrowding (Myers, Baer & Choi,
1996). About 86% of students were eligible for free or reduced-price lunches at school.
The majority of students (87%) reported being in good health. Most students (96-97%)
owned or borrowed a mobile phone and 50.4% used them to access the Internet. Students’
risk behaviors at T2 indicate that 29.9% were ever-smokers (even one or two puffs) and
33
56.8% had at least one drink of alcohol (other than for religious purposes). Roughly 40%
of students reported having at least one friend who smoked and/or drank alcohol.
In terms of social media activity (Table 1.2), students reported visiting Facebook
and Myspace about once a week. At T2, students’ Facebook use increased to about two to
three times a week while their Myspace use decreased to about once a month or less. At
T1, Myspace appeared to be the most popular venue for online friendships (on average 2.7
out of 5.5 nominated alters were Myspace friends), followed by Facebook (1.83 alters). On
average, 35% of students had at least one friend who talked about partying online and 21%
reported that their friends posted party/drinking pictures online.
The comparison between Facebook and Myspace user types revealed striking
differences in academic achievement, ethnicity, SES, risk behaviors, especially between
Facebook only and Myspace only users (Table 1.3). Facebook only users had higher grades
(66% vs. 25% As and Bs), spoke more English with friends (57% vs. 44%), but were less
likely to be Hispanic (24% vs. 88%), have reduced price school lunches (75% vs 92%), and
). were less likely to have ever smoked (9% vs. 43%) or drink alcohol (36% vs. 71%
Given these differences, Facebook and Myspace were assessed as separate
predictors in the following analyses. Table 1.4 displays the main effects and interaction
effects models for both smoking and alcohol use outcomes. The effects of SNS activity
(RQ1) were mixed. Having friends on either Facebook or MySpace was not significantly
associated with students’ reports of risk behaviors. However, if friends posted pictures of
themselves partying or drinking alcohol online, students were significantly more likely to
report that they smoke (β=.09, p<.001) and use alcohol (β=.05, p<.05). While Facebook use
34
did not exhibit significant effects on either risk behavior, higher levels of Myspace use was
associated with alcohol use (β=.06, p<.05).
As can be expected, students’ risk behaviors at T2 were the strongest indicators for
their behaviors at T2. Similarly, friend and parent influences were also significant for both
adolescent smoking (β =.08, p<.001; β =.05, p<.05 respectively) and drinking (β =.08,
p<.001; β =.07, p<.001 respectively). The number of friends nominated was negatively
associated with smoking (β =-.08, p<.001) but not alcohol use.
One significant negative interaction effect was found between “having friends who
post risky pictures of themselves online” and “friends’ smoking behaviors” (β=-.13, p<.05).
As depicted in Figure 1.1, the interaction effect indicates that the degree of association
between friends’ risky online behaviors and adolescents’ risky behaviors was moderated
by whether or not the adolescent’s close friend(s) drink alcohol. While adolescents with
drinking friends were at an elevated risk for drinking, exposure to risky online pictures
ol. appeared to pose a higher risk for adolescents whose close friend(s) do not drink alcoh
The dichotomized smoking models were not significantly different from the described
models using the 5-point outcomes. However, in the dichotomized alcohol use model,
online activities (i.e. posting pictures of partying and drinking, Myspace use), and the
interaction effect became insignificant. This suggests that there may be differential online
effects that apply to adolescents at varying levels of alcohol use.
Discussion
To our knowledge, this is the first study to apply social network analysis methods to
examine the influences of adolescent SNS activities on their smoking and alcohol use.
Social network analysis contributes to the understanding of peer influence processes by
35
accounting for the individual’s social contexts and the perceived norms within those
contexts. This study of egocentric networks used the characteristics of adolescents’
nominated friends and their shared online activities to help elucidate potential online
influence mechanisms.
Consistent with earlier research, friend and adolescent risk behaviors were strongly
associated (Simons-Morton & Farhat, 2010; Trucco et al., 2011). The study findings further
demonstrate that exposure to friends’ risky displays online significantly contributed to
adolescent smoking and drinking, while the frequency of SNS use and the number of online
friendships alone did not. Myspace use was also associated with higher levels of drinking.
These results suggest that friends’ online risky displays may be a viable source of peer
influence.
Only alcohol use was significantly associated with Myspace use. The significant
interaction effect between friends’ alcohol use and exposure to risky online portrayals of
partying and drinking suggests that this risk is magnified in the absence of face-to-face
drinking friends. Significance of these alcohol-related findings could be due to higher
prevalence of alcohol consumption requiring less power to demonstrate statistical
significance, or due to the social nature of drinking compared to smoking (15.5% vs. 6% in
our sample). Drinking behaviors may also be more easily modeled and learned than
smoking, and thus more readily transmitted through non-face-to-face contexts. Since
Internet access is almost ubiquitous for adolescents, online influences can occur at any time
of day, in any setting, in the company of others or in isolation. This further underscores the
importance for more research on the mechanisms of peer influence through SNSs.
36
In accordance with studies described earlier (boyd, 2011; Hargittai, 2007), Myspace
and Facebook users were markedly different and differentially associated with risk
outcomes. The significant associations between Myspace and risk behavior could have
been attributed to influences from its eclectic user-base, or that at-risk adolescents are
naturally drawn to SNSs that can be tailored to suit their preferences. Perhaps the
expectation that Facebook was related to “growing up” and a college audience (Robards,
2012), students may perceive risky online behaviors as less favorable. In either case, our
findings suggest that exposures to risky online displays are likely to contribute to biased
normative perceptions of risk behaviors.
The number of nominated friends was protective against risk behaviors. This might
appear to contradict previous studies that show an association between popularity and risk
(Valente et al., 2005). However, “close friendships” in this study were measured by the
number of ego’s outgoing nominations, which is substantively different from measures of
popularity, typically represented by the number of nominations received by ego (Fujimoto
& Valente, 2012).
Limitations
There are several limitations specific to this study. First, our findings are based on
adolescents’ reports of their friends’ risk and online behaviors. While these reports may be
prone to biases, studies have shown that one’s perceptions often provide more reliable
indicators for health outcomes than the reality (Champion & Skinner, 2008; Hoffman et al.,
2007; Valente, Fujimoto, Soto, et al.). Secondly, since our data were from 10
th
graders of
one Los Angeles County school district, our results may not be generalizable to the larger
adolescent population or to adolescents who were not surveyed or lost to follow-up.
37
Thirdly, our study focused on online friendships between existing close friends and could
not have fully captured all dimensions of these online friendships. The types of online
activity with friends may also vary in their effects on psychosocial or behavioral outcomes.
Lastly, while the overall effects of online influences were small, they are likely to increase
over time as SNSs become even more closely integrated with adolescents’ day-to-day
interactions. Finally, as a secondary data analysis study, interviews with adolescents or
parent figures were not possible. Such interviews in future studies would provide a
powerful context for learning about the reliability of reported behaviors and mechanisms
by which SNSs influence behavior.
Implications
Further research might examine how friendships may differ across online and
offline contexts and monitor specific types of online activities and discussion topics
between friends. Mediators such as perceived norms, attitudes, self-efficacy, or friendship
closeness should also be examined to inform a more robust model of online social influence
mechanisms.
Future health education interventions might consider incorporating modules to
teach adolescents about the harmful effects of posting risky behaviors online (Moreno,
2011) and how these displays can negatively affect their own friends. Strategies may
involve fostering norms that discourage or de-glamorize the posting of risky pictures since
others are likely to perceive them at face value whether or not they reflect one’s true
behavior. Online impressions (Walther, Van Der Heide, Kim, Westerman & Tong, 2008)
may bias perceived norms about risk behaviors by minimizing the appearance of negative
consequences and simultaneously increasing the spread of these risky beliefs. Teachers,
38
physicians, or peers may effectively relay messages to adolescents about the harmful
effects of risky online content or encourage students to leverage their close online
friendships to create ‘healthy’ online content to bolster favorable norms through SNSs.
The change in SNS use trends over the course of this study serves as a reminder that
technology advances occur rapidly, and that interventions must be adapted accordingly to
retain their appeal to adolescents. When utilizing SNSs for health promotion, public health
professionals should invest time in understanding the culture, norms, use patterns, and
user base of these sites to ensure that strategies and messages resonate with the intended
audience. While there are tremendous advantages to using social media for health
promotion, further studies are necessary to advance the theory of online influence
cial media interventions. mechanisms in order to inform the design of effective so

39

Table 1.1 Self Reported Sample Characteristics (%, 6 n=15 3)
T1 T2
Age (mean) 15.1 15.4
Gender
thnicity
Female 50.2 50.6
E

ic/Latino

66.4
2

65.3
2
Hispan
Asian/Asian-American
/Black
5.0 5.2
White
can
an
5.3
1.9
5.7
2.7 African-Ameri
c
anguage sp
Native Ameri 1.4 1.1
L oken at home
Only English
age equally
13.4 13.6
Mostly English
gu
19.1
5
14.1
19.9
5
14.1
English and another lan 0.9 0.3
Mostly another language
sehold)
Only another language
people in hou
-price lunch
2.5
.95
2.1
.95 SES (#rooms in home/#
ligible for f d
cademic Ac
E ree/reduce 86.1 87.0
A hievement
11.1

10.3 Mostly A's
Mostly A's and B's 24.0 27.6
Mostly B's 6.3 4.5
Mostly B's and C's 28.7
15.2
27.8
12.4
Mostly C's 6.7 7.4
Mostly C's and D's
and F's
Mostly D's
ostly D's

1.1
5.7
1.9
6.2 M
ealth Statu
Mostly F's 1.4 1.9
H s
ent
ood
Excell
g
18.2 20.7
Very 30.3
38.2
11.6
28.4
37.1
11.4
Good
Fair
obile phon
Poor 1.8 2.5
M e use
Own or borrow
ccess Internet
om

96.3
50.4
97.4
54.4
riends n
moking
A
F inated (mean)
le
5.52 5.28
S
Non-susceptib 59.0
22.0
62.2
22.0

Susceptible
Ever smoked
Past 30 day smoke
9.3
6.7
8.0
5.5
40
Daily smoke
ible
3.1 2.4
Alcohol
t

Non-suscep 35.3 38.0
Susceptible 4.7 5.2
Ever drink 30.6 32.3
Past 30 day drink 13.1 10.9
Binge drink
t least one friend smoked a cigarette
t least one friend had alcohol once/month
16.3
40.6
38.7
13.6
37.6
35.3
A
A

T ctivi N=1,536) able 1.2 Social Network Site A ty (
T1 T2
Facebook use (mean scores: 1=Never, 5=Very frequently)
5
55.9 (2.75) 78.5 (3.58)
Myspace use (mean scores: 1=Never, 5=Very frequently) 70.2 (2.72) 41.9 (1.62)
1.58 (2.21)
3.23 (2.50)
Friends on Myspace (mean, sd) 2.70 (2.61)
1.83 (2.37) Friends on Facebook (mean, sd)
At least one friend talks about partying online (%)
At least one friend posts party/drinking pictures online (%)
34.9
20.7
31.0
18.5

5
Facebook and Myspace use are not mutually exclusive.
41
Table 1.3 Comparisons of Mutually Exclusive e Gro Social M
F k
dia Use
Myspace
ups
Neither
aceboo
only only Both
Significance
N=202 N=254 N=446 N= 8 58 (X )
2
% % % %

Female 39 47 50 56 17.7***
Hispanic 75 24 87 67
.96
285.5***
( *
Speak only English with friends 52 57 45 54 26.3**
SES (mean)
duced price lunch
ent (As and Bs)
.88 1.2
75
.79 F)23.6**
1
Eligible for re 85 92 87 36.6***
Academic achievem 37
21
44
66
8
35
25
42
70
39
33
64
59.1***
91.4***
93.0***
Ever smoked
Ever drank alcohol
*p<.05, **p<.01,***p<=.001

42
Table 1.4 Associations between Online SNS Activity an Behaviors ( d Risk
king
N=1256
β)
ol Use
N=12
Smo T2 at Alcoh T2 at
26
Facebook friend .04 .05 .04 .04
Myspace friend .03 .02 -.03 <-.01
Friend posted pictures partying or drinking .09*** .08 .06* .16**
Friend talk about partying online -.03 -.03 .02 -.02
Facebook frequency -.02 -.02 -.01 -.01
MySpace frequency <-.01 <.01 .06* .06*
Smoking/drinking at T1 .67*** .67*** .53*** .53***
Friend smoking/alcohol use .08*** .08* .08*** .12**
Parent smoking/alcohol use .05* .05* .08*** .08***
Age -.05* -.05* -.04 -.04
Female -.05* -.05* .01 .01
Hispanic <.01 <.01 .02 .02
SES (#ppl/#rooms in home) .03 .03 .03 .03
Academic achievement -.08*** -.08*** -.06* -.06*
Health status .01 .01 <.01 <.01
Number of friends nominated -.12*** -.12*** -.04 -.05
School 1 .03 .03 .03 .02
School 2 .02 .02 .01 .01
School 3 .02 .02 .02 .02
School 5 <.01 <.01 <.01 -.01
Friend smoke/drink * Friends on Facebook -.01 <.01
Friend smoke/drink * Friends on MySpace .02 -.05
Friend smoke/drink * Friends post risky pictures
online .01 -.13*
Friend smoke/drink * Frie
online
nds talk about partying
<.01 .07
*p<.05, **p<.01,***p<.001

Figure 1.1 Interaction Effect between Friend Drinking Status and Risky Online
Exposure on Alcohol Intake
43

44
use outcomes were partially mediated by behavioral intentions.
Findings suggest that adolescents are not only influenced by the risk behaviors of
their friends, but also through the risk behaviors exhibited through online communication
channels. Although less in magnitude compared with IPI, it is important for health
intervention programs to begin exploring ways to educate students and families about the
CHAPTER 3 — A Comparison of In-person and Digital Social Influences
on Adolescent Smoking and Alcohol Use

Abstract
Social influence processes among friends are one of the strongest predictors of
smoking and alcohol use among adolescents. Intentions to perform the behavior often
precede the behavior itself, but less established are the pathways by which intentions are
formulated, whether through existing friendships or vicariously through different types of
media channels. This study compares the relative impact of in-person influences (IPI) and
digital influences (DI) on adolescent tobacco and alcohol intentions.
Data for this study consisted of students from five Southern California high schools.
Paper-pencil surveys were administered during October 2010 and April 2011, and a total
of 1563 students completed the survey at both waves. Personal and grade level network
data were collected, as well as adolescent tobacco and alcohol use and intentions, IPI, DI,
and subjective norms in relation to the risk behaviors.
Confirmatory factor analysis and structural equation models (SEM) were conducted
to test the hypothesized associations between IPI, DI, and behavioral outcomes for both
smoking and alcohol use. Results indicated that IPI and DI were positively associated with
intentions to perform those behaviors. Effects of in-person social influences to smoking and
alcohol
45
underlying effects of risky content viewed through online platforms, as well as ways to
harness positive media influences to facilitate the diffusion of healthy norms and practices.

46
Introduction
Adolescent Social Influences and Risk Behaviors
Multiple factors come into play when examining the risk factors associated with
adolescent and smoking and alcohol use (Mathur, Erickson, Stigler, Forster & Finnegan,
2013; Slater & Henry, 2013; Steinberg, 2008). Among these are social influence processes,
which have been demonstrated to be one of the strongest predictors of smoking and
alcohol use for this age group (Flay, 2009; Kobus, 2003; Steinberg, 2008; Valente, Fujimoto,
Soto, et al., 2013). Social influences are often conceptualized as perceived prevalence of
these behaviors among friends (Duan, Chou, Andreeva & Pentz, 2009), or actual exposure
to friends’ self-reported behaviors (Kobus, 2003).
Socio-cognitive theories such as the Theory of Reasoned Action/Planned Behavior
(TRA/TPB) (Fishbein & Ajzen, 1975) suggest that behaviors are manifest through one’s
intentions to perform the behavior. Many studies have found evidence of longitudinal
associations between intention and behavior. Less established are the pathways by which
intentions are formulated. Based on the TRA/TPB, immediate antecedents to behavioral
intentions are one’s attitudes toward the behavior and perceptions of positive or negative
outcomes (Montano & Kasprzyk, 2008). One’s perceived subjective norms or beliefs about
others’ attitudes towards the behavior, which are further classified into two types –
descriptive norms, which are the perceptions of what others do and injunctive norms,
which are the perceptions of what others think about the topic (Rimal & Real, 2003).
Recent studies suggest that descriptive norms are more strongly associated with
intentions (Olds, Thombs & Tomasek, 2005; Vitoria, Salgueiro, Silva & de Vries, 2011).
47
Social influences are also conceptualized as an observational learning processes
(Bandura, 1977) whereby individuals learn from role models who might be close friends,
family members, or a prominent figure from the media (Nash et al., 2005). Behaviors are
reinforced when a lack of negative consequences are shown in conjunction with the risk
behaviors. Whether fostered through face-to-face or social media channels, these
relationships have the potential to influence adolescents’ attitudes and subjective norms
regarding risky behaviors such as smoking and drinking alcohol.
Social Media and Health
To date, there is limited knowledge about the potential effects of social media
channels on the mechanisms of social influence and health behavior choices. Adolescents
are avid users of social media sites such as Facebook and Myspace for communication with
friends, family, and acquaintances (Lenhart et al., 2010). Thus their social networks are
dramatically expanded through “friendships” that are formed and maintained through
these online channels (Steinfield, Ellison & Lampe, 2008). More than 80% of adolescents
frequent these social network sites and use them for self expression, sharing of
photographs and for posting and following real-time status updates (Lenhart et al., 2011).
Research has shown that approximately half of these online profiles contain displays of risk
behaviors (Hinduja & Patchin, 2008); exposure to these displays are shown to be
associated with higher levels of actual risk behaviors (Moreno, Parks, Zimmerman, Brito &
Christakis, 2009).
As with traditional media influences, social media may act as a “superpeer”
(Strasburger, Jordan & Donnerstein, 2010) to adolescents through a more indirect process.
Studies have shown that television shows or movies can increase favorable attitudes
48
towards risk behaviors such as alcohol and tobacco use when the viewers form parasocial
relationship with the characters (Cin et al., 2009; Stoolmiller et al., 2012; Tickle, Hull,
Sargent, Dalton & Heatherton, 2006; Wills, Sargent, Stoolmiller, Gibbons & Gerrard, 2008).
The effects of social networking sites may yet be stronger than parasocial influences given
that SNSs are typically relationships that already exist within the face-to-face context
(Giles, 2002). Online social networking practices enable the capacity to rapidly
disseminate information, the good and bad. Given that adolescents are unlikely to post
negative consequences of their risk behaviors may result in biased normative perceptions
about prevalence and outcomes of these risk behaviors.
Current Study
The goal of this study is to compare the relative impact of in-person and digital
social influences on adolescent tobacco and alcohol intentions and use based on the TRA
and observational learning mechanisms (Figure 2.1). Given the prevalence of social media
in adolescents’ lives (Lenhart et al., 2010), it is important for public health professionals to
understand the degree to which this medium might add to our current understanding of
social influences and risk behaviors. Further understanding of these potential mechanisms
nt or delivery of new intervention strategies. of change may also inform the developme
are examined: Three specific hypotheses
H1. Exposure to risky in-person influences (IPI) is positively associated with subjective
norms that favor smoking/alcohol use and intentions towards current risk
behaviors.
49
H2. Exposure to risky digital influences (DI) is positively associated with subjective
norms that favor smoking/alcohol use and intentions towards current risk
behaviors.
H3. Smoking/alcohol norms and behavioral intentions mediates the relationship
between social influences and smoking/alcohol use outcomes.
Methods
Study Design and Data Collection
Data for this study were drawn from the first two waves of the Social Network Study
(Valente, Fujimoto, Unger, et al., 2013), which were collected in October 2010 and April
2011. The study took place at five comprehensive high schools in the El Monte Union High
School District (EMUHSD) located in the city of El Monte just east of the Los Angeles
metropolitan area with residents comprising 69% of Hispanic and 25% of Asian
racial/ethnic origin.
A total of 2,290 enrolled 10
th
grade students across all five high schools were
surveyed, of which 2,016 returned valid parental consent forms (88.0%) and 1,823 agreed
to participate in the study. Some 28 of these students did not provide student assent,
reducing the eligible pool to 1,795 students of whom 1,707 completed surveys (74.5%
overall participation rate) at the first wave of data collection. A total of 1,626 students
completed the survey at time two. Both waves of data were collected using paper and
pencil surveys during class on a regular school day.
Network Data Collection. Personal (egocentric) and grade level (sociometric)
network data were collected. For the egocentric network data, students were asked to
name up to “seven best friends regardless of where they live or go to school” and then
50
asked to provide demographic and risk behavior information about each of their friends.
For the grade level networks, students were asked to nominate up to 19 friends using a
printed roster with photographs and identification numbers of all the students in the tenth
grade (Valente, Fujimoto, Unger, et al., 2013).
Measures
Tobacco and Alcohol Use. The outcome variables for this study were students’ self-
reported tobacco and alcohol use. Smoking and alcohol use was assessed by students’ self-
reports of use based on established measures adopted from the 2011 National High School
Youth Risk Behavior Survey (Centers for Disease Control & Prevention, 2010). Smoking
was first coded into a composite score based on the responses to four questions about
smoking frequency. The items that were used to create the composite score included: age
smoked first whole cigarette (responses other than “never smoked” were coded as “ever
smoker”), ever tried cigarette smoking, even one or two puffs (“yes” was coded as “ever
smoker”), number of days smoked cigarettes during the past 30 days (responses other than
“zero days” were coded as “past month smoker”), and daily smoking for 30 days (responses
other than “zero days” were coded as “daily smoker”).
Alcohol use was coded the same way into a composite score with four response
categories. The items used to create the composite score included: age at first drink of
alcohol, number of days having at least one drink of alcohol, number of days having at least
one drink of alcohol in the past 30 days (responses other than “zero days” were coded as
“past month drinker”), and number of days having five or more drinks of alcohol in a row in
the past 30 days (responses other than “zero days” were coded as “past month binge
drinker”).
51
Tobacco and Alcohol Use Intentions. Smoking and drinking intention were assessed
using one item: “At any time in the next year, do you think you will smoke a cigarette?”
(1=no, definitely not, 2=maybe no, 3=maybe yes, 4=yes, definitely).
In-person Influences (IPI). Three measures were selected for the smoking social
influence latent variable, and the three parallel measures for alcohol use social influence.
Four of the six measures for smoking and alcohol use were derived from the egocentric
network data of students’ nominated best friends. The questions “has this person ever
smoked a cigarette/drink alcohol once a month?” and “have you ever smoked a
cigarette/drunk [sic] alcohol with him/her?” (0=no; 1=yes) were summed across all
nominated friends to determine the number of the nominated friends (range of 0-7) who
smoked/drank alcohol and performed these behaviors with the student. The measures
were subsequently recoded to a six point scale (0-5+) by collapsing the top three categories
with lower responses. The third measure, exposure to friends’ smoking and alcohol use
behaviors, was based on the sociometric network data at the grade level. Students’
exposure to smoking and alcohol use among friends in their grade level was calculated by
taking the average of their grade-level nominated friend’s self-reported tobacco and
alcohol use measures.
Digital Influences (DI). The two indicators that were used to represent the degree of
risky online influences were derived from the egocentric network data of the seven
nominated best friends. These included whether the number of friends who posted
pictures of themselves partying or drinking alcohol online, and the number of friends who
talked about partying online. Online risky behaviors of friends were assessed by asking
52
whether alters ever “posted pictures of themselves partying or drinking alcohol online” and
“talk about partying online.”
Smoking and Alcohol Subjective Norms. Three items were used to assess students’
perceived injunctive norms for smoking and alcohol use, which asked: how many of their
five best friends would “think it’s OK for someone your age to smoke,” “ever offer you a
cigarette,” and “be unfriendly towards you if you smoked.” Students responded with
1=none of them, 2=one or two friends, 3=three or four friends, and 4=all five friends.
Covariates. Three measures that have previously been shown to significantly affect
the relationships between social influences and risk outcomes were included in this study.
These included students’ gender (0=female, 1=male), ethnicity (0=non-Hispanic,
1=Hispanic), and academic performance (1=mostly F’s, 9=mostly A’s). Other covariates
including access to the Internet, socioeconomic status, and network centrality scores
(indegree, betweenness, and closeness) were tested but found to be insignificant.
Analysis Plan
All descriptive analyses were conducted using STATA 12.0 and statistical modeling
procedures were conducted using Mplus software version 6.0 (Muthén & Muthén, 2012).
Descriptive analyses were first performed to determine the characteristics of our sample.
Next, bivariate correlations between all indicator variables were conducted separately for
the smoking and alcohol use outcomes. Confirmatory factor analyses (CFA) were
performed to establish the latent measurement models of IPI, DI, and subjective norms
regarding tobacco and alcohol use practices.
Structural equation models (SEM) were conducted to test the hypothesized
associations between IPI, DI, and behavioral outcomes for both smoking and alcohol use.
53
Direct and indirect structural relationships between the factors were assessed and
exogenous covariates were then added and tested to determine the degree to which they
would add explanatory variance to the models. Full information maximum likelihood
(FIML) was used to generate robust estimates for data with missing responses. The
maximum likelihood MLR estimator was used to adjust for the non-normality of data
(Muthén & Muthén, 2012), and as such, the Chi-square test was not used for testing model
fit. Several other criteria were used to determine the fit of estimated models to the data
including the comparative fit index (CFI) of >.95, the root mean square error of
approximation (RMSEA) of <.06, and the standardized root-mean-square residual (SRMR)
of <.08 for detecting model misspecification (Kline, 2010). Post hoc modification indices
(MI) supplied by Mplus software are considered to improve the overall model fit.
Results
Sample Characteristics
Table 2.1 presents the descriptive demographic characteristics of our sample across
the two waves of data collection. Participants who responded to the survey were on
average 16 years old and were evenly distributed across both genders (51.2% female).
About two-thirds were of Hispanic/Latino racial ethnic origin and nearly one-fourth were
Asian, closely reflecting the El Monte City ethnic makeup. An indicator for socioeconomic
status (SES) was calculated by dividing the “number of rooms in the home” by the “number
of residents in the household.” Students reported on average having 3.2 rooms in their
home and a mean SES score of .93 (sd=.611). About two-thirds achieved grades that were
mostly B’s and C’s or higher.
54
Students’ risk behaviors as well as digital and in-person indicator variables are
presented in Table 2.2. Less than one-third (29.5%) reported ever smoking (even one or
two puffs) and 56.5% reported having had at least one drink of alcohol (other than for
religious purposes). About one-fifth (27.6%) reported having friends who posted risky
pictures online and 46.4% reported having friends who talked about partying and drinking
online.
Bivariate Correlations
Table 2.3 and 2.4 include the bivariate correlations of all manifest variables included
in this study as well as their means, standard deviations and ranges. As expected, smoking
and drinking outcomes were strongly associated with intentions use in this study (r=.57,
p<.001; r=.61, p<.001). Indicators of IPI were moderately correlated for smoking (r=.25-
.57, p<.001) and alcohol use (r=.38-.66, p<.001). The two indicators for DI were correlated
at r=.47 (p<.001) and the two indicators for injunctive norms were strongly correlated for
both outcomes (r=.54, p<.001; .76, p<.001).
Bivariate correlations between IPI, DI variables and alcohol use outcomes were very
similar to the correlations among smoking variables. Intention to use alcohol was the most
highly associated variable with alcohol use at time 2 (r=.634, p<.0001), followed by
injunctive normative beliefs about friends’ attitudes towards drinking (r=.544, r=.539), and
exposure to friends who report drinking alcohol (r=.316), all significantly correlated at
p<.001.
The correlations between the predictor variables and alcohol intentions were, on
average, similar but slightly stronger in association. Overall, bivariate correlations
55
between both IPI and DI were associated with students’ smoking and drinking status at
time 2.
Confirmatory Factory Analysis
An initial CFA indicated good fit for IPI, DI, and subjective norm latent variables in
both smoking and alcohol use models (CFI=.967, RMSEA=.059, SRMR=.039; CFI=.987,
RMSEA=.045, SRMR=.021, respectively). This resulted in a high correlation between the IPI
and “subjective norms” latent variables (smoking .89 and alcohol .77). Given that injunctive
norms have been considered as a measure of social influence, the indicators for “subjective
norms” were re-specified as part of the IPI latent variable (Figure 2.2).
The resulting CFA models revealed an improved fit for both smoking and alcohol use
models (CFI=.979, RMSEA=.047, SRMR=.028; CFI=.987, RMSEA=.043, SRMR=.023,
respectively). These modified CFA models were then used for subsequent addition of
structural effects. Cronbach’s alpha (α) for the smoking and alcohol use influence
indicators was α=.745 and α=.825 respectively, indicating satisfactory reliability.
Structural Equation Models
Structural models were specified for both smoking and alcohol use outcomes by
adding directional pathways from IPI and DI to behavioral intentions and then from
intentions to the risk outcomes at T2. Structural models were first tested before adding
exogenous covariates. These preliminary models resulted in an unexpected suppressor
effect indicated by negative direct effects of DI Æ behavioral intention and DI Æ T2 behavior
despite positive bivariate correlations between these variables (Table 2.4). The change in
sign may have been attributed to the strong correlation between the two latent social
influence variables (.69 for smoking and .76 for alcohol use), which then caused DI to act as
56
a suppressor of the effect from IPI. The negative coefficient suggests that DI may share
more of the unexplained variance from IPI than with intentions or behavior, therefore
suppressing the effect of IPI. A second indication of a suppressor effect was that the path
coefficients were greater than the bivariate correlation coefficients (Kline, 2010; Maassen
& Bakker, 2001). Given that IPI, DI and intentions were all taken at the same observation
period (T1), an alternate model was tested replacing the DI Æ intention and IPI Æ intention
causal paths with correlational paths, constraining DI Æ behavior to .001 (CFI=.979/.987,
RMSEA=.047/.043, SRMR=.028/.023).
Two final models (A and B) for both smoking and alcohol use outcomes were fitted
with covariates to account for the effects of gender, ethnicity, and academic achievement.
The first model (Figure 2.3, model A) retained the correlational pathways between IPI, DI
and intentions 001 (CFI=.923/.949, RMSEA=.064/.064, SRMR=.028/.023). A second model
(Figure 2.4, model B) was fitted to test the direct causal effects IPI Æ intention and
DI Æ intention by constraining the significant negative effect to .001 (CFI=.923/.944,
RMSEA=.063/.065, SRMR=.041/.038). Post hoc adjustments based on the Mplus
modification indices resulted in four additional correlations of indicator error terms for the
smoking models and three additions for the alcohol use models.
The paths between intentions at T1 were significantly associated with behaviors at
T2. Model A resulted in significant positive associations between IPI and both smoking
and alcohol intentions (.66 and .75, respectively). Model A also shows a positive
association between DI and intentions (.36 and .42, respectively). H3 was only tested in the
alternate model (B) with the causal associations between peer influences and intentions.
Though digital influence paths were not estimated due to suppression effects, results
57
suggest that in-person Æ T2 behavior is partially mediated by intention (indirect effect=.17)
and that the direct effect of IPI on T2 behavior was stronger (direct effect=.53).
Discussion
This study sheds light on the different types of peer influence mechanisms that lead
to higher levels of smoking and alcohol use over time. This study is the first to decompose
peer influence effects into in-person and digital domains of influence, which were validated
through confirmatory factor analyses and further tested with structural equation models
informed by the theory of reasoned action (Fishbein & Ajzen, 1975). The measurement of
digital peer influence sources is a timely topic of inquiry given the prevalence of adolescent
social media use (Lenhart et al., 2010) and the potential risks related to the inappropriate
display of online content by friends (Moreno, 2011).
The hypothesized relationships between peer influences and subjective norms were
unable to be tested due to the lack of valid measures for subjective norms. The resulting
structural models of this study, however, confirmed the relationship between influences
and intentions for both smoking and alcohol use. Social influences based on friends’ risk
behaviors (H1) as well as influences based on friends’ online risk behaviors (H2) were
positively associated with intentions to perform those behaviors. Effects of in-person social
influences to smoking and alcohol use outcomes were partially mediated by behavioral
intentions.
For both smoking and alcohol use models, digital effects were only directly
associated with behavioral intention from T1 with no significant direct effect on T2
behaviors. In-person effects, however, were extended to behaviors through both direct and
indirect effects. One possible explanation is that online relationships are confined in
58
cyberspace and that the exposures do not immediately translate into action. On the other
hand, behaviors may be more readily transmitted when friends gather in person whereby
adolescents may be offered a cigarette or an alcoholic beverage.
The meditational role of intentions between IPI and behavior are supported by the
study findings. However, there is stronger evidence that IPI may affect behavior directly.
These findings are consistent with dual-processing models such as the
Prototype/Willingness Model (Gibbons, Houlihan & Gerrard, 2009), which describes risk
behavior adoption as a reactive process without much cognitive processing or intentional
rationalization. Decisions are instead based on one’s willingness to be like prototypes they
admire who also engage in the behavior.
Mediation effects between DI and outcomes were not estimated due to suppression
effects. However, future studies may consider further exploration of the “willingness”
construct as a mediator of both IPI as well as DI processes on adolescent risk taking
(Gibbons et al., 2009). Since DI were only indirectly associated with behavior, improved
measures are needed to assess the underlying mechanisms that mediate the relationship
between digital media exposure and risk outcomes. Investigators may wish to adapt
measures from those used in media studies such as those used to assess movie alcohol
exposure (MAE) effects (Sargent, Worth, Beach, Gerrard & Heatherton, 2008) and the recall
of images based on the Prototype/Willingness model.
Confirmatory factor analyses based on the measures of this current study suggest
moderate discriminant validity between in-person and digital sources of peer influence.
Digital effects and in-person effects may be cumulative; however future studies may
investigate how the interactions of these effects affect intentions or other mediators of risk
59
behavior. Multiple measures over time and cross-lagged models may be used to shed light
on this. Furthermore, network methods to examine the effects and mechanisms of social
influences through digital social network channels (Rice & Karnik, 2012). For example,
whether socially marginalized students might be more susceptible to DI compared to
students who are more centrally positioned among the peers in their networks (Valente,
2010).
Limitations
Several limitations specific to this study should be noted. While positive
associations between our latent constructs were established through the structural models,
the causal associations between these constructs were clouded by an unexpected
suppression effect. This may have been attributed to measurement limitations due to
insufficient indicators to sufficiently discriminate between IPI and DI. The digital influence
construct was based on only two indicators that represented the degree to which they were
exposed to risky online content posted by their seven closest friends. While IPI may have
appeared to be stronger among best friends, potential sources of DI can be more
widespread. It is common for adolescents to be linked with upwards of hundreds of
“friends” most of which are weak ties, but may still exert a similar amount of face time
compared with online friends who are also closer friends from the offline context. Our
study focuses on DI exhibited by existing friends from the in-person context, however,
future studies may wish to examine a broader scope of online influences by incorporating
other types of online platforms or specific online applications through which adolescents
interact.
60
The measures used to operationalize in-person influences in this study included
peer influence as well as items that traditionally represent subjective norms (perceived
prevalence) and thus not able to be modeled as separate factors. Improved measures for
both descriptive and injunctive social norms should be encouraged (Manning, 2009; Rimal
& Real, 2005) to more effectively inform the development of intervention strategies.
Secondly, self-reported smoking and drinking measures were not bio-chemically
validated, which could introduce some biases in our findings. The use of network exposure
measures provide a more objective measure of influence based on peers’ reports of their
own behaviors. This is a significant strength over traditional measures of peer influence
based on the student’s reports of the number of their friends who perform the behavior.
Implications
Adolescents are not only influenced by the risk behaviors of their friends (IPI), but
also through the risk behaviors that are posted online by these friends (DI). Although less
in magnitude compared with IPI, it is important for health intervention programs to
incorporate modules to educate students and families about the underlying effects of risky
content viewed through online platforms, especially those posted by friends. Similarly,
students may be educated about online impression management practices such that online
posts are not necessarily accurate representations of reality. Adolescents should also be
informed about the negative effects of viewing risky content online and the consequences
that these postings may have on friends.
More importantly, as peer influences are shown to occur though online channels, it
is critical for health interventions to explore positive media influences that offer pro-social
mechanisms and facilitate the diffusion of healthy norms and practices. The potential for
61
positive behavioral change can be promising if health professionals learn how to effectively
harness the communication processes through online social media platforms.
62
Table 2.1 Study S haracteristics (N=1,707) ample C
%
Gender Female 51.2
Ethnicity
Hispanic/Latino
rican
6
2
7.8
Asian/Asian-Ame 2.6
White
k
5.1
Native American
African-American/Blac
sehold (mean)
2.6
1.8
3.2 Number of rooms in hou
SES (mean) .93
Academic Achievement
Mostly A's 10.7
Mostly A's and B's 23.3
Mostly B's 6.2
Mostly B's and C's 28.5
1
Mostly C's 6.5
Mostly C's and D's
and F's
5.4

Mostly D's
Mostly D's
Mostly F's
1.1
6.4
1.9

63
Table 2.2 Sm asurement Model Indicators (T1) oking and Alcohol Use Me
N=1,707
%
# Egocentric s who smoke friend
None 58.3
14.1
10.5
One
Two
Three 7.6
4 Four
Seven
Egocentric s whom I smoke with
.6
Five to
d
4.9
# frien
None 85.9
One 6.2
Two 3.7
Three 2.2
.8 Four
ive or more
rade friends (range 1-5)
F 1.2
Average smoking levels of 10
th
g

d
1.88 (.786)
# Egocentric s who drink frien
None 60.4
1
1
One 1.5
Two 0.8
Three 6.2
5.3 Four
Seven
Egocentric s whom I drink with
Five to
d
5.7
# frien
None 70.6
One 8.3
Two 7.4
Three 5.5
3.6 Four
ive or more

F 4.6
Average drinking levels of 10
th
grade friends (range 1-5)
2.80 (1.01)
# Friends post pictures of partying and drinking online
None 78.4
One 11.1
Two 4.6
Three or more 5.9
# Friends talk about partying online
None 63.6
One 13.9
Two 8.1
Three or more 14.4
64
Smoking inte ) ntion (12 months
No, definitely not 72.3
1
1
Maybe no 2.8
0 1 Maybe yes
lcohol inten
.
Yes, definitely
onths)
4.7
A tion (12 m
No, definitely not 49.5
Maybe no
aybe yes
12.4
24.8 M
2 Smoking
Yes, definitely 13.3
T
Never smoked
moke
7
2
0.6
Ever smoke
t 30 day s
1.7
5.5 Pas
2 Alcohol u
Daily smoke 2.3
T se
Never drank
drink
43.5

Ever drink
Past 30 day
Binge drink
32.4
11.0
13.1

Table 2.3 Pa Peer Influence and Smoking 0utcom Varia s irwise Bivariate Correlations between
2
e
9
ble
10 1 3 4 5 6 7 8 11 12
1 T2 Smk
2 Smk intention .53***
3 Frd smk .39*** .44***
4 Frd smk with .49*** .53*** .57***
5 Sch frd smk .27*** .28*** .34*** .25***
6 Frd Online pics .30*** .30*** .48*** .34*** .21***
7Frd Online party .18*** .28*** .39*** .23*** .17*** .47***
8 OK to smk .34*** .45*** .46*** .37*** .34*** .35*** .30***
9 Frd offer cig .42*** .52*** .44*** .47*** .32*** .32*** .24*** .54***

10 Female -.08*** .03 .06* -.04 .01 .13*** .06* .10*** .04
11 Hispanic .13*** .17*** .20*** .12*** .31*** .11*** .10*** .18*** .17*** .04
12 Academic ach. -.27*** -.31*** -.23*** -.22*** -.33 -.17*** -.14*** -.26*** -.28*** .06* -.28***
Mean .393 1.47 1.01 .295 1.88 .38 .73 1.66 1.43 .51 .68 6.14
Std. Dev. .696 .858 1.478 .866 .786 .827 1.105 .847 .704 .500 .469 1.967

Range 0-3 1-4 0-5

0-5

1-5

0-3

0-3

1-4

1-4

0-1 0-1 1-9
*p<.05; **p<.01; ***p<.001
Note: Frd=friend; smk=smoking; drk=drink; cig=cigarette
Shaded areas indicate items combined to form latent variables

65
66
Table 2.4 Pairwise Bivariate Correlations between Peer Influence and Drinking Outcome Variables
1 2 3 4 5 6 7 8 9 10 11 12
1 T2 Drk
2 Drk intention .61***

3 Frd drk .45*** .48***
4 Frd drk with .51*** .58*** .66***
5 Sch frd drk .36*** .40*** .38*** .39***
6 Frd Online pics .32*** .36*** .52*** .45*** .29***
7Frd Online party .28*** .30*** .42*** .36*** .23*** .47***
8 OK to drk .53*** .60*** .52*** .52*** .39*** .36*** .30***
9 Frd offer drk .53*** .63*** .52*** .57*** .42*** .38*** .33*** .76***

10 Female .08*** .11*** .13*** .10*** .12*** .13*** .06* .13*** .10***
11 Hispanic .22*** .22*** .23*** .21*** .39*** .11*** .10*** .25*** .25*** .04
12 Academic ach. -.27*** -.29*** -.25*** -.24*** -.34*** -.17*** -.14*** -.23*** -.25*** .06* -.28***
Mean .936 2.02 1.02 .769 2.8 .38 .73 2.03 1.92 .51 .68 6.14
Std. Dev. 1.032 1.13 1.535

1.415 1.006 .827 1.105 1.04 1.021 .500 .469 1.967
Range 0-3 1-4 0-5 0-5

1-5

0-3

0-3

1-4 1-4 0-1 0-1 1-9
*p<.05; **p<.01; ***p<.001
Note: Frd=friend; drk=drinking/dink
Shaded areas indicate items combined to form latent variables
Figure 2.1 Hypothesized Model of Digital and In-person Social Influences on
ehavior B

67
F
I

igure 2.2. Confirmatory Factor Analysis Models of Digital and In-person Social
nfluence Latent Variables for Adolescent Smoking and Alcohol Use

and coefficients on the right Note: Coefficients on the left indicate smoking influences
indicate alcohol use influences

68
Figure 2.3 Structural Equation Model (A) with Correlational Paths between Social
Influences and Intention

Note: Coefficients on the left indicate smoking influences and coefficients on the right
n-significant; dashed line indicates indicate alcohol use influences; ne=not estimated; ns=no
different smoking and alcohol use findings.

69
Figure 2.4 Structural Equation Model (B) for Social Influence Mechanisms with
Causal Paths between Social Influences and Intention.
70

Note: Coefficients on the left indicate smoking influences and coefficients on the right
indicate alcohol use influences; ne=not estimated; ns=non-significant; dashed line indicates
different smoking and alcohol use findings.

71
CHAPTER 4 — Smoking, Alcohol Use and Social Networking Sites:
Impact on Adolescent Friendship Selection and Influence

Abstract
The objectives of this study are to examine the effects of adolescent smoking,
alcohol use, and social network site use on the evolution of adolescent friends and peer
influence.
Longitudinal network data were collected during fall and spring of 2010-2011 from
1,435 10
th
grade students across five southern California high schools. A meta-analysis of
stochastic actor-based models was conducted to simultaneously estimate changes in
friendship ties and risk behaviors, and the effects of Facebook and Myspace use.
Significant shifts in adolescent smoking and drinking occurred despite little changes
in overall prevalence rates over six months. Drinkers had stronger tendencies towards
becoming friends with other drinkers and likewise influencing others to drink compared to
adolescent smokers. Adolescents who had similar Facebook and Myspace use habits
selected each other as friends and exposure to friends’ risky online pictures increased
adolescent smoking behaviors.
A greater focus on friendship selection mechanisms in school-based smoking and
alcohol use interventions is warranted. The use of social media platforms may facilitate
both selection and influence processes that effect adolescent normative risk perceptions.

72
Introduction
It is well established that adolescent risk behaviors such as smoking and alcohol use
are associated with the behaviors of their friends (Alexander et al., 2001; Ennett et al.,
2008; Hoffman et al., 2007; Kobus, 2003; Nash et al., 2005; Trucco et al., 2011). During
adolescence, peer relationships become stronger due to more shared activities and
opportunities for socialization (Steinberg, 2008). The desires for peer affirmation make
adolescents particularly susceptible to normative influences (Simons-Morton & Farhat,
2010; Urberg et al., 1997). These influence processes play a significant role in the adoption
of risk behaviors such as cigarette smoking and substance use, often related to maintaining
their social status or group membership (Cleveland, Feinberg, Bontempo & Greenberg,
2008; Oetting & Donnermeyer, 1998).
Adolescent Online Social Networking
Online social networking tools have increased the potential channels through which
adolescents interact and grow in intimacy. In the U.S., approximately 80% of adolescents
(ages 12-17) use social network sites (SNSs), and SNSs have gradually become the
preferred means of communication for teens across all racial and socioeconomic groups
(Lenhart, 2012; Lenhart et al., 2011). Adolescents frequently use SNSs to build
friendships, manage their self-presentations (boyd, 2007; Utz, 2010). and foster their sense
of social connectedness (Grieve, Indian, Witteveen, Tolan & Marrington, 2013).
There is growing concern, however, that SNSs may also serve as conduits of risk
behaviors through exposures to higher levels of alcohol use and smoking displayed by
friends (Huang et al., Under Review; Moreno et al., 2009). Influence mechanisms may
involve biased normative perceptions of risk (Leatherdale & Ahmed, 2010; Young & Jordan,
73
2013) and higher exposure to other unfavorable media sources
20
or social consequences
related to online encounters (Christofides, Muise & Desmarais, 2009; Festl & Quandt,
2013).
Peer selection and influence mechanisms
While the association between adolescent behaviors and their friends’ behaviors
have historically been attributed to peer influence mechanisms (Kobus, 2003), an
increasing body of research suggests that friendships are formed as a result of established
risk behaviors. Social network methods have been used to distinguish between the effects
of friendship selection and influence, but with mixed findings (Hoffman et al., 2007). Some
studies suggest that there is stronger support for selection effects (Knecht, Burk, Weesie &
Steglich, 2011; Mercken, Sleddens, de Vries & Steglich, 2013; Mercken, Snijders, Steglich,
Vertiainen & de Vries, 2010; Mercken, Steglich, Sinclair, Holliday & Moore, 2012; Mundt,
2013) while others find that selection and influence effects both contribute to the
association between adolescent and friend risk behaviors (Green et al., 2013; Kiuru, Burk,
Laursen, Salmela-Aro & Nurmi, 2010; Schaefer, Haas & Bishop, 2012; Steglich, Snijders &
West, 2006) (Appendix A).
Two social network principles provide the framework for selection and influence.
Homophily is the tendency for similar people to be drawn towards each other (McPherson,
Smith-Lovin & Cook, 2001). Adolescents who engage in risk behaviors are likely to be
attracted to each other (Mercken, Candel, Willems & de Vries, 2009) and consequently
these risk behaviors can become further reinforced as a result of shared activities. Social
contagion is the process by which information or ideas are diffused through a network by
contact or communication (Valente, 2010). The spread of behavior is conventionally
74
driven by opinion leaders situated in central network positions who exert influence
because others desire to be like them. Peer leaders in a school setting may exert both
positive and negative influences on tobacco and alcohol use based on the prevalence of
these behaviors and the norms in their peer environment (Valente et al., 2007).
One’s self-concept may also be defined by the social groups to which they belong as
adolescents are known to alter their behaviors to match the appropriate and group- and
self-defined norms (Hogg, Abrams, Otten & Hinkle, 2004), SNS content displayed by friends
may create the appearance of certain normative practices that others may desire to
emulate. Visual images or videos that friends post on SNSs may hasten this observational
learning process between friends that Bandura identified as early as 1978 (Bandura).
There is emerging evidence on the association between adolescent social media use
and higher risk behaviors but the mechanisms of influence are unclear. Studies that have
also utilized SNSs for health interventions have found modest results. However, recent
studies increasingly suggest that social media platforms can be effective in facilitating new
norms that encourage or discourage risk-taking (Chou, Prestin, Lyons & Wen, 2013).
The aim of this study is to examine whether SNS use changes friendships based on
similar risk behaviors (selection) or contributes to the spread of risk behaviors through
exposures to friends who exhibit these risk behaviors online (influence). Findings will shed
some light on social media influence mechanisms and potentially inform the design of
social media based interventions.
Four hypotheses are tested: (H1) selection effects will be stronger than influence
effects on adolescent smoking; (H2) selection effects will be stronger than influence effects
on adolescent alcohol use; (H3) similarity in SNS use will predict friendship selection; and
75
(H4) exposure to friends’ risky online content will exhibit peer influence effects on
adolescent smoking and alcohol use.
Methods
Study Design and Data Collection
Data were drawn from the first two waves of the Social Network Study (Valente,
Fujimoto, Unger, et al., 2013), a longitudinal network study of five southern California high
schools. Paper and pencil surveys were administered during class on a regular school day
in the Fall 2010 and Spring 2011. Of the total 2,290 enrolled 10
th
grade students, 2,016
returned valid parental consent forms (88.0%) with 1,823 agreeing to participate in the
study. Some 28 of these students did not provide student assent, reducing the eligible pool
to 1,795 students of whom 1,719 completed surveys at T1 and 1,620 at T2. A total of 1,563
students completed surveys at both time points.
Sociometric (saturated) network data were collected at the grade and classroom
levels; egocentric (personal) network data were collected at the community level. For
grade and class level networks, students were asked to identify five different types of peer
networks using ID numbers on a photo roster that included all students in their grade.
Students could nominate up to 19 best friends and up to seven students whom they
thought were most admired, likely to succeed, popular, and with whom they wished to
have a romantic relationship (Valente, Fujimoto, Unger, et al., 2013),
Measures
Dependent Network Variables. The network is specified as the dependent variable in
the modeling of selection effects. The sample for this study was initially limited to the 25
grade-level networks (5 network types X 5 schools). Four of the five network types
76
(admire, succeed, romantic and popular) generally had low densities (<.01) and high tie
turnover (Jaccard index <.2) so did not satisfy the conditions for model convergence (Kiuru
et al., 2010). Analyses were thus limited to the five friendship networks. Students who did
not provide nomination information at both waves were excluded from analysis, leaving a
final analytic sample of 1,435.
Dependent Behavioral Variables. Students’ self-reported tobacco and alcohol use
were specified as the dependent behavioral variables in the modeling of peer influence
effects. Smoking was coded into a 5-point composite score based on responses to six
questions on smoking frequency and intention: 12-month smoking intention (coded as
1=non-susceptible; 2=susceptible), age smoked first whole cigarette (3=ever smoked),
number of days smoked cigarettes during the past 30 days (4=past month smoker), and
daily smoking for 30 days (5=daily smoker). Similarly, a 5-point alcohol use score was
created using five items: 12-month drinking intention (1=non-susceptible; 2=susceptible),
age at first drink of alcohol except for religious purposes (3=ever drinker), number of days
having at least one drink of alcohol during the past 30 days (4=past month drinker) and
number of days having five or more drinks of alcohol in a row during the past 30 days
(5=past month binge drinker).
Covariates. A total of four social media use variables were tested in the model.
These included two items on the frequency of visits to SNSs Facebook and Myspace in the
past month (1=never, 2=rarely [once a month or less], 3=occasionally [once a week or less],
4=frequently [once every 2-3 days], and 5=very frequently [once a day or more]). Online risk
exposures were measured using two items: the number of friends who (1) posted pictures
of themselves partying or drinking alcohol online, and (2) talked about partying online.
77
These two measures were based on the reported characteristics of up to seven best friends
from students’ nominated egocentric networks. The total was calculated and then
dichotomized due to its skewed distribution. Demographic covariates included students’
age, gender, ethnicity (1=Latino and/or Hispanic ethnic origin, 2=other), academic
performance, parental smoking, and parental drinking status.
Analysis Plan
Stochastic actor-based (SAB) models were estimated using Rsiena (Simulation
Investigation for Empirical Network Analysis) version 4.0 (Ripley, Snijders & Preciado,
2013), a software package for estimating longitudinal co-evolution models of social
networks (Snijders et al., 2010) offered through the open source statistical system R (R
Development Core Team, 2011)
The primary advantage of SAB methods over previous methods for longitudinal
network analysis is the ability to simultaneously estimate network and behavioral
dynamics while accounting for endogenous network tendencies. These may include the
tendencies for friends to nominate each other as friends (reciprocity), the tendency for
friends of friends to become friends (transitivity), or the tendency for actors with similar
attributes to become friends (homophily) (Snijders, Steglich & Schweinberger; Snijders,
Steglich & Schweinberger, 2007; Snijders et al., 2010). Individuals within a bounded
network are “actors” who may change their ties based on behaviors of others around them
uence). (i.e. selection) or change their behavior based on their current network ties (i.e. infl
Three main assumptions distinguish SAB modeling from conventional methods for
estimating individual-level changes within a network (Steglich, Snijders & Pearson, 2010).
First, changes between measurement points are modeled according to a continuous-time
78
Markov process (Norris, 1998) to simulate likely unobserved developmental trajectories
between the measurement time points. Secondly, each actor is assumed to independently
make decisions about their changes in friendship ties or behaviors and do not conspire
with others about their decisions. Lastly, actors take micro-steps of change, only one
network tie or one level of ts. behavior at a time, reducing the variation between assessmen
Hypothesis testing. The study hypotheses were tested in two stages. First, a “risk
behavior model” was specified to simultaneously estimate selection and influence of
adolescent smoking and alcohol use. Since there is strong evidence for the co-occurrence
of adolescent smoking and alcohol use (Kiuru et al., 2010), both were included as
behavioral dependent variables to simultaneously control for the two behaviors. The
second “social media model” tests whether social network site use parameters add more
explanatory power to the network and behavioral dynamics above those established in the
first set of analyses.
Model Specification. Two main functions are used to govern network and behavior
changes in both model estimations for this study. The rate function represents the average
number of changes in the network and behavior between discrete time points. The
objective function includes parameters that guide the direction of these changes. Two
types of objective functions were then specified—the network objective function, which
estimates the change in friendship ties based on adolescent attributes and the behavior
objective function, which estimates the change in behavior based on friendships and
network structure.
A set of parameters (Table 3.1) were determined a priori based on prior literature
and theoretical justification, and further tested using a forward selection process based on

f Hispanic/Latino racial ethnic origin.
Students nominated on average 5.17 friends at T1 and 5.23 at T2. The average
network densities remained similar but fluctuated within schools. Overall network
centralization decreased slightly over time suggesting that peers formed new friendships
with more peripheral members in their grade and expanded their social circles over time.
Reciprocity increased over time, suggesting that friendship bonds were strengthened.
Jaccard indices, calculated as the ratio of new ties formed over total tie changes between
79
Neyman-Rao score-type tests for interdependent effects of alternative models against a null
model (Snijders et al., 2007). The two models were separately estimated for each school,
then combined using a meta-analysis in which the means and variances of all parameter
estimates across schools were tested (Snijders & Baerveldt, 2003).
Two types of tests were employed to determine the significance of effects: (1) a
likelihood-based method using the t-ratio (mean parameter estimate divided by its
standard error) under iterative weighted least squares modification (Snijders & Baerveldt,
2003) and (2) a Fisher-type combination test of one-sided p-values (Hedges, Olkin,
Statistiker, Olkin & Olkin, 1985). Between school differences were computed based on
approximate Χ
2
tests of parameter variances based on the Snijders-Baerveldt method
(Snijders & Baerveldt, 2003).
Results
Sample Characteristics
Network characteristics and risk behaviors of the five high schools are presented in
Table 3.2. Students were evenly distributed across gender with a mean age of 15 years.
Average self-reported academic grades were “mostly Cs” and about two-thirds reported
being o

as friends (p<.001), as were students who were both Hispanic/Latino (p<.001).
Students who had higher levels of smoking showed a trend toward receiving more
nominations ( χ
2
=18.27, p=.051) but did not send out more ties. Students with similar
smoking habits had a tendency to select each other as friends ( χ
2
=19.83, p=.031).
80
the two observation periods, ranged from .28 to .34. Scores above .2 indicated that the
number of stable ties were sufficient to accurately estimate effects (Snijders et al., 2010).
Tobacco and alcohol use remained steady over time, suggesting that approximately
equal number of students increased and decreased their risk behaviors. Students reported
that most parents did not smoke (70%) and that about half (51%) drank alcohol once a
week.
Coevolution of Friendship and Risk Behaviors
Two sets of meta-analysis outcomes are presented (Tables 3-4), but the conclusions
based on the Fisher’s one-sided test ( α/2 = .025) are perferred and discussed in more
detail since the five schools consist of the entire school district under study (Ripley et al.,
2013). Parameter variances between schools were assessed using χ
2
tests and 4 degrees of
freedom (N-1). With the exception of the endogenous structual parameters and similarity
in Hispanic/Latino ethnicity, there were little between-school differences.
As expected, parameter estimates of the three endogenous structural effects were
statistically signficant (p<.001) indicating that best friends are likely to reciprocate
friendships (reciprocity), become friends with someone who was a friend of an existing
friend (transitivity), and that limiting the number of nominations to 19 results in a network
density significantly less than 50% (negative outdegree). Of the covariates, students who
were the same gender and similar in academic achievement were likely to select each other
81
Effects of Social Network Site Use on Risk-related Selection and Influence
Risk behavior effects remained stable after the SNS effects were included into the
“social media model.” Controlling for the risk behavior effects, students who were similar
in Facebook and Myspace use were likely to select each other as friends ( χ
2
=41.55, p<.001;

Friendship selection effects were much stronger for alcohol use. Students who reported
more drinking had higher tendencies to receive ( χ
2
=39.22, p<.001) and send ( χ
2
=22.03,
p=.015) more nominations. Students with similar drinking behaviors were also
ignific s antly more likely to nominate each other as friends ( χ
2
=52.69, p<.001).
In terms of behavior dynamics, students changed their drinking status more
frequently than their smoking status (2.19 times vs. 1.38 times). The negative linear shape
effects indicate that the majority of students reported no use or low levels of tobacco and
alcohol use. The significant positive quadratic shape effects indicate that both smoking and
drinking behaviors were self-reinforcing, suggesting that those who exhibited higher risk at
T1 were likely to continue the risk behavior at T2 and those abstinent were likely to remain
bstine a nt.
Being female and having higher levels of academic acheivement were marginally
protective against smoking, while parental smoking significantly predicted smoking at T2
( χ
2
=27.90, p=.002). Gender, academic achievement, and being of Hispanic race/ethnic
origin were not significantly associated with alcohol use but having drinking parents
signficantly predicted adolescent drinking at T2 ( χ
2
=25.91, p=.004). There were no
influence effects due to smoking friends but adolescents with drinking friends had
signficantly higher tendencies to increase their drinking behaviors over the six months
( χ
2
=22.37, p=.013).
82
al., 2006).
When jointly considering the structural and behavioral effects, a more detailed view
of how these evolve in a secondary school emerges. Although the overall prevalence of
smoking and drinking did not increase over six months, there were signficant shifts (micro-

( χ
2
=53.69, p<.001 respectively). Friends who were both exposed to risky online postings
by friends were also likely to select each other as friends ( χ
2
=20.77, p=.003). Regarding
behavioral influence, Facebook and Myspace use did not signficantly predict smoking and
alcohol use. Effects of exposure to risky pictures online was significantly predictive of
adolescent’s increase in smoking ( χ
2
=22.37, p=.001) and exposure to friends talking about
partying online exhibited a trend towards increased alcohol use at T2 ( χ
2
=18.86, p=.042).
Discussion
In line with previous studies, selection effects were in general stronger than
influence effects for both smoking (H1) and alcohol (H2) (Kiuru et al., 2010; Knecht et al.,
2011; Mercken, Candel, Willems & de Vries, 2007; Mercken, Snijders, Steglich, Vartiainen &
de Vries, 2010; Mercken et al., 2012; Pearson, Steglich & Snijders, 2006; Pearson, Sweeting,
et al., 2006; Steglich et al., 2006). Adolescent drinkers had stronger tendencies to send and
receive friendship nominations from other tenth-graders, which indicates that students
who used alcohol were more outgoing and popular than those who did not. Students who
had similar alcohol use habits were also more likely to select each other as friends.
Smokers were slightly more popular than non-smokers and preferred to associate with
others who smoked. Signficant peer influence effects were found for alcohol use, which
suggests that norms favoring alcohol use may be more widespread among the adolescents
in our sample and regarded as more “social” compared to smoking (Pearson, Sweeting, et
83
steps of change) in behavior and friendships. In essence, there is a coalesence of behavior
such that friends increase their reciprocity and become friends of existing friends.
Likewise, friendship evolution occurs such that smokers become friends with smokers and
drinkers with drinkers. In sum, the behaviors create attractors for network dynamics,
partitioning friendship groups into those who embrace alcohol and tobacco use and those
who do not.
The findings also shed light on potential selection and influence mechanisms related
to adolescents’ use of social network sites. Students were more likely to become friends
with others who had similar Facebook and Myspace use habits (H3). This association was
also present between friends who were exposured to their friends’ online displays of
drinking and partying. If friends’ risky online displays indicate higher acceptance of
normative perceptions of risk behaviors, our results may suggest that students feel an
affinity toward others with similar normative perceptions of risk.
Use of Facebook and Myspace did not increase the tendencies for adolescents to
smoke or use alcohol while adolescents who were exposed to their friends’ risky online
displays reported an increase in smoking at T2 (H4). This supports previous findings that
SNS use itself does not appear to pose a risk, but it is the exchanged content within the SNS
that can influence behavior (Huang et al., Under Review). Pictures posted on SNSs can be
readily transmitted, which may promote biased perceptions of risk behaviors among
friends who are connected through online social networks.
The stronger association between online risk exposures and smoking compared to
alcohol use may have been due to signficant peer-to-peer influence effects associated with
alcohol use that were not present for smoking. Alternatively, since smoking prevalence is
84
generally lower and less normative, online mechansims may perhaps serve as a more
comfortable environment for adolescents to exhibit smoking peer influences. Given the
rising prevalence and expanded utility of SNSs among adolescents, these findings provide
timely evidence that peer influence effects may be transmitted through online displays of
risk behavior.
Limitations
Several limitations should be noted. First, since the study sample was drawn from
one Southern California high school district, findings may not be generalizable to the wider
U.S. adolescent population. Results from other school districts may vary based on the
prevalence of alcohol and tobacco consumption as well as the cultural mores of social
media use. Secondly, the combined use of egocentric and sociometric attributes may not
provide direct measures of friends’ online and offline exposures. Of note, however, is that
the self reports of friends’ displays used in this study can be considered as more proximal
to the student compared to exposures encountered through general Web browsing activity.
Future studies may consider direct measures of online risk behaviors of survey
participants in order to more directly estimate the selection and influence effects of these
exposures. Thirdly, friendships were restricted to students in the same grade level and
therefore may not account for all possible sources of influence. Higher smoking and
drinking rates were found in the egocentric community-level networks of our sample,
which were not limted by boundary specifications (Valente, Fujimoto, Unger, et al., 2013).
However, given that students could nominate up to 19 friends in their grade, our study
provides a considerably reliable representation of grade-level peer influences.
Implications
85
Despite the limitations, the presence of selection mechanisms is demonstrated
through this study and should be given more attention in the design and implementation of
school-based smoking and alcohol use prevention interventions. Interventions that
incorporate SNSs or similar platforms may enhance peer selection and peer influence
processes through higher levels of engagement and accelerated diffusion of positive social
norms. SAB modeling can be an effective way to examine the use patterns of existing SNSs
such as Facebook and Myspace and whether SNS activity might be indicative of social
norms around risk.
Future studies might incorporate more precise measurements of online risk
exposure to better inform the mechanisms of online social influences. Use of SNSs or other
online health applications could be modeled as a behavioral dependent variable (Pearson,
Steglich, et al., 2006) to answer questions regarding potential direct effects of online
participation on risk behaviors. SAB models can inform the efficacy of SNS-based
interventions by examining whether users of specific social media applications are
perceived as popular, whether users are effectively reaching out to others through the
applications, or whether there are any direct associations between intervention use and
health outcomes.
Continued application of SAB modeling is necessary to test additional social and
environmental factors that may contribute to selection and influence processes. Studies
with more than two observation waves are warranted to assess the dynamics of friendship
and behavior changes with time-varying covariates. Another area of exploration might be
the interactions between SNS use and the reciprocity of friendships (Mercken et al., 2007)
and the effects of reciprocated friendships on the degree of mutual influence.
86
Additional actor-based studies may consider estimating the effects of smoking and
alcohol use on each other (i.e. whether adolescent drinkers are more likely to also befriend
smokers), which might shed light on the simultaneous progression of alcohol and smoking
initiation (Leatherdale & Ahmed, 2010) and address high risk groups for which both
behaviors co-exist. Just as a calm sea masks the dramatic activities that occur below it, this
study shows that static behavioral prevalences can disguise the considerable dynamic
nfluence and selection activities that adolescents engage in over time. i

87
T Effects o ls able 3.1 Tested
Network Dynam
f Network and Behavior Coevolution Mode
on) ics (Selecti
Rate parameter-
friendship
The average number of changes in ties over time
Endogenous structural effects
Outdegree (density) Tendency to form friendship with someone
Reciprocity Tendency to form friendship with someone who previously s
em as friend
elected
th
Transitivity Tendency to form friendship with a friend of a current friend
Covariate effects
Similarity: drinking Tendency to form friendship with someone who has similar
drinking habits
Similarity: female Tendency to form friendship with someone of the same gender
Similarity: grades Tendency to form friendship with someone who reports the
academic grades
same
Similarity: online risky
displays
Tendency to form friendship with someone who has similar
exposures to online risky displays
Similarity: online partying Tendency to form friendship with someone who has similar
exposures to online discussions about partying
Similarity: Facebook Tendency to form friendship with someone who has similar
Facebook use habits
Similarity: online display
of partying
Tendency to form friendship with someone who has similar
Myspace use habits
Behavioral effects
Alter (popularity) The effect of behavior on the tendency of receiving ties
Ego (activity) The effect of behavior on the tendency of sending ties out
Similarity The effect of two people selecting each other as friends based on
similarity in behavior*
Behavior Dynamics (Influence)
Rate parameter-
behavior*
The average number of changes in behavior over time
Linear tendency Linear distribution of behavior
Quadratic tendency Quadratic distribution of behavior
Behavioral effects
Behavior alter average Effect of the average of friends' behaviors on own behavior
Female The effect of being female on the behavior
Grades The effect of grades on the behavior
Parent Smoke The effect of parent smoking on the behavior
Friends post risky
pictures online
The effect of friends' online posts of risky pictures on the behavior
Friends talk about
nline partying o
The effect of friends' online discussions about partying on the
behavior
Facebook The effect of Facebook use on the behavior
Myspace The effect of Myspace use on the behavior
*"Behavior" parameters refer to either smoking or alcohol use.
88
Table 3.2 Descriptive Sample Statistics
Mean T1 Range T1 Mean T2 Range T2
Network Characteristics
N 287.00 215-371
Average degree
lization
5.17
0.018
0
4.21-5.87 5.23
0
4.17-6.27
Density .015-.023
.031-.044
.233-
.019 .013-.025
.
.
Degree Centra
Reciprocity

.034
.259
0.031
.255
0
028-.037
229-.299
.28 - .34
.284
Jaccard index .302
Descriptive Characteristics
Tobacco use 1.79 1-5 1.75
Alcohol use 2.62 1-5 2.55
Age 15.10 14-18
Female
t-highest)
0.51 0-1
Hispanic 0.66 0-1
Grades (lowes 6.21 1-9
Parent smoke 0.36 0-2
Parent drink
isky pictures online
out partying online
0.62 0-2
Friends post r
b
0.41 0-1
Friends talk a 0.72 0-1
1-5
1-5

Facebook use 2.81
Myspace use 2.70

89
Table 3.3 Meta Analysis Results (Base model)
Snijders-Baerveldt method Fisher's combination 1-sided test (df=10)
parameter se Sd
χ
2

(right side) p
χ
2

(left side) p
Network Dynamics
Rate-friendship 12.875***1.0262.29*** 2764.20 0 1
Outdegree (density) -2.613*** 0.050 0.11*** 0 1 36635.90 <.001
Reciprocity 1.626***0.1050.24*** 2230.30<.001 0 1
Transitive triplets 0.347*** 0.037 0.08*** 1431.70 <.001 0 1
Female similarity 0.423*** 0.024 0.05 403.80 <.001 0 1
Grades similarity 0.441*** 0.049 0.11 91.00 <.001 0 1
Hispanic similarity 0.185+ 0.081 0.18*** 140.90 <.001 2.29 0.994
Smoking alter 0.032+ 0.012 0.03 18.27 0.051 2.58 0.990
Smoking ego -0.012 0.021 0.05 7.63 0.665 13.36 0.204
Smoking similarity 0.126 0.139 0.31 19.83 0.031 7.52 0.676
Drinking alter 0.050* 0.011 0.02 39.22 <.001 0.68 1
Drinking ego 0.027+ 0.012 0.03 22.03 0.015 2.64 0.989
Drinking similarity 0.372* 0.085 0.19* 52.69 <.001 1.09 1

Smoking Behavior Dynamics
Rate-smoking 1.377***0.0730.16 252.10<.001 0 1
Linear shape -1.088*** 0.0850.19 0 1 250.70 <.001
Quadratic shape 0.206***0.0180.04 56.56<.001 0.12 1
Average alter smoking 0.120 0.158 0.35 13.68 0.188 4.93 0.896
Female -0.198+0.0920.20 3.400.97 17.78 0.059
Grades -0.066*0.0200.05 2.760.986 18.37 0.049
Parent smoking 0.258+ 0.097 0.22 27.90 0.002 2.64 0.989
Hispanic -0.0160.1570.35 10.490.399 8.72 0.559
Drinking Behavior Dynamics
Rate-drinking 2.194***0.1390.31 350.50<.001 0 1
Linear shape -0.348** 0.0530.12 0 1 122.40 <.001
Quadratic shape 0.113* 0.027 0.06 45.40 <.001 2.29 0.994
Average alter drinking 0.216+ 0.094 0.21 22.37 0.013 2.13 0.995
Female 0.0370.0920.20 17.720.186 8.61 0.569
Grades -0.0180.0120.03 5.120.882 11.40 0.327
Parent smoking 0.134+ 0.057 0.13 25.91 0.004 2.52 0.991
Hispanic -0.0140.0490.11 7.500.677 7.95 0.633
***p<.001, **<.01, *<.05 (two-sided tests)
sd=between school standard deviation, significance based on appr
Note: bolded values indicate significant hypothesis test outcomes
oximate χ
2
test (df=4)
90
Table 3.4 Meta Analysis Results (SNS model)
Snijders-Baerveldt method Fisher's combination 1-sided test (df=10)
parameter se Sd
χ2
(right side) p
χ2
(left side) p
Network Dynamics
Rate-friendship 12.932***1.0382.32***2956.758<.001 <.001 1
Outdegree (density) -2.622*** 0.051 0.11*** 0.000 1 37417.370 0.001
Reciprocity 1.623***0.1060.24***2490.541<.001 <.001 1
Transitive triplets 0.342*** 0.037 0.08*** 1293.292 <.001 <.001 1
Female similarity 0.423*** 0.024 0.05 392.631 <.001 <.001 1
Grades similarity 0.417*** 0.050 0.11 80.056 <.001 0.019 1
Hispanic similarity 0.169 0.079 0.18*** 116.365 <.001 2.371 0.993
Online risk pic. similarity 0.062*** 0.008 0.02 20.773 0.023 1.427 0.999
Online party similarity 0.005 0.026 0.06 11.826 0.297 9.273 0.506
Facebook similarity 0.134*** 0.017 0.04 41.550 <.001 0.454 1
Myspace similarity 0.212*** 0.033 0.07 53.694 0.001 0.259 1
Smoking alter 0.031+ 0.013 0.03 17.863 0.057 2.858 0.985
Smoking ego -0.014 0.020 0.04 7.194 0.689 14.511 0.151
Smoking similarity 0.098 0.141 0.31* 18.043 0.057 8.803 0.551
Drinking alter 0.053* 0.013 0.03 38.157 <.001 0.649 1
Drinking ego 0.035+ 0.014 0.03 28.597 0.001 2.187 0.995
Drinking similarity 0.363* 0.096 0.21 54.117 <.001 1.809 0.998

Smoking Behavior Dynamics
Rate-smoking 1.360***0.0620.14 269.964<.001 <.001 1
Linear shape -1.111*** 0.078 0.17 0.003 1 206.139 0.001
Quadratic shape 0.167*** 0.013 0.03 36.224 <.001 0.489 1
Average alter smoking 0.024 0.120 0.27 9.221 0.511 6.441 0.777
Female -0.321+0.1320.30 2.6100.989 23.987 0.008
Grades -0.067*0.0190.04 2.5960.989 17.272 0.069
Parent smoking 0.260+ 0.113 0.25 26.279 0.003 3.323 0.973
Hispanic 0.0590.2120.47 12.3680.261 8.266 0.603
Online risk pictures 0.499+ 0.187 0.42 32.153 0.001 2.197 0.995
Online party 0.021 0.221 0.49* 16.148 0.095 11.559 0.316
Facebook 0.0260.0450.10 13.3530.205 7.489 0.679
Myspace -0.0160.071 0.16 10.400 0.406 13.120 0.217
Drinking Behavior Dynamics
Rate-drinking 2.165***0.1320.30 282.3090.001 <.001 1
Linear shape -0.351** 0.052 0.12 0.007 1 111.897 0.001
Quadratic shape 0.108* 0.029 0.07 41.842 0.001 2.457 0.991
Average alter drinking 0.199+ 0.085 0.19 19.080 0.039 2.761 0.987
Female 0.0150.0880.2011.7060.305 8.547 0.576
Grades -0.0120.0090.02 5.4050.863 9.768 0.461
Parent smoking 0.140+ 0.063 0.14 26.975 0.003 2.825 0.985
Hispanic 0.0050.0470.11 8.3480.595 6.745 0.749
Online risk pictures -0.002 0.058 0.13 7.083 0.718 8.363 0.593
Online party 0.142 0.090 0.20 18.863 0.042 4.819 0.903
Facebook 0.0170.0210.05 13.1010.218 5.673 0.842
Myspace 0.0330.0350.08 15.9890.100 6.621 0.761
***p<.001, **<.01, *<.05 (two-sided tests)
sd=between school standard deviation, significance based on appr
Note: bolded values indicate significant hypothesis test outcomes
oximate χ
2
test (df=4)
91
Appendix A. Studies on Peer Selection and Influence Using Stochastic Actor-based
M de o ling
Topic Sample Outcomes Reference
1

Smoking an
Alcohol
d Finnish 16 y/o; 9 schools;
N=1419
Both for alcohol; only
selection for tobacco
(Kiuru et
2010)
al.,
3 Substance
use
Teenage Friends and Lifestyle
3y/o Study (Scotland); 3 yrs; 1
Selection for both;
lcohol Influence for a
only
(Pearson,
, Sweeting, et al.
2006)
4 Alcohol Add Health; N=2563; 13
schools; 7-11 grade
Selection only (Mundt, 2013)
5 Alcohol Dutch sample of N=3041; 10-
15yo; 120 classrooms in 14
schools
Selection stronger (Knecht et al.,
2011)
6 Alcohol Teenage Friends and Lifestyle
Study; N=160; 13yo
Selection and
Influence
(Steglich et al.,
2006)
7 Smoking/
parenting
Belgian school; N=254; 14-16
yo;
Selection effect from
rnal smoking mate
beh.
(Mercken et al.,
2013)
8 Smoking Ad Health-2 schools with high
and low prevalence. 10-11
graders; N=1612;
Both (Green et al.,
2013)
9 Smoking Ad Health; N=509; 15 yo
school
;1 Both (Schaefer et al.,
2012)
10 Smoking British ASSIST study;11
schools; N=1716; 12 yo
Selection stronger (Mercken e
2012)
t al.,
11 Smoking 11 Finnish high schools;
N=1326
Selection stronger,
esp. in non-
reciprocated
friendships
(Mercken,
Snijders, Steglich,
et al., Vartiainen,
2010)
12 Smoking ESFA study; N=1163 from
control group; 9 schools;
N=13.6;
Selection stronger;
influence only among
girls
(Mercken,
Snijders, Steglich,
et al., Vertiainen,
2010)
13 Smoking ESFA study; N=7704; 6
European countries
Selection stronger (Mercken,
Snijders, Steglich
& de Vries, 2009)

92
CHAPTER 5 — Discussion and Conclusions

Summary of Overall Findings
The overall goal of this dissertation study was to delineate the associations between
two domains of social influences—the effects of their friends’ behavior and the effects of
the friends’ online behaviors—on adolescent smoking and alcohol use behaviors. Social
network analysis techniques were applied across the papers to generate measures of social
influence and social media influence in the context of existing friendships at the community
level and within their grade level at school. The analyses provide evidence on the less
studied effects of social network site use on adolescents and provide mechanisms by which
both sources of influence impact risk-taking behavior.
Study One took a multivariate approach to assess the impact of several measures
relating to adolescent use of social network sites (SNS). These included the frequency of
visits to specific SNSs (Facebook and Myspace), number for friends who use of these SNSs,
as well as specific exposures to risky content displayed by these named friends. Results
demonstrated that SNSs differed in function and effects on adolescent risk behaviors.
While no specific effects were attributed to the use of any specific SNS, the collective
display of risky content by friends was positively associated with higher levels of smoking
and alcohol use. The paper brings to light the significance of friends’ online risky
behaviors, and also confirmed the findings from previous studies regarding the strong
association between friend and adolescent risk behaviors (Hoffman et al., 2007; Simons-
Morton & Farhat, 2010).
93
As an extension to the findings from the first paper, Study Two was based on a
structural approach to examine the specific pathways of influence using a cognitive-
behavioral framework to test whether social norms and behavioral intentions mediated the
effects between social influences (in-person and digital) on behavior. A comparison of the
relative correlations between in-person and digital risk exposures revealed that having
friends who smoke or use alcohol and having friends who post content about these risk-
behaviors online were both positively associated with intentions and longitudinally
associated with behavior.
In accordance with the Theory of Reasoned Action, intentions were found to
mediate the relationship between in-person influences, but had a stronger direct effect on
both smoking and alcohol use. This has been found in other studies (Gibbons et al., 2009),
and lends further support for a dual-processing model, which posits that behaviors are in
certain situations preceded by rational intentions to perform the behavior, but can also
occur reactively without conscious processing or forethought.
Study Three takes an in-depth longitudinal approach to examine whether
associations between social influences and risk behaviors were attributed to actual
influences due to behaviors exhibited by friends, or whether the association was attributed
to friendship selection. Stochastic actor-based (SAB) modeling was followed by meta-
analyses of the five school-level networks. Findings indicated that friendship selection
effects were stronger than peer influence effects for both smoking and alcohol use, while
influence effects were comparatively more prominent for alcohol use. Secondly, digital
influence effects (exposure to friends’ risky online displays) were added to form a social
media model. This revealed that adolescent friendships were formed based on similar use
94
of SNSs Facebook and Myspace. Furthermore, exposure to friends’ risky online content
contributed to increases in drinking and smoking. This study also revealed that significant
changes in friendship ties and risk behaviors took place between the two observation
periods despite the appearance that prevalence did not change.
In sum, the three studies, using three different statistical methodologies may be
considered a “first look” at the interplay between online social networks, traditional in-
person peer friendship networks and their collective effects on adolescent smoking and
alcohol use.
Implications for Future Research
Given the dominance of social media use among adolescents, continued research is
necessary to provide a more comprehensive picture of the various components that make
up what this study has termed “risky digital influences,” and how an adolescent’s use of
these new media channels can affect their risk taking behaviors. Based on the study
findings, there are two main areas that would benefit from further research—improved
measures and a better understanding of the underlying mechanisms that drive digital
social influence effects.
In the three studies, online influences were operationalized as adolescent self-
reports of their friends’ risky online activities. However, social media effects can be much
broader. Future studies may consider more specific and direct measures such as the types
of online activity that adolescents engaged in (blog posts, photo-sharing, video-sharing,
gaming). Adolescents might also be surveyed about the type of content they post (pictures,
videos, status updates) and their tendencies to post risky content. Along with bounded
95
network data, as collected in this study, these measures can then be used to calculate
adolescents’ net exposure to their friends’ online activities.
Similarly, while this study focused on digital influences contained among existing
friends, influences from general online or Web 2.0 channels extend far beyond those
exhibited by friends. Future studies may wish to also examine the broader impact of the
Internet on the formation of social norms toward risk behaviors. The use and validation of
multiple measures would be beneficial especially in the formative stages of research in this
rea. a
Secondly, a better theoretical understanding of key influence pathways is necessary
to inform the development of effective intervention strategies to curtail the harmful effects
of social media exposure, as well as to harness its potential for health promotion. As an
extension to the current findings, one would hope to re-assess the effects of subjective
norms as a mediator of friend risk behaviors and risk outcomes. The relative effects
between digital and in-person social influences on descriptive and injunctive norms as well
potential feedback loops between the mediators, outcomes and predictors would be
informative for the design of new digital intervention tools. A further examination of the
dual-processing model as it relates to in-person and digital influences, may also lend
support to intervention approaches through mass-interpersonal communication networks,
by teasing out the similarities and differences between the two domains of influence and
associated processes.
Potential moderator effects might also be tested to determine whether the found
associations might vary by gender, racial/ethnic background, or other factors such as
96
friendship quality or social capital, or inform audience segmentation strategies and use of
more targeted intervention messages.
Lastly, in-person focus groups and in-depth interviews with adolescents and their
parents may provide valuable context in the formative stages of measurement
development or intervention design. Together, the findings here and directions for future
research are likely to contribute to a fuller understanding of online social influence
mechanisms for the adolescent population who are most at risk for experimenting or
adopting health compromising behaviors due to peer influences.
Implications for Future Practice
Findings from the three studies as well as future studies in this area are critical for
informing more theoretically driven social media based health promotion strategies (Chou
et al., 2013). Given that in-person social influences may be compounded by digital sources
of social influence and the importance placed on social acceptance during adolescence,
interventions may wish to address the natural discrepancies between online impressions
and reality (Strasburger, 2012).
A related area that warrants more attention is media literacy education, which has
been shown to be protective against adolescent substance use intentions and behaviors
(Primack & Hobbs, 2009). Of primary interest in relation to curtailing the negative effects
of broader social media influences is to teach adolescents how to critically examine and
distinguish between the representation-reality continuum of online presentations
(Primack & Hobbs, 2009). These lessons could be potentially adapted for addressing social
media literacy and could be incorporated into school-based or clinic/physician-based
97
interventions (Strasburger, 2012). A vibrant online social network presence may also be
an effective means to disseminate these messages broadly and efficiently.
Defined by Fogg (2008) as Mass Interpersonal Persuasion (MIP), this model
describes how individuals now have the tools to create and distribute persuasive messages
to the masses. The amount of online social network activity among adolescents today
present a potentially powerful mechanism for peer social influences to take place at an
exponential rate. Social network methods could be used to identify the central opinion
leaders in a school network to serve as role models or the identified to distribute the health
messages. Interventions should be designed to harness the innate ability of social media to
connect individuals and foster healthy dialogue and behaviors. School-based interventions
may consider developing pro-social group activities that foster self-reinforcing friendships
that foster more constructive activities.
Going forward with the task to develop a comprehensive strategy for addressing the
risks and benefits of adolescent use of social media, public health researchers should adopt
a transdisciplinary and shared perspective across different fields of research and practice.
With social media transecting almost all fields of study, this will be the key to making most
out of the tools that are currently at hand.

98
REFERENCES

Ahn, J. (2012). Teenagers and social network sites: Do off-line inequalities predict their
online social networks? First Monday, 17, 1-2.
Alexander, C., Piazza, M., Mekos, D. & Valente, T. (2001). Peers, schools, and adolescent
cigarette smoking. J Adolesc Health, 29(1), 22-30.
Avenevoli, S. & Merikangas, K. R. (2003). Familial influences on adolescent smoking.
Addiction, 98 Suppl(1), 1-20.
Bachman, J. G., O'Malley, P. M., Johnston, L. D., Schulenberg, J. E. & Wallace, J. M. (2011).
Racial/ethnic differences in the relationship between parental education and
substance use among U.S. 8th-, 10th-, and 12th-grade students: findings from the
Monitoring the Future project. J Stud Alcohol Drugs, 72(2), 279-285.
Backinger, C. L. & Augustson, E. M. (2011). Where There's an App, There's a Way? Am J Prev
Med, 40(3), 390-391. doi: 10.1016/j.amepre.2010.11.014.
Backstrom, L. (2011). Anatomy of Facebook. Retrieved July 12, 2012, from
facebook.com/note.ph 3859. http://www. p?note_id=1015038851924
e Hall. Bandura, A. (1977). Social Learning Theory. New Jersey: Prentic
Bandura, A. (1978). The self system in reciprocal determinism. American psychologist,
33(4), 344-358.
Bandura, A. (2002). Growing primacy of human agency in adaptation and change in the
electronic era. European Psychologist, 7(1), 2-16.
boyd, d. (2007). Why youth ♥ social network sites: The role of networked publics in teenage
social life. In D. Buckingham(Ed.). MacArthur Foundation Series on Digital Media and
Learning-Youth Identity, and Digital Media Volume (pp. 119-142). Cambridge, MA:
MIT Press.
boyd, d. (2011). White flight in networked publics? How race and class shaped American
teen engagement with MySpace and Facebook. In L. Nakamura and P. Chow-White
ce After the Internet (pp. 203-222). New York: Routledge. (Eds.). Ra
boyd, d. (2008). Taken out of context: American teen sociality in networked publics. PhD
rmation. Dissertation. University of California-Berkeley, School of Info
Bureau, U. S. C. (2010). American Fact Finder. Retrieved July, 2012
http://factfinder2.census.gov/faces/nav/jsf/pages/index.xhtml.
99
Centers for Disease Control & Prevention. (2010). Cigarette use among high school
students - United States, 1991-2009 MMWR Morb Mortal Wkly Rep (2010/07/09 ed.,
Vol. 59, pp. 797-801).
Champion, V. L. & Skinner, C. S. (2008). The health belief model. Health behavior and health
education: Theory, research, and practice, 4, 45-65.
Chou, W.Y. S., Prestin, A., Lyons, C. & Wen, K.Y. (2013). Web 2.0 for Health Promotion:
Reviewing the Current Evidence. Am J Public Health, 103(1), e9-e18.
Christofides, E., Muise, A. & Desmarais, S. (2009). Information disclosure and control on
Facebook: Are they two sides of the same coin or two different processes?
CyberPsychology & Behavior, 12(3), 341-345.
City of El Monte. Accessed July, 2012: http://www.ci.el-monte.ca.us.
Cleveland, M. J., Feinberg, M. E., Bontempo, D. E. & Greenberg, M. T. (2008). The role of risk
and protective factors in substance use across adolescence. Journal of Adolescent
Health, 43(2), 157-164.
Cruz, J. E., Emery, R. E. & Turkheimer, E. (2012). Peer network drinking predicts increased
alcohol use from adolescence to early adulthood after controlling for genetic and
shared environmental selection. Dev Psychol, 48(5), 1390-1402. doi:
10.1037/a0027515.
Dal Cin, S., Worth, K. A., Gerrard, M., Gibbons, F. X., Stoolmiller, M., Wills, T. A. & Sargent, J. D.
(2009). Watching and drinking: Expectancies, prototypes, and peer affiliations
mediate the effect of exposure to alcohol use in movies on adolescent drinking.
Health psychology: official journal of the Division of Health Psychology, American
Psychological Association, 28(4), 473-483.
El Monte Union High School District. Accessed July, 2012:
http://emuhsd.org/index.php/district/about_the_district.
Ennett, S. T. & Bauman, K. E. (1993). Peer group structure and adolescent cigarette
smoking: a social network analysis. J Health Soc Behav, 34(3), 226-236.
Ennett, S. T. & Bauman, K. E. (1994). The contribution of influence and selection to
adolescent peer group homogeneity: the case of adolescent cigarette smoking. J Pers
Soc Psychol, 67(4), 653-663.
Ennett, S. T., Bauman, K. E., Hussong, A., Faris, R., Foshee, V. A., Cai, L. & DuRant, R. H.
(2006). The peer context of adolescent substance use: Findings from social network
analysis. Journal of Research on Adolescence, 16(2), 159-186.
100
Ennett, S. T., Faris, R., Hipp, J., Foshee, V. A., Bauman, K. E., Hussong, A. & Cai, L. (2008). Peer
smoking, other peer attributes, and adolescent cigarette smoking: A social network
analysis. Prev Sci, 9(2), 88-98. doi: 10.1007/s11121-008-0087-8.
Festl, R. & Quandt, T. (2013). Social relations and cyberbullying: The influence of individual
and structural attributes on victimization and perpetration via the Internet. Human
101-126. communication research, 39(1),
Fishbein, M. & Ajzen, I. (Eds.). (1975). Belief, attitude, intention, and behavior: An
introduction to theory and research. MA: Addison-Wesely.
Flay, B. R. (2009). The promise of long-term effectiveness of school-based smoking
prevention programs: a critical review of reviews. Tob Induc Dis, 5(1), 7. doi:
10.1186/1617-9625-5-7.
Fogg, B. (2008). Mass interpersonal persuasion: An early view of a new phenomenon.
Persuasive Technology, 23-34.
Freeman, B. (2012). New media and tobacco control. Tob Control, 21(2), 139-144. doi:
10.1136/tobaccocontrol-2011-050193.
Fujimoto, K. & Valente, T. W. (2012). Decomposing the components of friendship and
friends' influence on adolescent drinking and smoking. Journal of Adolescent Health,
52(2), 136-143.
Gibbons, F. X., Houlihan, A. E. & Gerrard, M. (2009). Reason and reaction: the utility of a
dual-focus, dual-processing perspective on promotion and prevention of adolescent
health risk behaviour. Br J Health Psychol, 14(Pt 2), 231-248. doi:
10.1348/135910708X376640.
Giles, D. C. (2002). Parasocial interaction: A review of the literature and a model for future
research. Media psychology, 4(3), 279-305.
, 257-281. Giordano, P. C. (2003). Relationships in adolescence. Annual review of sociology
Granovetter, M. (1983). The strength of weak ties: A network theory revisited. Sociological
theory, 1(1), 201-233.
Green, H. D., Jr., Horta, M., de la Haye, K., Tucker, J. S., Kennedy, D. R. & Pollard, M. (2013).
Peer influence and selection processes in adolescent smoking behavior: a
comparative study. Nicotine Tob Res, 15(2), 534-541. doi: 10.1093/ntr/nts191.
Grieve, R., Indian, M., Witteveen, K., Anne Tolan, G. & Marrington, J. (2013). Face-to-face or
Facebook: Can social connectedness be derived online? Computers in Human
Behavior, 29(3), 604-609.
101
Hall, J. A. & Valente, T. W. (2007). Adolescent smoking networks: the effects of influence
and selection on future smoking. Addict Behav, 32(12), 3054-3059. doi:
10.1016/j.addbeh.2007.04.008.
Hargittai, E. (2007). Whose space? Differences among users and non-users of social
network sites. Journal of Computer-Mediated Communication, 13(1), 276-297.
Hedges, L. V., Olkin, I., Statistiker, M., Olkin, I. & Olkin, I. (1985). Statistical methods for
meta-analysis: Academic Press New York.
Hill, K. G., Hawkins, J. D., Catalano, R. F., Abbott, R. D. & Guo, J. (2005). Family influences on
the risk of daily smoking initiation. Journal of Adolescent Health, 37(3), 202-210.
Hinduja, S. & Patchin, J. W. (2008). Personal information of adolescents on the Internet: A
quantitative content analysis of MySpace. J Adolesc, 31(1), 125-146. doi:
10.1016/j.adolescence.2007.05.004.
Hoffman, B. R., Monge, P. R., Chou, C. P. & Valente, T. W. (2007). Perceived peer influence
and peer selection on adolescent smoking. Addict Behav, 32(8), 1546-1554. doi:
10.1016/j.addbeh.2006.11.016.
Hogg, M. A., Abrams, D., Otten, S. & Hinkle, S. (2004). The social identity perspective. Small
group research, 35(3), 246-276.
Huang, G. C., Unger, J. B., Soto, D., Fujimoto, K., Pentz, M. A., Jordan-Marsh, M. & Valente, T.
W. (under review). Peer influences: The impact of online and offline friendship
lth. networks on adolescent smoking and alcohol use. Journal of Adolescent Hea
Itō, M., Baumer, S., Bittanti, M., boyd, d., Cody, R., Herr, B., Horst, H. A., Lange, P. G.,
Mahendran, D., Martinez, K., Pascoe, C. J., Perkel, D., Robinson, L., Sims, C., and Tripp,
L. (2009). Hanging out, messing around, and geeking out: Kids living and learning
with new media. Cambridge: The MIT Press.
Johnston, L. D., O'Malley, P. M., Bachman, J. G. & Schulenberg, J. E. (2012). Monitoring the
Future: National results on adolescent drug use. Overview of key findings, 2011. Ann
Arbor: Institute for Social Research, The University of Michigan.
Kiuru, N., Burk, W. J., Laursen, B., Salmela-Aro, K. & Nurmi, J. E. (2010). Pressure to drink
but not to smoke: Disentangling selection and socialization in adolescent peer
d peer groups. J Adolesc, 33(6), 801-812. networks an
Kline, R. B. (2010). Principles and practice of structural equation modeling—Third Edition.
New York: The Guilford Press.
Knecht, A. B., Burk, W. J., Weesie, J. & Steglich, C. (2011). Friendship and alcohol use in early
adolescence: a multilevel social network approach. Journal of Research on
Adolescence, 21(2), 475-487.
102
friendships. Addiction, 102(9), 1483-1492. doi: 10.1111/j.1360-0443.2007.0190
Mercken, L., Candel, M., Willems, P. & de Vries, H. (2009). Social influence and selection
effects in the context of smoking behavior: changes during early and mid
adolescence. Health Psychol, 28(1), 73-82. doi: 10.1037/a0012791.
Kobus, K. (2003). Peers and adolescent smoking. Addiction, 98 Suppl 1, 37-55.
Krishna, S., Boren, S. A. & Balas, E. A. (2009). Healthcare via cell phones: a systematic
review. Telemedicine and E-Health, 15(3), 231-240.
Leatherdale, S. T. & Ahmed, R. (2010). Alcohol, marijuana, and tobacco use among Canadian
youth: do we need more multi-substance prevention programming? J Prim Prev,
31(3), 99-108. doi: 10.1007/s10935-010-0211-y.
oject. Lenhart, A. (2012). Teens, Smartphones & Texting. Pew Internet & American Life Pr
Lenhart, A., Madden, M., Smith, A., Purcell, K., Zickuhr, K. & Rainie, L. (2011). Teens,
t. kindness and cruelty on social network sites. Pew Internet & American Life Projec
Lenhart, A., Purcell, K., Smith, A. & Zickuhr, K. (2010). Social media & mobile internet use
among teens and young adults. Pew Internet & American Life Project.
Livingstone, S. (2008). Taking risky opportunities in youthful content creation: teenagers'
use of social networking sites for intimacy, privacy and self-expression. New Media
& Society, 10(3), 393-411.
Livingstone, S. & Brake, D. R. (2010). On the rapid rise of social networking sites: New
findings and policy implications. Children & society, 24(1), 75-83.
Maassen, G. H. & Bakker, A. B. (2001). Suppressor variables in path models definitions and
interpretations. Sociological Methods & Research, 30(2), 241-270.
Manning, M. (2009). The effects of subjective norms on behaviour in the theory of planned
ology, 48(4), 64 behaviour: A meta-analysis. British Journal of Social Psych 9-705.
Marsden, P. V. & Campbell, K. E. (1984). Measuring tie strength. Social forces, 63(2), 482-
501.
Mathur, C., Erickson, D. J., Stigler, M. H., Forster, J. L. & Finnegan, J. R., Jr. (2013). Individual
and neighborhood socioeconomic status effects on adolescent smoking: a multilevel
cohort-sequential latent growth analysis. Am J Public Health, 103(3), 543-548. doi:
10.2105/AJPH.2012.300830.
McPherson, M., Smith-Lovin, L. & Cook, J. M. (2001). Birds of a feather: Homophily in social
networks. Annual review of sociology, 415-444.
Mercken, L., Candel, M., Willems, P. & de Vries, H. (2007). Disentangling social selection and
social influence effects on adolescent smoking: the importance of reciprocity in
5.x.
103
Mercken, L., Sleddens, E. F., de Vries, H. & Steglich, C. E. (2013). Choosing adolescent
smokers as friends: The role of parenting and parental smoking. J Adolesc, 36(2),
383-392. doi: 10.1016/j.adolescence.2012.12.004.
Mercken, L., Snijders, T. A., Steglich, C. & de Vries, H. (2009). Dynamics of adolescent
friendship networks and smoking behavior: social network analyses in six European
countries. Soc Sci Med, 69(10), 1506-1514. doi: 10.1016/j.socscimed.2009.08.003.
Mercken, L., Snijders, T. A., Steglich, C., Vertiainen, E. & de Vries, H. (2010). Smoking-based
selection and influence in gender-segregated friendship networks: a social network
analysis of adolescent smoking. Addiction, 105(7), 1280-1289. doi: 10.1111/j.1360-
0443.2010.02930.x.
Mercken, L., Snijders, T. A. B., Steglich, C., Vartiainen, E. & de Vries, H. (2010). Dynamics of
adolescent friendship networks and smoking behavior. Social Networks, 32(1), 72-
81. doi: 10.1016/j.socnet.2009.02.005.
Mercken, L., Steglich, C., Sinclair, P., Holliday, J. & Moore, L. (2012). A longitudinal social
network analysis of peer influence, peer selection, and smoking behavior among
adolescents in British schools. Health Psychology, 31(4), 450-495.
Mesch, G. S. & Talmud, I. (2007). Similarity and the quality of online and offline social
relationships among adolescents in Israel. Journal of Research on Adolescence, 17(2),
455-465.
Mitchell, K. J. & Ybarra, M. (2009). Social networking sites: finding a balance between their
risks and benefits. Archives of Pediatrics and Adolescent Medicine, 163(1), 87.
Montano, D. E. & Kasprzyk, D. (2008). Theory of reasoned action, theory of planned
behavior, and the integrated behavioral model. Health behavior and health
education: Theory, research, and practice. San Francisco, CA: Jossey-Bass, 67-92.
Moreno, M. A. (2011). Youth and social media: Social networking sites and adolescent
health: New opportunities and new challenges. ISJLP, 7, 57-135.
Moreno, M. A., Briner, L. R., Williams, A., Brockman, L., Walker, L. & Christakis, D. A. (2010).
A content analysis of displayed alcohol references on a social networking web site.
1. Journal of Adolescent Health, 47(2), 168-175. doi: 10.1016/j.jadohealth.2010.01.00
Moreno, M. A., Brockman, L., Rogers, C. B. & Christakis, D. A. (2010). An evaluation of the
distribution of sexual references among "top 8" My Space friends. Journal of
. Adolescent Health, 47(4), 418-420. doi: 10.1016/j.jadohealth.2010.02.015
Moreno, M. A., Christakis, D. A., Egan, K. G., Brockman, L. N. & Becker, T. (2012).
Associations between displayed alcohol references on Facebook and problem
drinking among college students. Arch Pediatr Adolesc Med, 166(2), 157-163. doi:
10.1001/archpediatrics.2011.180.
104
Moreno, M. A., Egan, K. G. & Brockman, L. (2011). Development of a researcher codebook
for use in evaluating social networking site profiles. Journal of Adolescent Health,
49(1), 29-35. doi: 10.1016/j.jadohealth.2011.04.015.
Moreno, M. A., Jelenchick, L. A., Egan, K. G., Cox, E., Young, H., Gannon, K. E. & Becker, T.
(2011). Feeling bad on Facebook: Depression disclosures by college students on a
social networking site. Depression and Anxiety, 28(6), 447-455. doi:
10.1002/da.20805.
Moreno, M. A., Parks, M. R., Zimmerman, F. J., Brito, T. E. & Christakis, D. A. (2009). Display
of health risk behaviors on MySpace by adolescents: prevalence and associations.
Arch Pediatr Adolesc Med, 163(1), 27-34. doi: 10.1001/archpediatrics.2008.528.
Mundt, M. (2013). Social Network Analysis of Peer Effects on Binge Drinking among U.S.
Adolescents. In A. Greenberg, W. Kennedy & N. Bos (Eds.), Social Computing,
Behavioral-Cultural Modeling and Prediction (Vol. 7812, pp. 123-134). Berlin,
Heidelberg: Springer.
Muthén, L. K. & Muthén, B. O. (2012). Mplus User’s Guide. Seventh Edition. Los Angeles, CA:
Muthén & Muthén.
Myers, D., Baer, W. C. & Choi, S.-Y. (1996). The changing problem of overcrowded housing.
Journal of the American Planning Association, 62(1), 66-84.
Nash, S. G., McQueen, A. & Bray, J. H. (2005). Pathways to adolescent alcohol use: family
environment, peer influence, and parental expectations. J Adolesc Health, 37(1), 19-
16/j.jadoheal 28. doi: 10.10 th.2004.06.004.
ss. Norris, J. R. (1998). Markov chains. New York, New York: Cambridge University Pre
O'Keeffe, G. S. & Clarke-Pearson, K. (2011). The impact of social media on children,
adolescents, and families. Pediatrics, 127(4), 800-804.
Oetting, E. R. & Donnermeyer, J. F. (1998). Primary socialization theory: The etiology of
drug use and deviance. I. Subst Use Misuse, 33(4), 995-1026.
Olds, R. S., Thombs, D. L. & Tomasek, J. R. (2005). Relations between normative beliefs and
initiation intentions toward cigarette, alcohol and marijuana. J Adolesc Health, 37(1),
75. doi: 10.1016/j.jadohealth.2004.09.020.
Patchin, J. W. & Hinduja, S. (2010). Trends in online social networking: adolescent use of
MySpace over time. New Media & Society, 12(2), 197-216.
Pearson, M., Steglich, C. & Snijders, T. (2006). Homophily and assimilation among sport-
active adolescent substance users. Connections, 27(1), 47-63.
105
10.2105/AJPH.2012.300705.
Selwyn, N. (2009). Faceworking: exploring students' education-related use of Facebook.
Learning, Media and Technology, 34(2), 157-174.
Pearson, M., Sweeting, H., West, P., Young, R., Gordon, J. & Turner, K. (2006). Adolescent
substance use in different social and peer contexts: A social network analysis. Drugs:
Education, Prevention, and Policy, 13(6), 519-536.
Primack, B. A. & Hobbs, R. (2009). Association of various components of media literacy and
adolescent smoking. Am J Health Behav, 33(2), 192-201.
R Development Core Team. (2011). R: A language and environment for statistical
computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from
http://www.R-project.org.
Reich, S. M., Subrahmanyam, K. & Espinoza, G. (2012). Friending, IMing, and hanging out
face-to-face: Overlap in adolescents' online and offline social networks. Dev Psychol,
48(2), 356-368.
Reid, J., Manske, S. & Leatherdale, S. (2008). Factors related to adolescents' estimation of
), 81-93. peer smoking prevalence. Health Education Research, 23(1
Rice, E. & Karnik, N. S. (2012). Network science and social media. J Am Acad Child Adolesc
Psychiatry, 51(6), 563-565. doi: 10.1016/j.jaac.2012.02.008.
Rice, E., Monro, W., Barman-Adhikari, A. & Young, S. D. (2010). Internet use, social
networking, and HIV/AIDS risk for homeless adolescents. Journal of Adolescent
Health, 47(6), 610-613.
Rimal, R. N. & Real, K. (2003). Understanding the influence of perceived norms on
behaviors. Communication Theory, 13(2), 184-203.
Rimal, R. N. & Real, K. (2005). How Behaviors are Influenced by perceived norms: A test of
-414. the theory of normative social behavior. Communication Research, 32(3), 389
Ripley, R. M., Snijders, T. A. B. & Preciado, P. (2013). Manual for RSiena. Oxford, UK:
University of Oxford, Department of Statistics, Nuffield College.
Robards, B. (2012). Leaving MySpace, joining Facebook:‘Growing up’on social network
sites. Continuum, 26(3), 385-398.
Sargent, J. D., Worth, K. A., Beach, M., Gerrard, M. & Heatherton, T. F. (2008). Population-
based assessment of exposure to risk behaviors in motion pictures. Communication
Methods and Measures, 2(1-2), 134-151.
Schaefer, D. R., Haas, S. A. & Bishop, N. J. (2012). A dynamic model of US adolescents'
smoking and friendship networks. Am J Public Health, 102(6), e12-18. doi:
106
doi: 10.1542/peds.2011-3213.
Strasburger, V. C., Jordan, A. B. & Donnerstein, E. (2010). Health effects of media on children
and adolescents. Pediatrics, 125(4), 756-767. doi: 10.1542/peds.2009-2563.
Simons-Morton, B. G. & Farhat, T. (2010). Recent findings on peer group influences on
. adolescent smoking. J Prim Prev, 31(4), 191-208. doi: 10.1007/s10935-010-0220-x
Slater, M. D. & Henry, K. L. (2013). Prospective influence of music-related media exposure
on adolescent substance-use initiation: a peer group mediation model. J Health
Commun, 18(3), 291-305. doi: 10.1080/10810730.2012.727959.
Snijders, T. A. B. & Baerveldt, C. (2003). A multilevel network study of the effects of
delinquent behavior on friendship evolution. Journal of mathematical sociology,
27(2-3), 123-151.
Snijders, T. A. B., Steglich, C. E. G. & Schweinberger, M. (2007). Modeling the co-evolution of
networks and behavior. Longitudinal models in the behavioral and related sciences,
41-71.
Snijders, T. A. B., Van de Bunt, G. G. & Steglich, C. E. G. (2010). Introduction to stochastic
actor-based models for network dynamics. Social Networks, 32(1), 44-60.
rp LP. StataCorp. (2011). Stata Statistical Software: Release 12. College Station, TX: StataCo
Steglich, C., Snijders, T. A. B. & Pearson, M. (2010). Dynamic networks and behavior:
Separating selection from influence. Sociological Methodology, 40(1), 329-393.
Steglich, C., Snijders, T. A. B. & West, P. (2006). Applying SIENA: An illustrative analysis of
the coevolution of adolescents' friendship networks, taste in music, and alcohol
consumption. Methodology: European Journal of Research Methods for the Behavioral
and Social Sciences, 2(1), 48-56.
Steinberg, L. (2008). A social neuroscience perspective on adolescent risk-taking.
Developmental Review, 28(1), 78-106.
Steinberg, L. & Monahan, K. C. (2007). Age differences in resistance to peer influence. Dev
Psychol, 43(6), 1531-1543. doi: 10.1037/0012-1649.43.6.1531.
Steinfield, C., Ellison, N. B. & Lampe, C. (2008). Social capital, self-esteem, and use of online
social network sites: A longitudinal analysis. Journal of Applied Developmental
Psychology, 29(6), 434-445. doi: 10.1016/j.appdev.2008.07.002.
Stoolmiller, M., Wills, T. A., McClure, A. C., Tanski, S. E., Worth, K. A., Gerrard, M. & Sargent, J.
D. (2012). Comparing media and family predictors of alcohol use: a cohort study of
US adolescents. BMJ Open, 2, e000543. doi: 10.1136/bmjopen-2011-000543.
Strasburger, V. C. (2012). Pediatricians, schools, and media. Pediatrics, 129(6), 1161-1163.
107
Sussman, S., Dent, C. W., Mestel-Rauch, J., Johnson, C. A., Hansen, W. B. & Flay, B. R. (1988).
Adolescent nonsmokers, triers, and regular smokers' estimates of cigarette smoking
prevalence: When do overestimations occur and by whom? Journal of Applied Social
Psychology, 18(7), 537-551.
Sussman, S. & Sun, P. (2009). Youth tobacco use cessation: 2008 update. Tob Induc Dis, 5, 3.
doi: 10.1186/1617-9625-5-3.
Tickle, J. J., Hull, J. G., Sargent, J. D., Dalton, M. A. & Heatherton, T. F. (2006). A structural
equation model of social influences and exposure to media smoking on adolescent
smoking. Basic and Applied Social Psychology, 28(2), 117-129.
Trucco, E. M., Colder, C. R. & Wieczorek, W. F. (2011). Vulnerability to peer influence: a
moderated mediation study of early adolescent alcohol use initiation. Addict Behav,
36(7), 729-736. doi: 10.1016/j.addbeh.2011.02.008.
Unger, J. B. & Rohrbach, L. A. (2002). Why do adolescents overestimate their peers'
smoking prevalence? Correlates of prevalence estimates among California 8th-grade
students. J Youth Adolesc, 31(2), 147-153.
Urberg, K., Goldstein, M. S. & Toro, P. A. (2005). Supportive relationships as a moderator of
the effects of parent and peer drinking on adolescent drinking. Journal of Research
on Adolescence, 15(1), 1-19.
Urberg, K. A., Degirmencioglu, S. M. & Pilgrim, C. (1997). Close friend and group influence
-844. on adolescent cigarette smoking and alcohol use. Dev Psychol, 33(5), 834
U.S. Department of Health and Human Services, Centers for Disease Control and
Prevention, National Center for Chronic Disease Prevention and Health Promotion,
Office on Smoking and Health. Preventing tobacco use among youth and young adults:
A report of the Surgeon General. (2012). Atlanta, GA: Retrieved from
http://www.surgeongeneral.gov/library.
Utz, S. (2010). Show me your friends and I will tell you what type of person you are: How
one's profile, number of friends, and type of friends influence impression formation
on social network sites. Journal of Computer-Mediated Communication, 15(2), 314-
335.
Valente, T. W. (1996). Social network thresholds in the diffusion of innovations. Social
, 69-89. Networks, 18(1)
Valente, T. W. (2010). Social networks and health: Models, methods, and applications. New
York: Oxford Univ Press.
Valente, T. W., Fujimoto, K., Soto, D., Ritt-Olson, A. & Unger, J. B. (2013). A comparison of
peer influence measures as predictors of smoking among predominately
108
hispanic/latino high school adolescents. J Adolesc Health, 52(3), 358-364. doi:
10.1016/j.jadohealth.2012.06.014.
Valente, T. W., Fujimoto, K., Unger, J. B., Soto, D. W. & Meeker, D. (2013). Variations in
network boundary and type: A study of adolescent peer influences. Social Networks,
35, 309-316. doi: 10.1016/j.socnet.2013.02.008.
Valente, T. W., Ritt-Olson, A., Stacy, A., Unger, J. B., Okamoto, J. & Sussman, S. (2007). Peer
acceleration: effects of a social network tailored substance abuse prevention
program among high-risk adolescents. Addiction, 102(11), 1804-1815. doi:
10.1111/j.1360-0443.2007.01992.x.
Valente, T. W., Unger, J. B. & Johnson, C. A. (2005). Do popular students smoke? The
association between popularity and smoking among middle school students. J
016. Adolesc Health, 37(4), 323-329. doi: 10.1016/j.jadohealth.2004.10.
U.S. Census Bureau. (2010). American Fact Finder. Accessed July, 2012:
http://factfinder2.census.gov/faces/nav/jsf/pages/index.xhtml.
Vitoria, P. D., Salgueiro, M. F., Silva, S. A. & de Vries, H. (2011). Social influence, intention to
smoke, and adolescent smoking behaviour longitudinal relations. Br J Health
Psychol, 16(4), 779-798. doi: 10.1111/j.2044-8287.2010.02014.x.
Walther, J. B., Van Der Heide, B., Kim, S. Y., Westerman, D. & Tong, S. T. (2008). The role of
friends’ appearance and behavior on evaluations of individuals on Facebook: Are we
9. known by the company we keep? Human communication research, 34(1), 28-4
Wills, T. A. & Cleary, S. D. (1999). Peer and adolescent substance use among 6th-9th
graders: latent growth analyses of influence versus selection mechanisms. Health
Psychol, 18(5), 453-463.
Wills, T. A., Sargent, J. D., Stoolmiller, M., Gibbons, F. X. & Gerrard, M. (2008). Movie smoking
exposure and smoking onset: a longitudinal study of mediation processes in a
representative sample of US adolescents. Psychology of addictive behaviors: journal
. of the Society of Psychologists in Addictive Behaviors, 22(2), 269
Windle, M. (2003). Alcohol use among adolescents and young adults. Alcohol Res. Health,
27. Accessed online: http://pubs.niaaa.nih.gov/publications/arh27-1/79-86.htm.
Ybarra, M., Mitchell, K. & Espelage, D. (2012). Comparisons of bully and unwanted sexual
experiences online and offline among a national sample of youth. In Ö. Özdemir
(Ed.) Complementary Pediatrics (pp. 203-216). doi: 10.5772/33532.
Ybarra, M. L., Alexander, C. & Mitchell, K. J. (2005). Depressive symptomatology, youth
Internet use, and online interactions: A national survey. J Adolesc Health, 36(1), 9-
18. doi: 10.1016/j.jadohealth.2003.10.012.
109
Ybarra, M. L. & Mitchell, K. J. (2008). How risky are social networking sites? A comparison
of places online where youth sexual solicitation and harassment occurs. Pediatrics,
121(2), e350-e357.
Ybarra, M. L., Mitchell, K. J., Finkelhor, D. & Wolak, J. (2007). Internet prevention messages:
targeting the right online behaviors. Arch Pediatr Adolesc Med, 161(2), 138-145. doi:
10.1001/archpedi.161.2.138.
Young, S. D. & Jordan, A. H. (2013). The influence of social networking photos on social
norms and sexual health behaviors. Cyberpsychology, Behavior, and Social
Networking, 16(4), 243-247. doi:10.1089/cyber.2012.0080.
Young, S. D. & Rice, E. (2011). Online social networking technologies, HIV knowledge, and
sexual risk and testing behaviors among homeless youth. AIDS Behav, 15(2), 253-
260. doi: 10.1007/s10461-010-9810-0.

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Adolescent social networks, smoking, and loneliness

PDF

Social network influences on depressive symptoms among Chinese adolescents

PDF

Exploring the role of peer influence, linguistic acculturation, and social networks in substance use

PDF

Contextualizing social network influences on substance use among high risk adolescents

PDF

The role of depression symptoms on social information processing and tobacco use among adolescents

PDF

Social selection and influence of alcohol & marijuana implicit cognitions and behaviors: a longitudinal investigation of peer social network dynamics

PDF

Investigating factors that influence peer relationships and obesity during middle childhood

PDF

Cultural risk and protective factors for tobacco use behaviors and depressive symptoms among American Indian adolescents in California

PDF

Influences of specific environmental domains on childhood obesity and related behaviors

PDF

Evaluating social networks and impact of micro-influencers who promote e-cigarettes on social media

PDF

Social self-control and adolescent substance use

PDF

Friendship network position on adolescent behaviors: an examination of a broker position and the likelihood of alcohol and cigarette use

PDF

Exposure to alcohol advertising on television and alcohol use among young adolescents

PDF

Using a structural model of psychopathology to distinguish relations between shared and specific features of psychopathology, smoking, and underlying mechanisms

PDF

The formation and influence of online health social networks on social support, self-tracking behavior and weight loss outcomes

PDF

Diffusion of a peer‐led suicide preventive intervention in secondary schools: strategies to increase effectiveness of peer‐led interventions

PDF

Risk factors associated with smoking initiation among Chinese adolescents: a matched case-control study

PDF

Anxiety symptoms and nicotine use among adolescents and young adults

PDF

The social groups approach to quitting smoking: An examination of smoking cessation online and offline through the influence of social norms, social identification, social capital and social support

PDF

Psychosocial and behavioral ractors associated with emotional eating in adolescents

Asset Metadata

Creator Huang, Grace C. (author)

Core Title A network analysis of online and offline social influence processes in relation to adolescent smoking and alcohol use

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Preventive Medicine (Health Behavior)

Publication Date 08/10/2013

Defense Date 05/15/2013

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag adolescent health,alcohol use,friendship,OAI-PMH Harvest,peer influence,Smoking,social media,social network analysis

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Valente, Thomas W. (committee chair), Chou, Chih-Ping (committee member), Jordan-Marsh, Maryalice (committee member), Pentz, Mary Ann (committee member), Unger, Jennifer B. (committee member)

Creator Email gracehan@usc.edu,ritgrace@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-320760

Unique identifier UC11293718

Identifier etd-HuangGrace-1996.pdf (filename),usctheses-c3-320760 (legacy record id)

Legacy Identifier etd-HuangGrace-1996.pdf

Dmrecord 320760

Document Type Dissertation

Format application/pdf (imt)

Rights Huang, Grace C.

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA