University of Southern California Dissertations and Theses

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

(USC Thesis Other)

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

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content EXPLORING THE ROLE OF PEER INFLUENCE, LINGUISTIC
ACCULTURATION, AND SOCIAL NETWORKS IN SUBSTANCE USE

by

Raquel Fosados

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PREVENTIVE MEDICINE)

August 2007

Copyright 2007 Raquel Fosados
ii
DEDICATION
I dedicate this work to Michael D. Myers II, his endless love, support, and
encouragement are what brought me to this point. Thank you for encouraging me to
apply to the program, for convincing me that I deserve this, for making several trips
to the biomedical library, for making sure I eat and sleep properly, and for being so
willing to accompany me to several conferences. Mostly, thank you for putting
things into perspective and for making sure I laugh often. Thank you for always
being my center, my shoulder to lean and often cry on, for being my rational thought
in an irrational world. Your patience, love, and understanding are worth more to me
than you will ever know. Thank you for being my biggest cheerleader and for taking
this new road with me. I love you so much. Forever.

“Throughout the centuries there were men who took first steps, down new roads,
armed with nothing but their own vision.” – Ayn Rand

My vision includes you, standing right next to me.

iii
ACKNOWLEDGEMENTS
I wish to thank my committee members (Drs. Lourdes Baezconde-Garbanati,
Chih-Ping Chou, Harry Pachon, and Steve Sussman) and my committee chair (Dr.
Tom Valente) for their encouragement, guidance, advice, and for helping me
establish the foundation upon which I now stand on. Thank you for always listening,
no matter how small. Lourdes, thank you for being such a caring and understanding
person. Chih-Ping, thank you for always knowing the answers to my statistical
questions. Harry, thank you for accepting me as I am, for allowing me to flourish and
express myself, and for understanding why I have chosen the path I have. Steve,
thank you for your quick response to my many questions, and for your understanding
of the external forces that often plague and sometimes attempt to derail this process.
Most especially, thank you for your great sense of humor! Tom, thank you for letting
me spread my wings and allowing me to forge my own path; that has made this
entire process worth its weight in gold. I have learned so much under your guidance,
by your example, and from your words. Thank you so very much.
I also wish to thank Ruth Westphal for her never-ending optimism and
guidance. Thank you Ruth for taking me under your wing long ago, for believing in
me, and teaching by example that “achievement of your happiness is the only moral
purpose of your life, and that happiness, not pain or mindless self-indulgence, is the
proof of your moral integrity, since it is the proof and the result of your loyalty to the
achievement of your values” (by Ayn Rand). To Tom G. Palmer, who will probably
never know how much his words at Rancho Bernardo have and continue to influence
iv
me. To Dr. Charlotte Twight, thank you for planting the seed that is to be my next
adventure. To my scrapping buddies: Camey Cresor, Lisa Fortune, Judy Hernandez,
Monica Hunter, and Shetara Oliver, thank you for your encouragement, being my
escape, my rest and relaxation, and for being as crazy about scrapbooking as I am. I
love you all, thank you for being such great friends! To Alicia Gonzalez, what can I
say other than thank you for everything. We have been through a lot, and these
experiences have shaped our lives in ways we could have never predicted back in
high school. We are so much better for it! Remember to take it one day at a time, and
always do what makes you happy. I love you girl. Thanks for being a great friend!
To my fellow Ph.D. classmates, thank you for the laughs and chats. There is nothing
stronger than common suffering to bond a group of people together! To Marny
Barovich and James Pike, without you I would have been completely lost, thank you.
I also wish to thank members of my family. To Emma Jean “Jeanie” Myers,
thank you for your love and support, and of course, for Michael. To my nieces and
nephews, always remember that “The question isn't who is going to let me; it's who
is going to stop me” (by Ayn Rand). Soar far my little ones; I will be here to make
your landing soft should you stumble. To my brother, Daniel J. Fosados, I love you;
don’t worry, things will get better. To my parents Ignacio and Gabriela Fosados:
Gracias por todo su apoyo, los quiero mucho. Most importantly, thank you for
bringing me to this country, where the American dream has become my reality. This
truly is the greatest country on the planet!
v
TABLE OF CONTENTS
Dedication ii
Acknowledgements iii
List of Tables vi
List of Figures vii
Abstract viii
Chapter 1: Introduction 1

Chapter 2: Linguistic acculturation, peer social influence, and 10
substance use in a sample of high- and low-risk youth

Chapter 3: Exploring the association between linguistic acculturation 35
and substance use among Hispanic continuation high
school youth: Evidence for social influence as a mediator

Chapter 4: The effects of network composition and tie strength on 62
substance use: Applying social network methods to a
sample of high-risk youth

Chapter 5: Conclusion 91

References 96
vi
LIST OF TABLES
1.1. Sample size characteristics 9

2.1 Demographic, language preference, and substance use 22
differences by school type

2.2. Regression weights on lifetime and current substance 25
use variables, stratified by school

2.3 Associations between peer social influence and linguistic 28
acculturation, stratified by school type

2.4. Regression weights on lifetime and current substance use 29
variables, stratified by school type

3.1. Selected sample characteristics by language preference 48

3.2. Regression weights of lifetime and current substance use 52

3.3 Regression weights on several outcomes 54

4.1 Demographics, substance use characteristics, and network 76
composition stratified by ethnicity

4.2 Associations between drug network composition and 80
substance use

4.3 Associations between strength of tie and substance use 82

vii
LIST OF FIGURES
1.1 Theoretical model of substance use 9

2.1 Lifetime and current substance use as a function of 24
linguistic acculturation among Hispanic continuation
high school students in Southern California

2.2 Lifetime and current substance use as a function of 24
linguistic acculturation among Hispanic regular high
school students in Southern California

3.1 Lifetime and current substance use as a function of 53
linguistic acculturation among Hispanic continuation
high school youth in California

viii
ABSTRACT
Students attending alternative/continuation high schools report significantly
higher substance use rates than regular high school students. Peer influence is often
implicated as a significant correlate of substance use, with peer relationships being
one of the primary factors involved in whether or not youth decide to engage in and
maintain these behaviors. This dissertation argues for the need to examine peer social
influence as a potential mediator for the association between acculturation on
substance use, and the need to apply social network methods when investigating peer
influence and substance use.
The first study used a cross-sectional baseline survey of 1720 Hispanic youth
attending regular (N=1138) and alternative/continuation (N=582) high schools
throughout Southern California. Results indicate linguistic acculturation was
associated with increased risk for lifetime and current use of alcohol, cigarettes,
marijuana, and hard drug use among the regular high school sample. Among
alternative/continuation school students, acculturation was associated with risk for
lifetime cigarettes, marijuana, hard drugs and current cigarette and marijuana use.
Results of mediation analyses indicated that peer social influence either partially or
completely mediated the relationship between linguistic acculturation and substance
use.
The second study used cross-sectional data from 714 Hispanic youth
attending alternative/continuation high schools. Results indicated a greater level of
English speaking proficiency (linguistic acculturation) was associated with greater
ix
substance use. Peer social influence was also a significant predictor of substance use.
Mediation analyses revealed peer social influence either partially or completely
mediated the relationship between linguistic acculturation and several substance use
outcomes.
The third study examined associations between network composition (drug
network and strength of tie) and substance use among 968 alternative/continuation
high school youth. Results indicate that the number of substance users in peer
networks (drug network composition) was associated with risk for lifetime use of
alcohol, cigarettes, and marijuana, controlling for several covariates. Strength of tie
(being closer to first or second named peer vs. third-fifth named peer) was also
associated with risk for lifetime cigarette, marijuana, and cocaine use, controlling for
several covariates. These findings suggest that peer influence may be an important
variable influencing the relationship between acculturation and substance use.
1
CHAPTER 1: INTRODUCTION

There is evidence to suggest that acculturation, peer influence, and peer
networks are important with regards to the adoption and maintenance of substance
use among youth. However, little attention has been placed on the role of such
factors within the context of a population at high-risk for substance use. Students
attending alternative/continuation high schools (CHS) are considered high-risk
because of several characteristics, including poor academic performance, truancy,
emotional or functional problems, substance use, and being relatively problem-prone
(Simon et al., 1994; Sussman et al., 1995c). Substance use is much higher among this
population compared to youth attending regular high school (RHS; Sussman et al.,
1995c), reporting use greater than the national average (Grunbaum et al., 2001;
Simon et al., 1994). Furthermore, CHS students are also at risk for substance abuse
(Sussman et al., 1995c), more likely to use more than one type of drug, and they tend
to initiate substance use before the age of 13 (Brener and Wilson, 2001).
Limited research exists on the role of ethnicity among this population. What
does exist points to some ethnic differences in the use of certain substances. For
instance, White students attending CHS, on average, report highest rates of use.
However, Latinos report higher lifetime and current cocaine use compared to White
and Black students (Grunbaum et al., 1999). Even though ethnicity has been shown
to account for some of the variability in substance use, attention has been placed on
the role of culture, especially acculturation, in substance use. Results generally
2
indicate that increasing levels of acculturation are associated with greater substance
use (Amaro et al., 1990; Burnam et al., 1987; Cherpitel, 1992; De la Rosa, 2002;
Epstein et al., 1998, 2001; Farabee et al., 1995; Marsiglia and Waller, 2002; Nielsen
and Ford, 2001; Unger et al., 2000; Vega et al., 1998b). Others have established that
place of birth, used as a proxy for acculturation, is also significantly associated with
substance use (Ebin et al., 2001; Farabee et al., 1995; Vega et al., 1993), with
foreign-born Latinos at less risk for substance use (Caetano et al., 1998), but their
risk increases the longer they remain in the United States (Caetano, 1987; Epstein et
al., 1996). However, such research conducted among high-risk youth attending
alternative/continuation high schools remains limited.
Communication, including interpersonal interactions and the exchange of
information, are crucial to acculturation (Hsu et al., 1993; Kim, 2005). It is through
interactions with the host society and its members that newcomers develop insight
about their new environment (Hsu et al., 1993). In fact, the more frequently an
individual associates with people in the host country, the greater the adjustment to
the host society (Hsu et al., 1993). A study conducted by Hsu et al. (1993) reported
that the acculturation level of peers has more influence on one’s own acculturation
than length of residence in the United States. These results indicate that the degree of
acculturation among peers in an individual’s immediate social network plays a
central role (Hsu et al., 1993). Given that peers exert a great amount of influence, it
is no wonder that peer groups are considered especially important antecedents that
3
may contribute to other behaviors, including substance use (Ennett and Bauman,
1994; Sieving et al., 2000).
Peer groups and the influence they exert have been used to explain many
adolescent behaviors (Ennett and Bauman, 1993; Jessor and Jessor, 1977; Kandel,
1978). When it comes to smoking and drug use behaviors, peer relationships are
considered one of the primary factors involved in whether or not adolescents decide
to engage in and maintain these behaviors (Ennett and Bauman, 1993, 2000; Kobus,
2003; Valente, 2003). Research confirms this by reporting similarities (also known
as homophily) in substance use patterns between adolescents and their peers (Kirke,
2006): 87). In other words, adolescents who use substances are likely to have
substance using friends (Bauman et al., 1984; Ennett and Bauman, 1993, 1994;
Ennett et al., 1994; Rai et al., 2003; Sieving et al., 2000; Sussman et al., 1990; Unger
et al., 2001; Windle, 2000).
The method by which peer groups influence substance use behaviors has
been attributed to peer pressure (Urberg et al., 1991). Also known as influence, this
suggests that peer groups are important antecedents that may contribute to substance
using behaviors (Ennett and Bauman, 1994; Sieving et al., 2000). These behaviors
may occur as a result of peers modeling substance use, peers making substances
more readily available, peers exerting mutual influence to use substances, and/or
peer norms that encourage substance use (Bandura, 1977; Gaughan, 2003; Kobus,
2003; Oetting and Donnermeyer, 1998; Perry and Jessor, 1985; Sieving et al., 2000).
4
There is another process known as selection that can also help explain
similarities in substance use patterns among peer groups (Bullers et al., 2001).
Selection occurs when adolescents purposefully select and keep friends based on
similar attitudes, beliefs and behaviors as their own, including that of substance use
(Ennett and Bauman, 1994; Kandel, 1978; Sieving et al., 2000). In fact, Kandel
(1985) reported that the formation of adolescent friendships are first based on
sociodemographic characteristics such as age and gender, followed by behaviors
such as illicit drug use. Selection can also occur when friendships are discontinued
when peer substance use become dissimilar (Sieving et al., 2000; Urberg et al.,
2003). Although most of the focus has been on influence and its role in adolescent
substance use, studies have increasingly reported that both influence and selection
equally contribute to similarities in substance use patterns among peer groups
(Ennett and Bauman, 1994; Hall and Valente, 2007; Hoffman et al., 2007; Kandel,
1985; Kirke, 2004, 2006; Valente et al., 2004).
Whether individuals are influenced by, or purposefully select peers with
similar substance using behaviors, the role peers play on substance use is very
important. There is potential significance in applying social network theory, which
focuses on the relational ties between individuals within a social system (Wasserman
and Faust, 1994). Social network theory provides a unique method for studying peer
networks because social network analysis allows for the direct investigation of
relationship patterns within networks (Kirke, 2004; Wasserman and Faust, 1994).
Despite the growing body of research on the social networks of youth and the role
5
peer networks have on substance use, to date little research has been devoted to
examining the social networks of high-risk youth.

Specific Aims
This project is being carried out in order to examine peer influence on
substance use within the context of acculturation and social networks among youth
attending alternative/continuation high schools throughout Southern California. The
specific aims of this project include:

1. To determine if linguistic acculturation is associated with substance use
2. To determine if the association between linguistic acculturation and
substance use is mediated by peer social influence
3. To determine if there are significant ethnic differences in substance use and
peer social influence, as measured by network composition
4. To determine if peer social influence, as measured by drug network
composition, is associated with risk for substance use
5. To determine if peer social influence, i.e. strength of tie, is associated with
risk for substance use

Theoretical Framework
There are a number of theoretical frameworks that have been applied in order
to explain how peers and social interactions affect behaviors such as alcohol,
6
tobacco, and drug use (Ennett and Bauman, 1994; Kobus, 2003). However, those that
focus on social processes, like friend selection, interpersonal influence and
behavioral imitation may provide insight into understanding substance use within the
context of these processes (Kobus, 2003). Theories specifically relevant to this
project include social learning theory (Bandura, 1977, 1986) and social network
theory (Wasserman and Faust, 1994). What these theories have in common is that
they assume youth look to relevant peers in order to determine behaviors that are
acceptable.
Social learning theory (SLT) suggests that behavior, perceptions of behavior,
and the environment all interact to influence one another (Bandura, 1977; Hoffman
et al., 2006). SLT takes into consideration both social processes and cognitive
mediation in order to understand the acquisition and maintenance of behavior
(Kobus, 2003). According to this theory, behaviors are learned through observing
others modeling the behavior, all the while associating rewards or punishments with
the behavior (Bandura, 1986). Indeed, the theory implies that adolescents may
become interested in using substances by simply observing others receive rewards
resulting from substance use (Valente et al., 2004). In so doing, adolescents take into
account those perceived rewards or consequences in deciding whether or not to
engage in drug use.
Keeping in mind that observation can constitute watching someone from a
distance or being shown how to use substances by one’s close friends (Valente et al.,
2004), adolescents are more likely to imitate the behavior from peers whom they
7
have the most contact with. Additionally, closer bonds as well as bonds formed
earlier in life are important in the social learning process (Kobus, 2003). Once the
adolescent has used substances, experience with the substance provides the
adolescent with direct information about the rewards and punishments of its use
(Kobus, 2003).
Social network theory focuses on the relational ties between individuals
within a social system (Wasserman and Faust, 1994). Social network theory assumes
that individuals within social systems interact with each other in varying degrees,
with such interactions serving as significant sources of information and support
(Kobus, 2003). Operating as significant sources in each other’s decision-making,
these interactions are seen as conduits through which information is transmitted
throughout the social system (Kobus, 2003). Although a social system (or social
network) is one whose population, more or less, can be identified by specific
boundaries (e.g. a community, a school, students in a classroom, etc.), social network
theory places extreme importance on the relationship ties among the interacting
members of the social network.
Figure 1.1 illustrates the proposed relationships between peer influence and
substance use. This model suggests that several sociodemographic variables
influence the way youth acculturate and respond to peer social influence to use
substances. In this model, acculturation is associated with peer social influence and
substance use (outcome). Peer social influence is also directly associated with
8
substance use. Finally, peer social influence mediates the relationship between
acculturation and substance use.
This dissertation examined the role of peer influence on substance use in two
ways, using two separate datasets (see Table 1.1 for sample size characteristics).
The first study examined peer social influence as a potential mediator for the
relationship between acculturation and substance use among a sample of Hispanic
regular and alternative/continuation high school students using data from 18 schools.
The second study also tests peer social influence as a potential mediator for the
association between acculturation and substance use, however the data come from a
recent study of 14 alternative/ continuation high schools. This study was restricted to
Hispanic students. The final study uses social network methods to examine the
association between peer influence and substance use. Peer social influence was
conceptualized as the network composition of peers, where having substance users in
one’s network, and being closer to first and second named peers would be associated
with substance use. The dissertation concludes with a summary of the studies and a
discussion of the potential study implications.

9
Table 1.1. Sample size characteristics

Dataset 1 Dataset 2
Study One Study Two Study Three
Alternative/
Continuation
High School
Students (n=582)

Regular
High School
Students (n=1138)

Alternative/
Continuation
High School
Students (n=714)

Alternative/
Continuation
High School
Students (n=968)

Gender
SES
Age
Ethnicity
Peer
Social
Influence
Acculturation
Substance
Use
Figure 1.1. Theoretical model of substance use
10
CHAPTER 2: LINGUISTIC ACCULTURATION, PEER INFLUENCE, AND
SUBSTANCE USE IN A SAMPLE OF HIGH- AND LOW-RISK YOUTH

CHAPTER 2 ABSTRACT
Aims: To test peer social influence as a potential mediator for the relationship
between linguistic acculturation and various substance use outcomes. Design,
participants and measurements: A cross-sectional baseline survey of 1720 Hispanic
adolescents attending regular (N=1138) and continuation (N=582) high schools in
Southern California. The Marin linguistic acculturation scale was used to assess
acculturation. Data on peer social influence, lifetime and current (past 30-day)
substance use (alcohol, cigarettes, marijuana, and hard drugs) were available.
Findings: Results indicate that, as in previous studies, continuation high school
(CHS) students use significantly higher levels of different substances than regular
high school (RHS) students. Linguistic acculturation was associated with increased
risk for lifetime and current use of alcohol, cigarettes, marijuana, and hard drug use
among the RHS sample. Among the CHS sample, it was associated with risk for
lifetime cigarettes, marijuana, hard drugs and current cigarette and marijuana use.
Results of mediation analyses indicate that peer social influence mediated the
relationship between acculturation and lifetime and current alcohol use, and lifetime
cigarette and hard drug use among CHS students, and current alcohol use among
11
RHS students. Conclusion: These results provide some evidence that peer social
influence may be an important variable influencing the relationship between
acculturation and substance use.

INTRODUCTION
Half of all American secondary school students have tried at least one illicit
substance by the time they reach graduation (Johnston et al., 2005). Students
attending continuation high schools (CHS) are at higher risk for illicit drug use
(Grunbaum and Basen-Engquist, 1993; Grunbaum et al., 2001; Grunbaum et al.,
2000), reporting rates three to five times higher than regular high school (RHS)
students (Sussman et al., 1995c). For example, 53.9% of students attending
continuation schools engage in past 30-day marijuana use, compared to 26.2% of
RHS students (Grunbaum et al., 2001). Similar differences are indicated for past 30-
day cocaine and alcohol use, and lifetime use of alcohol, marijuana, cocaine, and
illegal steroids (Grunbaum et al., 2001). In addition, when these high-risk youth use
drugs, they are more likely to use more than one type of drug (Beauvais and Oetting,
1986).
Researchers have long stated that individuals of different ethnic groups and
cultures exhibit varying patterns of substance use. Whites, on average, tend to report
the highest lifetime substance use rates, followed by Hispanics, then Blacks
(Beauvais and Oetting, 2002). Recently, Hispanic youth have equaled, and in some
instances surpassed the substance use rates of Whites. Among a nationally
12
representative sample of 12
th
graders, Hispanics report the highest rates of crack,
heroin (both IV and non-IV use), and Rohypnol use (Johnston et al., 2005). In some
parts of the U.S., alcohol, crystal methamphetamine, ecstasy (MDMA), and cocaine
use is higher among Hispanic than Black and White youth (Centers for Disease
Control and Prevention, 2004; Johnston et al., 2001; National Institute of Drug
Abuse and University of Michigan, 2004). Among a nationally representative sample
of CHS students, Hispanics report the highest lifetime and current cocaine use,
compared to White and Black students (Grunbaum et al., 1999).

Acculturation and substance use
With increasing cultural diversity in the U.S., opportunities for intercultural
interactions exist. These interactions involve the interchange of cultural behaviors,
attitudes (Unger et al., 2002), social behaviors, and customs (Unger et al., 2004).
Known as acculturation, as individuals incorporate the mainstream culture,
acculturation can lead to changes in speech, attitudes, and social behavior (Berry,
2005; Unger et al., 2002). Therefore, increasing interactions with individuals in the
host country may result in greater acculturation (Kim, 2005). Some have argued that
these interpersonal interactions are crucial to acculturation; one of the most salient
forms in the cultural learning process (Hsu et al., 1993; Kim, 2005).
While the literature has produced some equivocal findings on association
between acculturation and substance use, generally, greater levels of acculturation
are associated with greater risk for smoking, alcohol, and substance use (Amaro et
13
al., 1990; Blake et al., 2001; Burnam et al., 1987; Cherpitel, 1992; De la Rosa, 2002;
Epstein et al., 1998, 2001; Landrine and Klonoff, 2004; Marsiglia and Waller, 2002;
Nielsen and Ford, 2001; Unger et al., 2000; Vega et al., 1998b). Researchers have
hypothesized that this relationship may be due to greater availability, access, and
acceptability of recreational drug use in the U.S. (Escobar, 1998). Others have
attributed it to marginalization and discrimination (Alaniz, 2002; Felix-Ortiz and
Newcomb, 1995; Portes and Rumbaut, 2001), while others ascribe it to a greater
susceptibility to peer pressure among more acculturated adolescents (Wall et al.,
1993).
Acculturation is a complex process, involving several dimensions. One
important aspect in the acculturation process involves improved speaking
proficiency. As youth acculturate to the U.S. culture, their English speaking
proficiency improves. This has the potential to affect interactions with peers (Unger
et al., 2000), allowing acculturating youth to communicate with peers they once had
been unable to. In so doing, they may access a larger and more diverse network of
peers, one that reaches beyond their initial immigrant peer group (Marsiglia and
Waller, 2002; Wall et al., 1993). Furthermore, as youth acculturate, they may be
more likely to spend time with highly acculturated or U.S.-born peers who have
incorporated the mainstream norms of the American youth culture (Unger et al.,
2000). Greater association with U.S.-born peers may expose acculturating youth to
relatively more pro-substance peer influences (Rai et al., 2003; Rice et al., 2003;
Unger et al., 2000; Urberg et al., 1997; Windle, 2000) and perhaps allow them to
14
perceive substance use as normative. Acculturating youths may also find themselves
more often in circumstances where peers offer or are using substances because they
now can communicate in English with others that have access to drugs in the
environment (Unger et al., 2000). Research supports this hypothesis, suggesting that
level of acculturation has the potential to affect the degree to which youth are
influenced by their peers (Wall et al., 1993).
Peer social influence
Peer groups, and the influence they exert, have been used to explain many
adolescent behaviors (Ennett and Bauman, 1993; Jessor and Jessor, 1977; Kandel,
1978). Research has established that peer influence is a primary factor involved in
whether or not youth decide to engage in and maintain substance use (Ennett and
Bauman, 1993, 2000; Hoffman et al., 2006; Jessor and Jessor, 1977; Kobus, 2003;
Sieving et al., 2000; Sussman et al., 1995a; Valente, 2003; Valente et al., 2004).
According to social learning theory, learning occurs through modeling, which is
based on the direct observation and imitation of role models’ behavior, or through
vicarious learning and reinforcement (Bandura, 1977). So, substance use behaviors
may occur as a result of peers’ modeling substance use, peers making substances
more readily available, peers exerting mutual influence to use substances, and/or
peer norms that encourage substance use (Bandura, 1977; Gaughan, 2003; Kobus,
2003; Oetting and Donnermeyer, 1998; Perry and Jessor, 1985; Sieving et al., 2000).
Studies conducted among specific ethnic groups also confirm this relationship
15
(Dornelas et al., 2005; Epstein et al., 1999a; Epstein et al., 1999b; Gritz et al., 2003;
Unger et al., 2001).
To date, relatively little research has been conducted examining potential
mediators for the association between acculturation and substance use among both a
low-risk and high-risk youth population. The current study builds on prior literature
by investigating how peer social influence mediates the relationship between
acculturation and various lifetime and current substance use outcomes among
Hispanic students in both regular and continuation high schools throughout Southern
California. Study hypotheses include: 1) linguistic acculturation is associated with
increased risk for substance use and 2) this association is mediated by peer social
influence.

METHODS
Sample and data collection
A total of 18 high schools (9 RHS; 9 CHS) from two Southern California
counties (Los Angeles and Ventura) were recruited to participate in this study. Data
on demographic items, current living situation, language-based acculturation,
behavioral items such as current (past 30-day) and lifetime use of substances
(cigarettes, alcohol, marijuana, and hard drugs), peer social influence, and other
constructs were collected at pre-test, post-test, and one-year follow-up. Prior to pre-
test survey administration, parental written or verbal consent, and/or student assent
were obtained. Participation was completely voluntary, and confidentiality was
16
emphasized throughout the length of the study. The Institutional Review Board at the
University of Southern California approved all study procedures. A total of 2,751
students (1,902 RHS students; 849 CHS students) completed the pre-test survey. The
sample for this study was restricted to those who completed the pre-test survey and
who self-identified as Latino or Hispanic (including Mexican American, Mexican,
Central American, South American, and others), yielding a final sample size of 1,720
(1,138 RHS; 582 CHS). Although participants who reported speaking another
language other than English were not asked about the other language, almost 90% of
our sample is of Mexican origin; it is likely that their other language is Spanish.
Census data indicate that 65.4% of Californians over the age of 4 who speak another
language at home speak Spanish (Shin and Bruno, 2003). Furthermore, Mexican
census data indicate that less than 7% of the population speak an indigenous
language (INEGI, 2006), indicating that Mexican immigrants are more likely to
speak Spanish than an indigenous dialect. No program effects were analyzed in this
study.

Measures
Substance use
Measured separately for both lifetime and current (last 30 days) use, we
asked participants “How many times have you tried each of the drugs below?” The
drug category responses include: cigarettes; alcohol; marijuana (weed); cocaine
(crack); hallucinogens (LSD, Acid, mushrooms); stimulants (ice, speed,
17
amphetamines); inhalants (rush, nitrous, glue); ecstasy (MDMA, XTC, Adam); and
other (depressants, PCP, steroids, heroin, etc.). Responses were provided on an 8-
point rating scale ranging from 0 to more than 100 times in increasing intervals (e.g.
0 times; 1-10 times; 11-30 times; 31-50 times; 51-70 times; 71-90 times; 91-100
times; more than 100 times). The reliability and predictive validity of this format
have been previously established (Sussman et al., 1995a; Sussman et al., 1998). By
averaging across responses to cocaine, hallucinogens, stimulants, inhalants, ecstasy,
and other drugs, a hard drug use index was created (O = 0.85 for lifetime use; O =
0.92 for current use). A higher mean value on cigarette, alcohol, marijuana, and hard
drug use indicates greater use of those substances.

Linguistic acculturation
Four questions were used to assess linguistic acculturation in our sample.
They included questions about language most often read and spoken; language
usually spoken with friends; language preference in watching and listening to
movies, TV, and radio shows; and language most often spoken at home. Responses
were on a five-point Likert scale, ranging from only English to only another
language (not English). This scale is adapted from the short form of Marin and
colleagues’ acculturation scale, which has been previously shown to have good
validity and reliability (Marin et al., 1987). Mean scores ranged from 1 to 5, with
lower scores indicating English preference, a marker for higher levels of
acculturation (O = 0.86). This variable was reverse recoded to where a higher mean
18
value indicates greater English proficiency. Although this unidimensional
acculturation scale measured one aspect of acculturation, a major component of
acculturation is language use. Language accounts for a considerable portion of the
variance in several acculturation measures (Cherpitel and Borges, 2002; Cuellar et
al., 1980; Epstein et al., 1996; Lessenger, 1997; Marsiglia and Waller, 2002; Vega et
al., 1998a) and there is sufficient evidence to suggest that language is an efficient
proxy for acculturation (Marsiglia and Waller, 2002; Norris et al., 1996).
Peer social influence
One type of question was used to assess peer social influence: In the last
month, how may of your five closest friends have used each of the drugs below. This
question was asked separately for cigarettes, alcohol, marijuana, and hard drugs.
Response items were given on a six point scale in increasing intervals of one,
beginning with zero friends (Sussman et al., 2000; Sussman et al., 1995b).

Covariates
Other variables found to be associated with substance use behaviors were
included as covariates, including gender, age, and socioeconomic status (SES).
Socioeconomic status was measured separately for fathers and mothers with: What is
the highest grade completed by your (father or mother)? Response options included:
1=not completed elementary school (8
th
grade); 2=not completed high school (12
th

grade); 3=completed high school (received diploma); 4=some college or job training
19
(1 to 3 years); 5=completed college (4 years); 6=completed graduate school; and
7=don’t know. Don’t know was recoded to missing. These items were averaged in
order to create one SES score, where a higher score indicates greater education level
and SES (O=0.69).

Data Analysis
Frequencies, chi-square, and Student t-test analyses were used to analyze
differences between the two types of schools. After checking for regression
assumptions, we used square root transformation to change the substance use
(dependent) variables (Netter et al., 1990). We conducted mediated associations by
using the method described by Baron and Kenny (Baron and Kenny, 1986). We first
ran the multilevel linear regression models testing whether linguistic acculturation
(independent variable) was significantly associated with each of the lifetime and
current substance use outcomes, without the hypothesized mediating variable (peer
social influence). Next, we examined whether linguistic acculturation was associated
with peer social influence (potential mediating variable and the dependent variable in
this model). Finally, we examined whether linguistic acculturation was significantly
associated with each of the lifetime and current substance use outcomes, controlling
for peer social influence in the model. Mediation is supported if (a) linguistic
acculturation is significantly associated with substance use; (b) linguistic
acculturation was significantly related to the mediator (peer social influence); (c)
peer social influence was significantly associated with substance use; and (d) the
20
previously significant direct relationship between linguistic acculturation and
substance use became weaker when controlling for peer social influence. If the
relationship between linguistic acculturation and substance use became non-
significant after controlling for peer social influence, the mediation effect would be
considered complete; if the association remained significant but its effect was
reduced, then the mediation effect would be considered partial (Baron and Kenny,
1986). We evaluated the mediation pathway with the Sobel test, examining the
hypothesis that the indirect pathway from acculturation to substance use through peer
social influence is different from zero (Sobel, 1982). This last step was conducted
using the SAS programming developed by Dudley, Benuzillo, and Carrico (2004),
testing the mediation effect and the proportion of the variance attributed to the
mediation effect (Beck et al., 2005). Mediation analyses were conducted with
multilevel linear regression, using individuals as the unit of analysis and taking into
account the random-effects resulting from data collected within schools (Singer,
1998). Regression analyses were stratified by type of school (continuation or regular
high school), controlling for age, gender, and SES. A one-tailed alpha of 0.05 was
used to determine level of significance and analyses were conducted with the
Statistical Analysis System software (SAS Institute, 1990).

21
RESULTS
Demographic characteristics
Table 2.1 illustrates the demographic characteristics of the overall population,
and by type of school. The overall mean age was 15.4 years (SD=1.3), with a mean
SES of 2.8 (SD=1.2), indicating that on average parents completed a high school
diploma. A slight majority of the sample was male (51.9%), and spoke English only
(57%). When examining differences between the two types of schools, continuation
high school students were significantly more likely to be older, be of lower SES,
male, English only speakers, have more peer social influences to use alcohol,
cigarettes, marijuana and hard drugs, and use more of all substances compared to
regular high school students.

Associations between linguistic acculturation and substance use
Table 2.2 shows the associations between linguistic acculturation and several
lifetime and current substance use outcomes by type of school. The table presents
unstandardized beta coefficients of these associations, adjusting for age, gender,
SES, and random-effects of schools. These associations are also illustrated in Figure
2.1 and 2.2. Among continuation high school students, for each one unit increase in
the level of English speaking proficiency, the risk significantly increases by 12% for
lifetime cigarette, 22% for lifetime marijuana, 7% for lifetime hard drugs, 12% for
22
Table 2.1. Demographic, language preference, and substance use differences by
school type
Overall Regular Continuation
High School High School
(n=1720) (n=1138) (n=582)
Age (Mean, SD)** 15.4 (1.3) 14.8 (0.9) 16.7 (0.8)

SES (Mean, SD)* 2.8 (1.2) 2.8 (2.7) 2.5 (2.6)

Gender (n,%)*
Male 892 (51.9) 559 (49.2) 333 (57.2)
Female 826 (48.1) 577 (50.8) 249 (42.8)

Language Preference (n, %)**
English Only 980 (57.0) 619 (54.4) 361 (62.0)
Bilingual 586 (34.1) 399 (35.1) 187 (32.1)
Spanish Only 154 (8.9) 120 (10.5) 34 (5.8)

Peer Social Influences To Use:
Alcohol**
0 Friends 505 (32.5) 401 (37.7) 104 (21.3)
1-2 Friends 347 (22.4) 264 (24.8) 83 (16.9)
3-5 Friends 700 (45.1) 398 (37.5) 302 (61.8)
Mean (SD)** 2.3 (2.0) 2.0 (1.9) 3.0 (2.0)

Cigarettes**
0 Friends 823 (53.6) 669 (64.0) 154 (31.4)
1-2 Friends 342 (22.3) 217 (20.8) 125 (25.5)
3-5 Friends 370 (24.1) 159 (15.2) 211 (43.1)
Mean (SD)** 1.4 (1.8) 0.9 (1.5) 2.3 (2.0)

Marijuana**
0 Friends 620 (41.1) 501 (48.8) 119 (24.8)
1-2 Friends 301 (20.0) 197 (19.2) 104 (21.7)
3-5 Friends 586 (38.9) 329 (32.0) 257 (53.5)
Mean (SD)** 2.0 (2.1) 1.6 (1.9) 2.8 (2.0)

Hard Drugs**
0 Friends 1063 (70.4) 791 (76.3) 272 (57.4)
1-2 Friends 255 (16.9) 155 (15.0) 100 (21.1)
3-5 Friends 192 (12.7) 90 (8.7) 102 (21.5)
Mean (SD)** 0.8 (1.5) 0.6 (1.3) 1.2 (1.7)

Substance Use (n, %)
Lifetime
Alcohol**
Never 444 (26.2) 367 (32.6) 77 (13.5)
1 - 10 times 651 (38.4) 481 (42.7) 170 (29.8)
11 - 30 times 230 (13.5) 138 (12.3) 92 (16.1)
31 or more times 372 (21.9) 140 (12.4) 232 (40.6)

23
Table 2.1, Continued

Cigarettes**
Never 922 (54.6) 739 (66.0) 183 (32.1)
1 - 10 times 491 (29.1) 299 (26.7) 192 (33.7)
11 - 30 times 85 (5.0) 33 (3.0) 52 (9.1)
31 or more times 191 (11.3) 48 (4.3) 143 (25.1)

Marijuana**
Never 888 (52.7) 734 (65.7) 154 (27.0)
1 - 10 times 393 (23.3) 235 (21.0) 158 (27.8)
11 - 30 times 102 (6.0) 51 (4.6) 51 (9.0)
31 or more times 303 (18.0) 97 (8.7) 206 (36.2)

Hard drugs**
Never 1259 (76.6) 940 (86.1) 319 (57.7)
1 - 10 times 242 (14.7) 113 (10.4) 129 (23.3)
11 - 30 times 65 (4.0) 14 (1.3) 51 (9.2)
31 or more times 78 (4.7) 24 (2.2) 54 (9.8)

Current
Alcohol**
Never 984 (58.0) 736 (65.5) 248 (43.2)
1 - 10 times 556 (32.8) 331 (29.5) 225 (39.2)
11 - 30 times 101 (5.9) 37 (3.3) 64 (11.1)
31 or more times 56 (3.3) 19 (1.7) 37 (6.5)

Cigarettes
Never** 1403 (83.0) 1022 (91.3) 381 (66.7)
1 - 10 times 169 (10.0) 79 (7.0) 90 (15.8)
11 - 30 times 39 (2.3) 10 (0.9) 29 (5.1)
31 or more times 80 (4.7) 9 (0.8) 71 (12.4)

Marijuana**
Never 1293 (76.6) 948 (84.8) 345 (60.5)
1 - 10 times 231 (13.7) 123 (11.0) 108 (19.0)
11 - 30 times 62 (3.7) 21 (1.9) 41 (7.2)
31 or more times 102 (6.0) 26 (2.3) 76 (13.3)

Hard drugs**
Never 1507 (90.8) 1041 (94.6) 466 (83.4)
1 - 10 times 115 (6.9) 46 (4.2) 69 (12.3)
11 - 30 times 19 (1.2) 4 (0.4) 15 (2.7)
31 or more times 18 (1.1) 9 (0.8) 9 (1.6)

Note: Student t-test or chi-square analyses were used to test for differences between types of school.
* p<.05; ** p<.001

24
Figure 2.1. Lifetime and current substance use as a function of linguistic
acculturation among Hispanic continuation high school students in Southern
California
0
10
20
30
40
50
60
70
80
90
Spanish only Spanish more
than English
English and
Spanish
equally
English more
than Spanish
English only
Linguistic Acculturation
Percent who have ever tried substances
Lifetime Alcohol
Lifetime Cigarette
Lifetime Marijuana
Lifetime Hard Drugs
Current Alcohol
Current Cigarette
Current Marijuana
Current Hard Drugs
Figure 2.2. Lifetime and current substance use as a function of linguistic
acculturation among Hispanic regular high school students in Southern California
0
10
20
30
40
50
60
70
80
Spanish only Spanish more
than English
English and
Spanish
equally
English more
than Spanish
English only
Linguistic Acculturation
Percent who have ever tried substances
Lifetime Alcohol
Lifetime Cigarette
Lifetime Marijuana
Lifetime Hard Drugs
Current Alcohol
Current Cigarette
Current Marijuana
Current Hard Drugs
25
Table 2.2. Regression weights on lifetime and current substance use variables, stratified by school type
Lifetime Use Current Use
Alcohol Cigarettes Marijuana Hard Drugs Alcohol Cigarettes Marijuana Hard Drugs
Continuation High School
Age 0.09* 0.09* -0.02 0.02 0.04 0.08* -0.05 -0.02
SES -0.04 0.04 -0.01 0.00 -0.01 0.04 -0.02 0.01
Female -0.00 -0.10 -0.06 -0.01 -0.09 -0.07 -0.16 0.00
Acculturation 0.07 0.12* 0.22** 0.07* 0.02 0.12* 0.21** 0.02
ICC: 0.02 0.07 0.03 0.05 0.01 0.11 0.03 0.02
Regular High School
Age 0.06* 0.09** 0.08** 0.02** 0.02 0.04* 0.01 0.01
SES -0.01 -0.01 -0.02 -0.00 -0.02 -0.00 -0.01 -0.00
Female 0.06 -0.05 -0.01 0.01 0.07 -0.02 -0.02 0.01
Acculturation 0.13** 0.11** 0.19** 0.04** 0.06* 0.04* 0.08** 0.02*
ICC: 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.01
Note: Multilevel models adjust for random-effects of schools.
ICC: Intraclass correlation
* p<.05; ** p<.01
26
current cigarette, and 21% for current marijuana use. Among regular high school
students, all of the substance use outcomes were significantly associated with
acculturation.
Table 2.3 presents the association between linguistic acculturation and peer
social influence. Among continuation high school students, acculturation was
significantly associated with peer influences to use alcohol (b=0.33; p<.05),
cigarettes (b=0.28; p<.05), marijuana (b=0.56; p<.01), and hard drugs (b=0.33;
p<.01). Among regular high school students, acculturation was significantly
associated with peer influences to use alcohol (b=0.28; p<.01), cigarettes (b=0.17;
p<.05), and marijuana (b=0.41; p<.01).
Table 2.4 shows the final mediation models, with lifetime and current
substance use as dependent variables (models for each outcome was tested
separately; 8 models), acculturation as the independent variable, controlling for peer
social influence (the mediator), age, SES, gender, and random-effects of schools.
Among continuation high school students, peer social influence was a significant
predictor for all substance use outcomes, with greater peer influence associated with
greater substance use. This table also indicates that the significant effects of
acculturation observed in Table 2.2 became weaker and non-significant for the
outcomes of lifetime cigarette and lifetime hard drug use, indicating complete
mediation. The regression weights for acculturation, without the mediator, decreased
after including peer social influence into the model, but remained significant for the
outcomes of lifetime marijuana, current cigarettes, and current marijuana use,
27
indicating partial mediation. Among regular high school students, peer social
influence was significantly associated with greater substance use (all outcomes). In
addition, the effect of acculturation weakened and became non-significant for the
outcome of current alcohol use, indicting complete mediation. The regression
weights for acculturation decreased after adding peer social influence in the model,
but they remained significant for lifetime alcohol, lifetime cigarettes, lifetime
marijuana, current cigarette, and current marijuana use, indicating partial mediation.
According to the criteria set by Baron and Kenny (1986), peer social influence met
the conditions for being a mediator for the relationship between acculturation and
several substance use outcomes. Results of the Sobel test indicated that the indirect
effects were significant.

28
Table 2.3. Associations between peer social influence and linguistic acculturation,
stratified by school type

Dependent Variables
Peer social influence to use:
Alcohol Cigarettes Marijuana Hard Drugs
b b b b
Predictors
Continuation High School
Age 0.20 0.09 -0.02 -0.19*
SES -0.05 0.08 -0.08 -0.07
Female 0.29 -0.02 0.08 0.28
Acculturation 0.33* 0.28* 0.56** 0.33**

ICC: 0.04 0.03 0.03 0.00

Regular High School
Age 0.15* 0.14* 0.12 0.08*
SES 0.12 0.05 -0.01 0.01
Female 0.19 0.01 0.16 -0.01
Acculturation 0.28** 0.17* 0.41** 0.05

ICC: 0.02 0.01 0.00 0.01

Note: Multilevel models adjust for random-effects of schools
ICC: Intraclass correlation.
* p<.05; ** p<.01
29
Table 2.4. Regression weights on lifetime and current substance use variables, stratified by school type
Lifetime Use Current Use
Alcohol Cigarettes Marijuana Hard Drug Alcohol Cigarettes Marijuana Hard Drug
Continuation High School
Age 0.03 0.06 -0.04 0.05 0.04 0.07 -0.02 -0.01
SES -0.04 0.01 -0.01 0.01 0.00 0.01 0.00 0.00
Female -0.04 -0.10 -0.07 -0.05 -0.12* -0.05 -0.15* -0.00
Acculturation 0.01 0.05 0.14* 0.03 0.00 0.10* 0.16* 0.01
Peer Social Influence† 0.07** 0.21** 0.21** 0.12** 0.12** 0.14** 0.15** 0.05**
Intraclass Correlation 0.01 0.04 0.03 0.05 0.00 0.08 0.01 0.02
Sobel Value 4.22** 3.28** 5.20** 2.22* 4.18** 3.50** 4.95** 2.22*
% total effect mediated 43% 31% 46% 19% 72% 38% 46% 31%
Regular High School
Age 0.04 0.06** 0.05* 0.02* 0.00 0.02 0.01 0.00
SES -0.03* -0.02 -0.03 -0.01 -0.03* -0.01 -0.00 -0.00
Female 0.01 -0.05 -0.04 0.01 0.03 -0.02 -0.02 0.01
Acculturation 0.09** 0.10** 0.11** 0.04** 0.02 0.03* 0.04* 0.01*
Peer Social Influence† 0.18** 0.18** 0.18** 0.08** 0.13** 0.10** 0.10** 0.04**
Intraclass Correlation 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Sobel Value 2.17* 2.67* 3.59** 2.97* 2.20* 2.81* 3.56** 2.88*
% total effect mediated 59% 47% 45% 51% 75% 38% 37% 69%
Note: Multilevel modeling adjusts for random-effects of schools.
† Indicates peer social influence to use respective substance.
* p<.05; ** p<.01
30
DISCUSSION
Our results indicate that linguistic acculturation (greater English speaking
proficiency) is associated with greater risk for lifetime use of cigarettes, marijuana,
and hard drugs and current use of cigarettes and marijuana among Hispanic
continuation high school students. Among Hispanic regular high school students,
linguistic acculturation was associated with greater use of all lifetime and current use
of alcohol, cigarettes, marijuana, and hard drugs. Our findings are consistent with
those for Hispanic youths attending regular (Ebin et al., 2001; Epstein et al., 1998,
2001; Landrine et al., 1994; Marsiglia et al., 2004; Marsiglia and Waller, 2002;
McQueen et al., 2003; Unger et al., 2000; Vega et al., 1998b) and continuation high
schools (Fosados et al., under review-a). Acculturation was found to have a strong
impact on all peer social influence variables in the continuation high school sample,
and an impact on all except for peer social influence to use hard drugs among the
regular high school sample. Peer social influence was also found to have a strong
association with all of the substance use outcomes, for both continuation and regular
high school students. This is also consistent with the literature (Ennett and Bauman,
1993, 2000; Hoffman et al., 2006; Jessor and Jessor, 1977; Kobus, 2003; Sieving et
al., 2000; Sussman et al., 1995a; Valente, 2003; Valente et al., 2004).
Peer social influence mediates the impact of acculturation on lifetime and
current substance use. According to the criteria set by Baron and Kenny (1986), peer
social influence met the conditions for being a mediator for the relationship between
acculturation and some substance use outcomes. The significant association between
31
linguistic acculturation and substance use (lifetime cigarettes and hard drugs among
continuation high school students) became non-significant when peer social
influence was added to the model. For several of the remaining substance use
outcomes, especially among regular high school students, evidence for partial
mediation was observed. Although peer social influence was measured slightly
different than our study, our results are consistent with other studies examining
mediation of social influence on the association between acculturation and substance
use among Hispanic youth attending regular (Epstein et al., 2003; Samaniego and
Gonzales, 1999; Unger et al., 2000) and continuation high schools (Fosados et al.,
under review-a). These findings suggest that peer social influence may help explain
the association between acculturation and substance use.
As Hispanic youths acculturate, as indicated by greater English speaking
proficiency, they may become more susceptible to peer influences. Explanation for
this occurrence may be found in language and cultural differences. For example,
Spanish speakers may be limited in the number of peers they communicate with,
inadvertently forming a tight knit group of other Spanish speakers. Since research
indicates that immigrant groups are less likely to use substances than their U.S.-born
counterparts, these peer groups may be protected from pro-substance using peer
influences (Marsiglia and Waller, 2002; Unger et al., 2000). Also, because Hispanic
cultures tend to be more collectivist (Delgado-Gaitan, 1994) with a strong loyalty to
and an emphasis on family (Marin and Marin, 1991; Sabogal et al., 1987),
32
unacculturated Hispanic youth may be more influenced by their parents and their
stance against substance use, and less so by their peers (Wall et al., 1993; Wills et al.,
2004).

Limitations and conclusion
There are several limitations to this research. First, this study measured one
aspect of acculturation, language preference, instead of implementing a scale that
measured several dimensions involved in the acculturation process. Yet, there is
enough evidence to suggest that language is an important part of the acculturation
process, serving as an efficient proxy for acculturation (Epstein et al., 1996;
Marsiglia and Waller, 2002). Future research should consider using multi-
dimensional acculturation scales in order to help determine which dimensions are
more specifically associated with substance use.
Second, our results are based on self-reported acculturation, peer social
influence, and substance use. Despite this fact, because our data were confidential,
the underreporting bias tends to diminish (Sussman et al., 1995b). Previous research
has reported the validity of self-reported substance use (Graham et al., 1984; Harrell,
1997), and all of our measures have been reliably used in other studies (Marin et al.,
1987; Sussman et al., 2000; Sussman et al., 1995b).
Third, our results are based on cross-sectional data; therefore no causality can
be inferred. It is possible that peer social influence is a result of substance use, where
individuals desiring to use substances purposefully seek out substance users as peers.
33
Or, acculturation could mediate the relationship between social influence and
substance use; unacculturated youth may deliberately seek out certain types of peers
in an effort to understand and acculturate to the new environment or due to a desire
to learn English. Research is needed to test these alternate hypotheses. Even though
cross-sectional data have been used in the literature to test mediation (Barsevick et
al., 2006; Beck et al., 2005; Unger et al., 2000), only longitudinal research can shed
light on the order of precedence in a relationship (Collins et al., 1987). It can also
allow for the examination of several aspects of the mediational model, including
temporal precedence and stability across time (MacKinnon et al., 2007).
Finally, our results may only be generalizable to Hispanic youth attending
regular and continuation high schools in Southern California. Because the majority
of Hispanics in Southern California area are of Mexican descent, these results may
not generalize to Hispanics from different countries of origin (e.g. Guatemala, Puerto
Rico, etc.).
Despite these limitations, these findings have possible implications. For
instance, school-based substance use prevention programs may be more effective if
they are developed with consideration of the impact peer social influences have on
acculturating youth. As youth acculturate, their interactions with their peers change,
giving them access to a more diverse network of peers who may influence substance
use behaviors. If programs can help ease the transition from a limited to a broader
peer network, all the while maintaining the protective cultural factors, perhaps
substance use can be prevented as youths acculturate. These findings also provide
34
new insight about one possible mechanism for explaining the relationship between
acculturation and substance use. These results show that acculturation influences
peer social influence, and peer social influence affects substance use. Future research
is needed to better understand the relationship among acculturation, peer networks,
peer influence, and substance use and to determine if other factors not measured in
this study may contribute to substance use.

35
CHAPTER 3: EXPLORING THE ASSOCIATION BETWEEN LINGUISTIC
ACCULTURATION AND SUBSTANCE USE AMONG HISPANIC
CONTINUATION HIGH SCHOOL YOUTH: EVIDENCE FOR
SOCIAL INFLUENCE AS A MEDIATOR

CHAPTER 3 ABSTRACT
Research suggests that acculturation increases the risk of substance use
among Hispanic adolescents. However, this relationship is not well understood
among high-risk youth. This study examined associations between acculturation and
several substance use indicators among a sample of 714 Hispanic youth attending
continuation high schools throughout Southern California. Peer social influence was
assessed as a potential mediator. Acculturation, measured by language use, was
associated with increased risk of lifetime alcohol, marijuana and current alcohol,
cigarettes, marijuana, and hard drug use, controlling for age, SES, and gender.
Results of mediation analyses indicate that peer social influence mediated the
relationship between acculturation and lifetime alcohol, and current alcohol,
cigarettes, and hard drug use. Evidence for partial mediation was observed with
lifetime and current marijuana use. These results provide evidence that peer social
influence is an important mediating variable that should be considered when
examining the relationship between acculturation and substance use.
36
INTRODUCTION
Immigrants to a new country are oftentimes influenced by and may
incorporate the values, norms, practices, behaviors, language, and customs of the
host country. This process, known as acculturation, has the potential to influence
individual preferences and attitudes about what constitutes normative behavior
(Warner et al., 2006). Hispanic immigrants to the U.S. are no different. Not only are
they exposed to a wealth of new behaviors, some of these new behaviors may place
them at an elevated risk for several adverse health outcomes (Black and Markides,
1993; Vega et al., 1998a), gradually shifting their risk behaviors to resemble those of
the host country (Unger et al., 2000). Such is the case with substance use. Hispanics
living in the U.S. engage in higher rates of substance and alcohol use than their
counterparts living in the country of origin (Caetano and Mora, 1988; Cherpitel and
Borges, 2001; De la Rosa et al., 1990). As a result, Hispanics are disproportionately
affected by alcohol-related problems (e.g. cirrhosis of the liver, cirrhosis mortality;
Caetano, 2003), HIV/AIDS, and other infectious diseases resulting from substance
use (National Institute on Drug Abuse, 2004).
In previous years, Hispanic youth have reported substance use rates
somewhere between White and Black youth. Recently, however, Hispanic youth
have equaled, and in some instances, surpassed the substance use rates of Whites.
For instance, among a nationally representative sample of 12
th
graders, Hispanics
reported the highest rates of crack, heroin (both IV and non-IV use), and Rohypnol
use (Johnston et al., 2005). In some parts of the U.S., crack and ecstasy (MDMA) use
37
was highest among Hispanic than Black and White youth (Johnston et al., 2001).
Hispanics have also reported some of the highest rates of lifetime alcohol use
(Centers for Disease Control and Prevention, 2004).
Substance use rates are higher for youth attending alternative/continuation
high school (CHS) than those in regular high school (RHS). CHS students are
significantly more likely to report illicit drug use than students attending regular high
school (Grunbaum and Basen-Engquist, 1993; Grunbaum et al., 2001; Grunbaum et
al., 2000), reporting rates three to five times higher (Sussman et al., 1995c). In
addition, when these high-risk youth use drugs, they are more likely to use more than
one type of drug (Beauvais and Oetting, 1986). Although comparatively less research
is conducted with continuation high school students than with regular high school
students, fewer still are the studies that have examined ethnic differences among the
former. Yet, studies show that Hispanics may be at greater risk for substance use. A
study conducted among a nationally representative sample of continuation high
school students indicated that Hispanics reported the highest lifetime cocaine use
(46.4%), compared to White (43.8%) and Black (5.7%) students (Grunbaum et al.,
1999). Similar patterns were observed with current cocaine use (19.4%; 17.7%;
3.6%; respectively) and having tried cocaine before the age of 13 (5.7%; 4.0%;
3.1%; respectively). Another study conducted among continuation high school
students indicated that being Hispanic was significantly predictive of hard drug use
(McCuller et al., 2001).

38
Acculturation and substance use
Research on Hispanic substance use has emphasized the importance of
acculturation. Defined as the “dual process of cultural and psychological change that
takes place as a result of contact between two or more cultural groups and their
individual members”(Berry, 2005: 698), the interchange that occurs when different
cultures come into contact with one another can lead to changes in speech, social
behavior, attitudes, and customs (Berry, 1998, 2005; Unger et al., 2000; Unger et al.,
2002). Some have argued that these interpersonal communications are crucial to
acculturation, indicating it as one of the most salient forms in the cultural learning
process (Hsu et al., 1993; Kim, 2005). As such, increasing interactions with
individuals in the host country may result in greater acculturation (Kim, 2005).
Although the literature has produced some inconsistent findings, generally, greater
levels of acculturation are associated with greater risk for smoking (Landrine et al.,
1994; Unger et al., 2000), alcohol and substance use (Amaro et al., 1990; De la Rosa,
2002; Epstein et al., 1998, 2001; Farabee et al., 1995; Marsiglia and Waller, 2002;
Nielsen and Ford, 2001; Vega et al., 1998b).
Researchers have hypothesized that the relationship between acculturation
and substance use may be due to marginalization and discrimination, or conflicts that
may occur due to adolescents and parents progressing through the acculturation
process differently (Alaniz, 2002; Felix-Ortiz and Newcomb, 1995; Portes and
Rumbaut, 2001). Others have attributed it to greater availability, access, and
acceptability of recreational drug use in the U.S. culture (Escobar, 1998), or to a
39
greater susceptibility to peer pressure among more acculturated adolescents (Wall et
al., 1993). Since acculturation to the U.S. culture often involves improved English
speaking proficiency, in so doing, youths’ interactions with their peers change
(Unger et al., 2000). Acculturating youth may be able to communicate with peers
they once had been unable to, potentially accessing a larger or more diverse network
of peers that reaches beyond the immigrant peer group (Wall et al., 1993). As a
result, youth with greater English speaking proficiency may come across more pro-
substance peer influences, including perceiving substance use (including alcohol and
tobacco) as a normative behavior (Rai et al., 2003; Rice et al., 2003; Unger et al.,
2000; Urberg et al., 1997; Windle, 2000). Furthermore, more acculturated youth may
associate more with highly acculturated and/or U.S.-born peers, who are more likely
to have incorporated the mainstream norms of the American adolescent culture
(Unger et al., 2000). Hence, as youth acculturate, their association with U.S.-born
peers may increase, causing them to perceive substance use as normative, and they
may perhaps find themselves in circumstances where peers offer or are using
substances (Unger et al., 2000). Empirical evidence suggests that level of
acculturation may affect the degree to which youth are influenced by their peers
(Wall et al., 1993). Compared to low acculturated youth, highly acculturated youth
were more likely to have friends who smoke (Deosaransingh et al., 1995; Keefe and
Newcomb, 1996). In addition, alcohol-using Hispanics were less likely to have
friends of the same ethnic group than were non-alcohol users (Kandel and Davis,
1991; Krohn et al., 1996; Krohn and Thornberry, 1993), suggesting this may be a
40
manifestation of acculturation whereby greater acculturation, as represented by
having more friends of different ethnic groups, may be related to increased alcohol
use (Canino et al., 2003; Krohn and Thornberry, 1993).

Peer influence
Several studies and reviews of the literature have established that peer
influence is a significant predictor of adolescent smoking and substance use (Ennett
and Bauman, 1993; Hoffman et al., 2006; Jessor and Jessor, 1977; Kandel, 1978;
Kobus, 2003; Sieving et al., 2000; Sussman et al., 1995a; Valente et al., 2004).
According to social learning theory, learning occurs through modeling, which is
based on the direct observation and imitation of role models’ behavior, or through
vicarious learning and reinforcement (Bandura, 1977). Substance use behaviors may
occur as a result of peers’ modeling substance use, peers making substances more
readily available, peers exerting mutual influence to use substances, and/or peer
norms that encourage substance use (Bandura, 1977; Gaughan, 2003; Kobus, 2003;
Oetting and Donnermeyer, 1998; Perry and Jessor, 1985; Sieving et al., 2000). Peer
relationships are considered the primary factor involved in whether or not youths
decide to engage in and maintain substance use (Ennett and Bauman, 1993, 2000;
Kobus, 2003; Valente, 2003), with peers reinforcing substance use behaviors
(Epstein et al., 1999a). This is indicated in part by substance using youth being more
likely to have substance using friends (Alexander et al., 2001a; Bauman et al., 1984;
Ennett and Bauman, 1993, 1994; Ennett et al., 1994; Rai et al., 2003; Sieving et al.,
41
2000; Sussman et al., 1990; Unger et al., 2001; Windle, 2000). Studies conducted
among specific ethnic groups also confirm this relationship (Dornelas et al., 2005;
Epstein et al., 1999a; Epstein et al., 1999b; Gritz et al., 2003; Unger et al., 2001).
To date, relatively little research has been conducted on the association
between acculturation and substance use among a high-risk youth population
attending continuation high school. One study reported that depressed,
unacculturated, and anxious high-risk Hispanic youth who perceived peer social
influence to use substances were more likely to engage in hard drug use (McCuller et
al., 2001). The current study builds on prior literature by investigating how social
influence mediates the relationship between acculturation and various lifetime and
current substance use outcomes among Hispanic continuation high school students in
Southern California. Study hypotheses include: 1) there are significant differences in
substance use by acculturation level (specifically English language preference youth
will have higher use rates than Spanish language preference youth); 2) linguistic
acculturation is associated with increased risk for substance use; and 3) the
association between language preference and substance use is mediated by peer
social influence.

METHODS
Sample and Data Collection
Data in this study are a subset of baseline data from a substance abuse
prevention intervention program conducted in fourteen Southern California
42
continuation high schools (N = 980). The purpose of the intervention was to test
whether an evidenced-based substance abuse prevention intervention (Towards No
Drug Use) can be made more effective by incorporating interactivity and by
structuring teams based on social networks. Surveys were conducted at baseline, pre-
test, and post-test. Prior to baseline survey administration, parental consent and
student assent were obtained from all participants. The Institutional Review Board at
the University of Southern California approved all study procedures. The sample for
this study was restricted to those who completed the baseline survey and who self-
identified as Latino or Hispanic (including Mexican American, Central American,
and others), yielding a final sample size of 714. Although participants were not asked
about their country of origin, it is likely that many of our respondents are of Mexican
origin because almost 78% of Hispanics in California are of Mexican descent (U.S.
Census Bureau, 2001). It is also likely that among those youth who reported
speaking another language other than English they speak Spanish. Census data
indicate that 65.4% of Californians over the age of 4 who speak another language at
home speak Spanish (Shin and Bruno, 2003). Furthermore, Mexican census data
indicate that less than 7% of the population speak an indigenous language (INEGI,
2006), indicating that Mexican immigrants are more likely to speak Spanish than an
indigenous dialect.

43
Measures
Students completed baseline survey before the beginning of the intervention.
The instrument was a 134-item questionnaire, with items designed to assess various
demographic characteristics, multiple substance use indicators, linguistic
acculturation, peer social influence, and psychological variables known to be
predictive of substance use. The substance abuse prevention intervention from which
these data are a subset of are described in greater detail elsewhere (Valente et al., in
press; Valente et al., submitted). All the measures described below were included in
this study.

Substance use
Measured separately for lifetime and current (past 30 day) use, we asked
participants “How many times have you used each of the drugs below in your
(lifetime or past 30 days)?” The drug categories include: alcohol; tobacco;
marijuana; cocaine or crack; ecstasy; hallucinogens; stimulants or amphetamines;
tranquilizers; opiates; inhalants or vapors; other club/party drugs. Responses were
given on an 11-point rating scale ranging from 0 to 91+ times in increasing intervals
(e.g. 0 times, 1 to10 times, 11 to 20 times, 21 to 30 times, etc.). The reliability and
predictive validity of this format has been previously established (Sussman et al.,
1995a; Sussman et al., 1998). By averaging across responses to cocaine, ecstasy,
hallucinogens, stimulants/amphetamines, tranquilizers, opiates, inhalants, and other
club/party drugs (GHB, ketamine, Rohypnol), a hard drug use index was created (O =
44
0.86 for lifetime use; O = 0.90 for current use). A higher mean on alcohol, tobacco,
marijuana, or hard drug use indicates greater use of those substances.

Linguistic acculturation
Four questions were used to assess linguistic acculturation in our sample.
They included questions concerning language most often read and spoken; most
often spoken at home; language spoken with friends; and the language respondents
preferred to watch movies, TV, and radio shows. Responses were on a five-point
Likert scale with response categories including: only English, English more than
another language, English and another language equally, another language more than
English, another language only (not English). This scale has been previously shown
to have good validity and reliability (Marin et al., 1987). Mean scores ranged from 1
to 5, with lower scores indicating English preference, a marker for higher levels of
acculturation (O = 0.84). This variable was reverse recoded to where a higher mean
value indicates greater English proficiency. Although the acculturation process is a
complex one, involving multiple dimensions, a major component of acculturation is
language use, accounting for a considerable portion of the variance in several
acculturation measures (Cuellar et al., 1980; Epstein et al., 1996; Lessenger, 1997).
In addition, there is sufficient evidence to suggest that language may play an
important role in acculturation and is an efficient proxy for acculturation (Marsiglia
and Waller, 2002; Norris et al., 1996).

45
Peer social influence
Three questions were used to assess peer social influence. They include, how
many of your five best friends have ever tried drugs, use drugs at least once a month,
and would think it is ok for someone to use drugs (Sussman et al., 2000; Sussman et
al., 1995b). Response items were given on a four point scale: none of them, 1 or 2
friends, 3 or 4 friends, or all 5 friends. These items were averaged to create one peer
social influence score (O = 0.83).

Covariates
Variables found to be associated with substance using behaviors were
included as covariates, including gender, age, and socioeconomic status.
Socioeconomic status (SES) was measured with the question: What is the highest
grade completed by your (father or mother)? This question was asked separately
about fathers and mothers. Response options included: 1=not completed elementary
school (8
th
grade); 2=not completed high school (12
th
grade); 3=completed high
school (received diploma); 4=some college or job training (1 to 3 years);
5=completed college (4 years); 6=completed graduate school; and 7=don’t know.
Don’t know was recoded to missing. In order to create one SES score, both of these
items were averaged, where a higher score indicates greater education level and SES
(O = 0.70).

46
Data Analysis
Frequencies, chi-square analyses, and analyses of variance (ANOVA) with
Duncan grouping were used to analyze differences among language preference
groups. After checking for regression assumptions, the substance use variables
(dependent variables) were transformed by their natural log. We conducted mediated
associations by employing the method described by Baron and Kenny (1986). The
outcome variables include lifetime and current substance use, linguistic acculturation
as the predictor, and peer social influence is the potential mediator. We first ran the
multilevel linear regression models testing whether linguistic acculturation
(independent variable) was significantly associated with each of the lifetime and
current substance use outcomes, without the hypothesized mediating variable (peer
social influence). Next, we examined whether linguistic acculturation was associated
with peer social influence (potential mediating variable and the dependent variable in
this model). Finally, we examined whether linguistic acculturation was significantly
associated with each of the lifetime and current substance use outcomes, controlling
for peer social influence in the model. Mediation is supported if: (a) linguistic
acculturation was significantly associated with substance use; (b) linguistic
acculturation was significantly related to the mediator (peer social influence); (c)
peer social influence was significantly associated with substance use; and (d) the
previously significant direct relationship between linguistic acculturation and
substance use became weaker when controlling for peer social influence. If the
relationship between linguistic acculturation and substance use became non-
47
significant after controlling for peer social influence, the mediation effect would be
considered complete; if the association remained significant but its effect was
reduced, then the mediation effect would be considered partial (Baron and Kenny,
1986). We evaluated the mediation pathway with the Sobel test, examining the
hypothesis that the indirect pathway from acculturation to substance use through peer
social influence is different from zero (Sobel, 1982). This last step was conducted
using the SAS programming developed by (Dudley et al., 2004), testing the
mediation effect and the proportion of the variance attributed to the mediation effect
(Beck et al., 2005). Mediation analyses were conducted with multilevel linear
regression, using individuals as the unit of analysis and taking into account the
random-effects resulting from data collected within schools (Singer, 1998).
Regression analyses also controlled for age, gender and SES. A one-tailed alpha of
0.05 was used to determine level of significance and analyses were conducted with
the Statistical Analysis System software (SAS Institute, 1990).

RESULTS
Demographic Characteristics and Substance Use
Table 3.1 illustrates the demographic characteristics of the population overall
and by language preference. The overall mean age was 16.3 years (SD=1.3), with
mean SES of 2.4 (SD=1.1; indicating almost half-way between parents not
completing high school and parents completing a high school diploma), and the
majority of the sample was male (59.2%). When examining differences among the
48
Table 3.1. Selected sample characteristics by language preference (n=714)
Overall English Bilingual Spanish
Only (n=415) (n=246) Only (n=53)
Age (Mean, SD)†* 16.3 (1.3) 16.3 (1.4)
a
16.4 (1.3)
a
16.8 (1.2)
b
SES (Mean, SD)†** 2.4 (1.1) 2.7 (1.1)
a
2.2 (1.0)
b
2.0 (0.6)
b
Gender (n,%)
Male 404 (59.2) 227 (59.3) 144 (58.5) 33 (62.3)
Female 278 (40.4) 156 (40.7) 102 (41.5) 20 (37.7)

Substance Use (n, %)
Lifetime
Alcohol
Never 129 (20.8) 64 (19.5) 51 (21.0) 14 (28.6)
1 – 10 times 159 (25.7) 72 (2.0) 72 (29.6) 15 (30.6)
11 – 30 times 112 (18.1) 59 (18.0) 44 (18.1) 9 (18.4)
31 or more times 220 (35.5) 133 (40.5) 133 (31.3) 11 (22.4)
Mean (SD)† 3.9 (2.4) 4.4 (3.1)
a
3.9 (1.3)
a
3.3 (2.7)
a

Cigarette*
Never 219 (35.7) 119 (36.7) 79 (32.9) 21 (42.9)
1 – 10 times 195 (31.8) 90 (27.8) 87 (36.3) 18 (36.7)
11 – 30 times 77 (12.6) 39 (12.0) 33 (13.7) 5 (10.2)
31 or more times 122 (19.9) 76 (23.5) 41 (17.1) 5 (10.2)
Mean (SD)†* 3.2 (3.1) 3.5 (3.4)
a
3.0 (2.9)
ab
2.4 (2.3)
b

Marijuana**
Never 206 (33.2) 92 (28.2) 92 (37.9) 22 (44.0)
1 – 10 times 132 (21.3) 66 (20.2) 49 (20.2) 17 (34.0)
11 – 30 times 68 (11.0) 34 (10.4) 30 (12.3) 4 (8.0)
31 or more times 214 (34.5) 135 (41.3) 72 (29.6) 7 (14.0)
Mean (SD)†* 3.5 (3.0) 4.4 (3.4)
a
3.6 (3.2)
a
2.6 (2.5)
b

Hard drugs
Never 324 (53.1) 165 (51.1) 128 (53.8) 31 (63.3)
1 – 10 times 131 (21.5) 67 (20.7) 53 (22.3) 11 (22.4)
11 – 30 times 50 (8.2) 31 (9.6) 15 (6.3) 4 (8.2)
31 or more times 105 (17.2) 60 (18.6) 42 (17.6) 3 (6.1)
Mean (SD)†* 1.5 (1.2) 1.6 (1.3)
a
1.5 (1.0)
ab
1.3 (0.7)
b

Current
Alcohol*
Never 245 (39.9) 123 (37.7) 99 (41.3) 23 (47.9)
1 – 10 times 225 (36.6) 120 (36.8) 86 (35.8) 19 (39.5)
11 – 30 times 86 (14.0) 49 (15.1) 34 (14.2) 3 (6.3)
31 or more times 58 (9.5) 34 (10.4) 21 (8.7) 3 (6.3)
Mean (SD)†* 2.1 (1.6) 2.3 (1.9)
a
2.2 (1.9)
a
1.8 (1.1)
b

49
Table 3.1, Continued

Cigarette
Never 385 (63.2) 202 (62.4) 146 (61.6) 37 (77.1)
1 – 10 times 129 (21.2) 62 (19.1) 61 (25.7) 6 (12.5)
11 – 30 times 40 (6.6) 26 (8.0) 12 (5.1) 2 (4.2)
31 or more times 55 (9.0) 34 (10.5) 18 (7.6) 3 (6.2)
Mean (SD)† 2.0 (2.1) 2.2 (2.4)
a
1.9 (2.0)
a
1.6 (1.4)
a

Marijuana*
Never 359 (58.7) 174 (53.5) 147 (61.5) 38 (79.2)
1 – 10 times 117 (19.1) 67 (20.6) 43 (18.0) 7 (14.5)
11 – 30 times 58 (9.5) 33 (10.2) 24 (10.0) 1 (2.1)
31 or more times 78 (12.7) 51 (15.7) 25 (10.5) 2 (4.2)
Mean (SD)†* 2.0 (2.0) 2.5 (2.4)
a
2.1 (2.1)
a
1.5 (1.4)
b

Hard drugs
Never 449 (74.1) 233 (72.1) 172 (73.2) 44 (91.7)
1 – 10 times 93 (15.3) 54 (16.7) 36 (15.3) 3 (6.2)
11 – 30 times 29 (4.8) 15 (4.7) 13 (5.5) 1 (2.1)
31 or more times 35 (5.8) 21 (6.5) 14 (6.0) 0 (0.0)
Mean (SD)† 1.2 (0.8) 1.2 (0.8)
a
1.2 (0.8)
a
1.0 (0.1)
a

† ANOVA with Duncan grouping was used to test for differences among language preference
groups (means with the same letter are not significantly different from one another).
* p<.05; ** p<.001

50
language preference groups, there were small but significant differences in age, with
Spanish speakers being slightly older (16.8 years [SD=1.2]) than bilingual (16.4
years [SD=1.3]) and English speakers (16.3 years [SD=1.4]; p<0.05). Also, English
speakers were more likely to report parents completing their high school diploma
(mean: 2.7 [SD=1.1]) than bilingual (mean: 2.2 [SD=1.0]) and Spanish speakers
(mean: 2.0 [SD=0.6]; p<.001). Although there were no significant gender differences
among the language preference groups, there were significant differences in lifetime
cigarette, marijuana, hard drug use and in current alcohol and marijuana use among
the among the language preference categories, with the differences being greatest
between the English and Spanish groups. On average, as English proficiency
increased, so did self-reported substance use.

Associations between substance use and language preference
Table 3.2 shows the associations between linguistic acculturation and several
lifetime and current substance use indicators. The table presents the unstandardized
betas of these associations, adjusted for age, SES, gender, and random-effects of
schools, but not for peer social influence, the hypothesized mediating variable. These
associations are also depicted in Figure 3.1. For each increase in the level of English
language proficiency, the risk for lifetime alcohol use increases by 11% (p<.05).
Significant associations were also observed with lifetime marijuana (b=0.18; p<.01),
current alcohol (b=0.08; p<.05), current cigarette (b=.08; p<.05) current marijuana
(b=0.15; p<.01), and current hard drug use (b=0.03; p<.05).
51
The top half of Table 3.3 presents the association between linguistic
acculturation and peer social influence. Linguistic acculturation was significantly
associated with peer social influence. So, with each increase in the level of English
language proficiency, the risk for peer social influences to use substances increases
by 44% (p<.05). Age (b=0.34; p<.01) and being female (vs. male: b=-0.54; p<.05)
were also significantly predictive of peer social influence to use substances.
The bottom half of Table 3.3 depicts the final mediation models, with lifetime and
current substance use variables as dependent variables (models were tested
separately for each outcome), acculturation as the independent variable, controlling
for peer social influence (the mediator), age, SES, gender, and random-effects of
schools. Peer social influence was a significant predictor for all of the substance use
outcomes, with greater peer influence associated with greater lifetime and current
substance use. This table also showed that the significant effects of acculturation
observed in Table 2.2 became weaker and non-significant for the outcomes of
lifetime alcohol, current alcohol, current cigarette, and current hard drug use,
indicating complete mediation. The regression weights for acculturation, without the
mediator, decreased after including peer social influence into the model, but
remained significant for the outcomes of lifetime and current marijuana use,
indicating partial mediation. According to the criteria set by Baron and Kenny
(1986), peer social influence met the conditions for being a mediator for the
relationship between acculturation and several substance use outcomes. Results of
52
Table 3.2. Regression weights of lifetime and current substance use
Lifetime Use Current Use
Predictors Alcohol Cigarettes Marijuana Hard Drug Alcohol Cigarettes Marijuana Hard Drug
Age 0.05 0.01 0.07* 0.02 0.00 0.01 0.01 -0.00
SES -0.03 0.03 -0.01 0.04* -0.03 0.02 -0.02 0.01
Female -0.07 -0.01 0.01 -0.05 -0.10 -0.02 0.07 0.00
Acculturation 0.11* 0.07 0.18** 0.03 0.08* 0.08* 0.15** 0.03*
Intraclass Correlation: 0.05 0.02 0.08 0.03 0.08 0.01 0.02 0.01
Note: Multilevel modeling adjusts for random-effects of schools
* p<.05; ** p<.01
53
Figure 3.1. Lifetime and current substance use as a function of linguistic
acculturation among Hispanic continuation high school youths in California
Legend: llifealc llifecig llifemari llifehard
Lifetime Substance Use (log)
0
1
2
3
Linguistic Acculturation
1 23 45
Legend: lcurralc lcurrcig lcurrmari lcurrhard
Current Substance Use (log)
0
1
2
3
Linguistic Acculturation
1 23 45
Note: Linguistic Acculturation 1=Spanish only; 2=Spanish more than English; 3=English and Spanish
equally; 4=English more than Spanish; 5=English only
54
Table 3.3. Regression weights on several outcomes
Predictors ICC
Age SES Female Acculturation Peer Social Influence
Dependent Variable
Peer social influence 0.34** -0.15 -0.54* 0.44* 0.02
Dependent Variables
Lifetime Use
Alcohol
a
0.02 -0.01 -0.01 0.08 0.11** 0.04
Cigarettes
b
-0.03 0.06 0.04 0.03 0.11** 0.02
Marijuana
c
0.01 0.02 0.11 0.13** 0.17** 0.05
Hard Drugs
d
-0.00 0.05** -0.02 0.01 0.01** 0.01
Current Use
Alcohol
e
-0.03 -0.02 -0.06 0.05 0.06** 0.07
Cigarettes
f
-0.02 0.03 0.02 0.03 0.08** 0.01
Marijuana
g
-0.02 -0.00 0.12* 0.10* 0.10** 0.01
Hard Drugs
h
-0.01 0.02* 0.01 0.02 0.02** 0.00
Note: Multilevel modeling adjusts for random-effects of schools; ICC = Intraclass correlation. * p<.05; ** p<.01
a
Sobel value = 2.66, p < .01, percentage of total effect mediated = 41%.
b
Sobel value = 2.53, p < .05, percentage of total effect mediated = 46%.
c
Sobel value = 2.71, p < .01, percentage of total effect mediated = 36%.
d
Sobel value = 2.56, p = .01, percentage of total effect mediated = 46%.
e
Sobel value = 2.68, p < .01, percentage of total effect mediated = 47%.
f
Sobel value = 2.64, p < .01, percentage of total effect mediated = 40%.
g
Sobel value = 2.73, p < .01, percentage of total effect mediated = 33%.
h
Sobel value = 2.38, p < .05, percentage of total effect mediated = 25%.
55
the Sobel test indicated that the indirect effects were significant, accounting for 25%-
47% of the total mediated effect, depending on the type of substance (dependent
variable).

DISCUSSION
Our results indicate that greater English speaking proficiency (linguistic
acculturation) is associated with an increased risk of lifetime use of alcohol and
marijuana, and current (past 30-day) use of alcohol, cigarettes, marijuana, and hard
drugs among a sample of Hispanic youth attending continuation high schools in
Southern California. Our findings are consistent with those for Hispanic youths
attending regular high school (Ebin et al., 2001; Epstein et al., 1998, 2001; Landrine
et al., 1994; Marsiglia et al., 2004; Marsiglia and Waller, 2002; McQueen et al.,
2003; Unger et al., 2000; Vega et al., 1998b). This study builds on previous studies
by using a sample of youth already at high-risk for substance use, and by identifying
peer social influence as a mediator for the association between linguistic
acculturation and substance use. Acculturation was found to have a strong impact on
peer social influence, and peer social influence had a strong impact on all of the
substance use outcomes. The significant associations between linguistic acculturation
and substance use (lifetime alcohol and current alcohol, cigarette, marijuana, and
hard drugs) became non-significant when peer social influence was added to the
model. This finding suggests that peer social influence may, to some extent help
explain the association observed between acculturation and substance use.
56
Our results suggest that as Hispanic youth become more proficient in
speaking English and acculturate to the American youth culture, they may become
more susceptible to peer influence. One explanation for this occurrence may be
related to language and culture. Researchers have hypothesized that language
barriers and being an immigrant may contribute to the relatively homogeneous make-
up of their peer networks (Schweizer et al., 1998). Although our sample is made up
of students of varying degrees of English language proficiency, language may act as
a “vehicle for boundary maintenance,

both supporting the unique cultural identity of
its members

and acting as a barrier to "out-group" interaction and infiltration” (Burr
and Mutchler, 2003: 83). As such, the social networks of Spanish speakers (e.g.
recent immigrants) are comprised mostly of others who speak the same language
(Burr and Mutchler, 2003). However, as Hispanic youths acculturate, their social
networks may expand, giving them greater access to or greater peer influences to use
substances. Conversely, Spanish only speakers may be limited in the number of
peers they can communicate with, forming a relatively tight knit group that interacts
with other Spanish speakers. Non-acculturated peers may inadvertently form a closed
peer group that may protect against pro-substance use peer influences (Unger et al.,
2000). Or, these peer groups may be less inclined to exert pressure on one another to
use substances (Wall et al., 1993). Further, non-English speaking, low-acculturated
families tend to live in ethnic enclaves and attend ethnic-specific social events and
places of worship (Portes, 1990), placing their children in contact with other youths
with similar cultural backgrounds and beliefs (Unger et al., 2000). Doing so perhaps
57
continues to reinforce strong counter-mandates against substance use seen in ethnic
minority groups (Dornelas et al., 2005).
There are also cultural differences with regard to peer versus parent
orientation (Wall et al., 1993; Wills et al., 2004). For instance, in individualistic
societies such as the U.S., adolescence typically involves a separation from parents,
relying more on peers for cues as to appropriate norms and behaviors (Harris, 1995).
Collectivist cultures, on the other hand, may not see this shift because they tend to be
more parent-oriented as opposed to peer-oriented. The Hispanic culture tends to be
more collectivist (Delgado-Gaitan, 1994), with a strong emphasis on family (Marin
and Marin, 1991; Sabogal et al., 1987), and individuation from parents is not an
integral part in collectivist cultures (Brooks et al., 1998). So, unacculturated Hispanic
youths may be more influenced by their parents and less influenced by their peers.
Indeed, evidence exists to support this hypothesis (Brooks et al., 1998; Triandis,
1995; Wall et al., 1993).
To our knowledge, this is the first study conducted among a sample of
Hispanic continuation high school students examining peer social influence as a
potential mediator for the relationship between acculturation and substance use.
However, despite social influence being measured differently, our findings are
consistent with studies conducted with Hispanic youths attending regular schools
(Epstein et al., 2003; Samaniego and Gonzales, 1999; Unger et al., 2000). One study
conducted among a representative sample of California youth attending regular
schools reported that best friends’ smoking (sometimes used in the literature as
58
social influence) mediated the relationship between linguistic acculturation and
smoking among Hispanic youth (Unger et al., 2000). Another study reported that
peer drinking mediated the relationship between linguistic acculturation and
polydrug use in a sample of Hispanic adolescents (Epstein et al., 2003). Finally,
another study reported that negative peer hassles (which included pressure to use
drugs) partially mediated the relationship between acculturation (a combination of
linguistic acculturation and generation status) and delinquency (including drug use)
in a sample of Mexican-American junior high school students (Samaniego and
Gonzales, 1999).

Limitations and Implications
There are several limitations to this research. First, our study only measured
one aspect of the acculturation process. Acculturation is a multi-dimensional process,
involving other attitudes and behaviors (Berry, 2005), which were not assessed.
Because of this limitation, the association is more specifically between linguistic
acculturation (an important aspect of acculturation) and substance use. Nonetheless,
there is sufficient evidence to suggest that language use may be an efficient proxy for
acculturation (Epstein et al., 1996; Marsiglia and Waller, 2002). Future research is
needed where multi-dimensional acculturation scales are included. Doing so may
further our understanding of which acculturation dimensions are associated with
adolescent substance use (Unger et al., 2000).
59
Second, our results are based on youths’ self-report of substance use,
linguistic acculturation and peer social influence, which may have been biased.
However, because our results were confidential, the underreporting bias tends to
diminish. In fact, previous research has reported the validity of self-reported
substance use in survey research (Graham et al., 1984; Harrell, 1997). In addition,
with respect to peer social influence, it should be noted that we measured students’
perceptions of how many of their five best friends use substances and not actual peer
substance use. It is possible that participants over or under estimated the number of
peers who use substances, especially if they themselves have not observed it, but
have been told of peer’s substance use. Nonetheless, this measure has been reliably
used in other studies (Sussman et al., 2000; Sussman et al., 1995b), and some
substance use prevention interventions have targeted perceived substance use norms
in an attempt to affect actual drug use (Botvin et al., 1994; Botvin et al., 1995; Skara
et al., 2005).
Third, these results are cross-sectional therefore no causality can be inferred.
It is possible that peer social influence is a result of substance use (i.e., a desire to
use substances may influence choice of peers), as opposed to the cause. Or,
acculturation could be the mediator of the association between social influence and
substance use. Unacculturated youth looking for acceptance, or youth with good
social skills and/or a desire to learn English may deliberately seek out certain types
of peers (i.e. more acculturated peers, peers that can help them fit in or help translate
the new environment), in turn influencing their own acculturation process. Or,
60
because extended kin tend to live within the same city (84.2% of Hispanics reported
kin living on the same block as they did; Schweizer et al., 1998), and frequent
contact with extended family members is high (Hispanics reported 73.3% of their
social networks is made up of kin; Bernard et al., 1990), perhaps influences to use
substances from U.S.-born acculturated familial peers may prevail over the
protective effects of low acculturation. Future research is needed to test these
alternate hypotheses and to examine the mediation hypothesis prospectively.
Finally, our results may be generalizable only to other Hispanic high risk
youth, especially those attending continuation high schools in Southern California.
Because the majority of Hispanics in the Southern California area are of Mexican
descent, these results may not generalize to Hispanics from different countries of
origin (e.g. Cuba, Guatemala, and Puerto Rico).
Despite these limitations, these findings have potential implications for
school-based substance use prevention programs. Such programs may be made more
effective if they are developed with the acculturation level of the intended audience
in mind, noting that as youth acculturate, their interactions with their peers may
change. Or, with the consideration that the impact peer social influences have on
acculturating youth may vary depending on the level of acculturation. For example,
less acculturated youth may affiliate more with other youth of similar acculturation
level. But substance using behaviors may begin when less acculturated youth move
toward greater acculturation and affiliate more with other acculturated youth.
Although it may not be feasible or cost-effective to tailor programs to unacculturated
61
youth attending large urban schools, whose students are predominantly acculturated,
it may be important to consider tailoring substance use intervention programs to
acculturating youth, whose risk for peer social influences to use substances may
increase with greater levels of acculturation. Future research should also evaluate the
effectiveness of substance use prevention programs among Hispanic youths from
different countries of origin, and with different levels of acculturation.
These results provide new insight about one possible method accounting for
the association between linguistic acculturation and substance use. These results
indicate that acculturation has a strong impact on peer social influences, and these
peer social influences may encompass pro-substance using norms or peer pressure to
use substances. These then may have an impact on substance use. This, and other
factors not measured in this study, may explain the association between linguistic
acculturation and substance use.

62
CHAPTER 4: THE EFFECTS OF NETWORK COMPOSITION AND TIE
STRENGTH ON SUBSTANCE USE: APPLYING SOCIAL NETWORK
METHODS TO A SAMPLE OF HIGH-RISK YOUTH

CHAPTER 4 ABSTRACT
Research suggests that having substance users in one’s network increases the
risk of substance use among youth. However, this relationship is not well understood
in general and not well understood in conditions where the behavior is widespread.
This study examined associations between network composition (drug network and
strength of tie) and substance use among a sample of 968 youth attending
alternative/continuation high schools in Southern California. Data on drug network
composition, strength of tie, and lifetime substance use (alcohol, cigarettes,
marijuana, and cocaine) were available. Results indicate significant ethnic
differences in substance use and network composition, with Whites reporting highest
substance use and generally highest number of substance using peers. Blacks, on
average, reported lowest use and fewer number of substance using peers. The
number of substance users in peer networks (drug network composition) was
associated with risk for lifetime use of alcohol, cigarettes, and marijuana, controlling
for several covariates. Strength of tie (being closer to first or second named peer vs.
third-fifth named peer) was also associated with risk for lifetime cigarette, marijuana,
and cocaine use, controlling for several covariates. Social network methods may
63
assist our understanding of how network composition and peer substance use
patterns affect substance use and the formation of peer networks.

INTRODUCTION
It is estimated that half of all American secondary school students have tried
at least one illicit substance by the time they reach graduation (Johnston et al., 2005).
In a nationwide study of high school students, almost 45% reported having had one
or more alcoholic drinks and nearly 28% reported any cigarette use in the past 30-
days (Grunbaum et al., 2004). With respect to illicit substances, 22.4% reported past
marijuana use, 11.5% reported lifetime use of ecstasy, 7.6% reported lifetime
methamphetamine use, and 4.1% reported past cocaine use (Grunbaum et al., 2004).
The situation is worse for students attending alternative/continuation high
schools (CHS) as they report higher illicit drug use prevalence rates than students
attending regular high school (Grunbaum and Basen-Engquist, 1993; Grunbaum et
al., 2001; Grunbaum et al., 2000; Sussman et al., 1995c). For example, students
attending CHS were significantly more likely than regular high school students
(RHS) to report past 30-day cigarette (70.1% vs. 36.3%), alcohol (64.9% vs. 50.8%),
marijuana (53.9% vs. 26.2%), and cocaine use (15.9% vs. 3.3%; Grunbaum et al.,
2001). In addition, when these high-risk youth use drugs, they are more likely to use
more than one type of drug (Beauvais and Oetting, 1986) and to report early
initiation into alcohol and drug use (Brener and Wilson, 2001).
64
Researchers have long stated that individuals of different ethnic groups and
cultures exhibit varying patterns of substance use. Whites, on average, tend to report
the highest lifetime substance use rates, followed by Hispanics, then Blacks
(Beauvais and Oetting, 2002). Recently, Hispanic youth have equaled, and in some
instances surpassed the substance use rates of Whites. In some parts of the U.S.,
alcohol, crystal methamphetamine, ecstasy (MDMA), and cocaine use is higher
among Hispanic than Black and White youth (Centers for Disease Control and
Prevention, 2004; Johnston et al., 2001; National Institute of Drug Abuse and
University of Michigan, 2004). Among a nationally representative sample of CHS
students, Hispanics reported higher lifetime cocaine use (46.4%) compared to White
(43.8%) and Black (5.7%) students (Grunbaum et al., 1999). Similar patterns were
observed with current cocaine use (19.4%; 17.7%; 3.6%; respectively) and having
tried cocaine before the age of 13 (5.7%; 4.0%; 3.1%; respectively). Another study
indicated that being Hispanic was significantly predictive of hard drug use (McCuller
et al., 2001).

Peer influence
Peer groups and the influence they exert have been used to explain many
adolescent behaviors (Ennett and Bauman, 1993; Jessor and Jessor, 1977; Kandel,
1978). When it comes to smoking and drug use behaviors, peer relationships are
considered one of the primary factors involved in whether or not adolescents decide
to engage in and maintain these behaviors (Ennett and Bauman, 1993, 2000; Kobus,
65
2003; Valente, 2003). Research confirms this by reporting similarities (also known
as homophily) in substance use patterns between adolescents and their peers (Kirke,
2006): 87). In other words, adolescents who use substances are likely to have
substance using friends (Bauman et al., 1984; Ennett and Bauman, 1993, 1994;
Ennett et al., 1994; Rai et al., 2003; Sieving et al., 2000; Sussman et al., 1990; Unger
et al., 2001; Windle, 2000).
The method by which peer groups influence substance use behaviors has
been attributed to peer pressure (Urberg et al., 1991). Also known as influence, this
suggests that peer groups are important antecedents that may contribute to substance
using behaviors (Ennett and Bauman, 1994; Sieving et al., 2000). These behaviors
may occur as a result of peers modeling substance use, peers making substances
more readily available, peers exerting mutual influence to use substances, and/or
peer norms that encourage substance use (Bandura, 1977; Gaughan, 2003; Kobus,
2003; Oetting and Donnermeyer, 1998; Perry and Jessor, 1985; Sieving et al., 2000).
There is another process known as selection that can also help explain
similarities in substance use patterns among peer groups (Bullers et al., 2001).
Selection occurs when adolescents purposefully select and keep friends based on
similar attitudes, beliefs and behaviors as their own, including that of substance use
(Ennett and Bauman, 1994; Kandel, 1978; Sieving et al., 2000). In fact, Kandel
(1985) reported that the formation of adolescent friendships are first based on
sociodemographic characteristics such as age and gender, followed by behaviors
such as illicit drug use. Selection can also occur when friendships are discontinued
66
when peer substance use become dissimilar (Sieving et al., 2000; Urberg et al.,
2003). Although most of the focus has been on influence and its role in adolescent
substance use, studies have increasingly reported that both influence and selection
equally contribute to similarities in substance use patterns among peer groups
(Ennett and Bauman, 1994; Hall and Valente, 2007; Hoffman et al., 2007; Kandel,
1985; Kirke, 2004, 2006; Valente et al., 2004).

Peer influence and social networks
There are a number of theoretical frameworks that have been applied to
explain how peers and social interactions affect behaviors such as alcohol, cigarette,
and drug use (Ennett and Bauman, 1994; Kobus, 2003). However, those that focus
on social processes, like friend selection, interpersonal influence, and behavioral
imitation may provide insight into understanding substance use within the context of
these processes (Kobus, 2003). Social learning theory (SLT) suggests that behavior,
perceptions of behavior, and the environment all interact to influence one another
(Bandura, 1977; Hoffman et al., 2006). According to SLT, behaviors are learned
through observing others modeling the behavior, all the while associating rewards or
punishments with the behavior (Bandura, 1986). Closer bonds as well as bonds
formed earlier in life are important in the social learning process (Kobus, 2003), and
adolescents are more likely to imitate the behavior from peers whom they have the
most contact with (Valente et al., 2004).
67
Social network theory also takes into account social processes, but the focus
is on the relational ties between individuals within a social system (Wasserman and
Faust, 1994). Social network theory assumes that individuals within social systems
interact with each other in varying degrees, with such interactions serving as
significant sources of information and support (Kobus, 2003). These interactions are
also seen as conduits through which information is transmitted throughout the social
system (Kobus, 2003). Although a social system (or social network) is one whose
population, more or less, can be identified by specific boundaries (e.g. a community,
a school, students in a classroom, etc.), social network theory places extreme
importance on the relationship ties among the interacting members of the social
network.
The composition of social networks has the potential to affect members’
behaviors, perceptions, and attitudes because of the varying levels of interaction
among peers (Valente et al., 2004; Wasserman and Faust, 1994). Although social
network studies conducted among high-risk youth have not been completed often,
research conducted among adults indicate that the composition of social networks
has the potential to influence substance use (Gogineni et al., 2001; Latkin et al.,
2004; Latkin et al., 1999; Schroeder et al., 2001). For instance, adults reporting drug
user in their social network is associated with a greater use of illicit substances and
drug overdose (Gogineni et al., 2001; Latkin et al., 2004; Schroeder et al., 2001).
Among adults undergoing drug treatment, those who have substance using network
members were at an increased risk of drug relapse (Havassy et al., 1995). Further,
68
adults who achieved abstinence from drug use were more likely to report fewer drug
users in their network (Latkin et al., 1999; Wasserman et al., 1998).
Another aspect related to network composition has to do with the strength of
tie (or closeness) between network members. A study conducted with sociometric
friendship pairs of Dublin youth reported that youth and their peer networks were
more likely to have similar current alcohol and drug use rates with their close
friends, compared to those who were not close (Kirke, 2006: 82). Another study
conducted among injection drug users participating in the Baltimore Needle
Exchange Program reported that participants engaged in more risk taking, i.e. needle
sharing, with strong-tie close friends (first or second friend named) than with weak-
tie ones (third to fifth friends named; Valente and Vlahov, 2001). Although much
network research has been devoted to reporting the strength of weak ties
(Granovetter, 1973; Granovetter, 1982), many researchers have realized that strong
ties have the capability to exert more influence (Krackhardt, 1992).
Social network theory provides a unique method for studying peer networks
because social network analysis allows for the direct investigation of relationship
patterns within networks (Kirke, 2004; Wasserman and Faust, 1994). Despite the
growing body of research on adolescent social networks and the role peer networks
have on substance use, to date little research has been devoted to examining the
social networks of high-risk youth. One study conducted among runaway and
homeless high-risk youth reported that youth without a social network were
significantly more likely to report current illicit substance use (Ennett et al., 1999).
69
In addition, among youth who did report a social network, their networks tended to
be relatively small, comprised of friends, and the presence of one illicit substance
user in the network significantly increased the risk of current alcohol and illicit
substance use (Ennett et al., 1999). The current study will add to the literature by
investigating the network composition characteristics associated with lifetime
substance use among a sample of high-risk alternative high school youth in Southern
California. Study hypotheses include: 1) there are significant differences in network
composition and substance use outcomes by ethnicity; 2) substance users in one’s
network is positively associated with increased risk for substance use; and 3)
strength of tie is associated with risk for substance use (homophily of use will be
greater for closer friends, being named first or second friend, than for less close
friends, being named third, fourth or fifth). By studying friendship choices through
social network analysis, a greater understanding may be achieved of how friendship
patterns reflect choices made in part by substance use characteristics.

METHODS
Sample and Data Collection
Data in this study are from a substance abuse prevention intervention
program conducted in fourteen Southern California alternative/continuation high
schools. The purpose of the intervention was to test whether an evidenced-based
substance abuse prevention intervention (Towards No Drug Use) can be made more
effective by incorporating interactivity and by structuring teams based on social
70
networks (Valente et al., in press; Valente et al., submitted). Surveys were conducted
at baseline, pre-test, and post-test. Prior to baseline survey administration, active
parental consent and student assent were obtained from all participants. The
University Institutional Review Board approved all study procedures. The sample for
this study was restricted to those who completed the baseline survey (N = 980).

Measures
The instrument was a 134-item questionnaire, with items designed to assess
demographic characteristics, substance use, and psychological variables known to be
predictive of substance use. The survey collected data on two types of friendship
social networks: egocentric and sociometric. This study focused on sociometric
friendship networks. All the measures described below were included in this study.

Substance use
To assess substance use we asked participants, “How many times have you
used each of the drugs below in your lifetime?” The drug categories included in this
study are: alcohol, cigarettes, marijuana, and cocaine. Responses were given on an
11-point rating scale ranging from 0 to 91+ times in increasing intervals (e.g. 0
times; 1 to10 times; 11 to 20 times; 21 to 30 times; etc.). The reliability and
predictive validity of this format has been previously established (Sussman et al.,
1995a; Sussman et al., 1998). All the data on participant and peer substance use were
71
self-reported by them, minimizing the potential exaggeration of perceived substance
use whenever participants are asked about their peers’ use (Kirke, 2006: 35).

Network composition
The sociometric data collection technique provides measures of an entire
social network because all members of the network are interviewed (Valente et al.,
2004). Participants were provided with a class roster and asked to list their five best
friends in the classroom as well as their friend’s respective roster number. We then
calculated the data on the following variables based on friends’ self-report: age,
gender, linguistic acculturation, and substance use (alcohol, cigarette, marijuana,
cocaine). We created a measure representing average network size (range 0-5, where
0=no peers named in network to 5=five peers named in network), network age
(average age of network members), percent male (the percentage of male network
members), network acculturation level (average acculturation level of network
members), and drug network composition (number of substance users in participant’s
network separately by alcohol, cigarettes, marijuana, and cocaine where 0=no
network member used respective substance and 5=all five network members used
respective substance). Using the drug network composition variables, a measure for
strength of tie was created using friendship rank, where 0=third-fifth friend named
and 1=first or second friend named (Valente and Vlahov, 2001).

72
Demographic characteristics
Variables found to be associated with substance use were included as
covariates, including gender, age, ethnicity, socioeconomic status, and acculturation.
Ethnicity was self-reported, with the following response categories: Asian or Asian
American including Chinese, Japanese, and others; Black or African American;
Hispanic or Latino including Mexican American, Central American, and others;
White, Caucasian, Anglo, European American, not Hispanic; American
Indian/Native American; mixed ethnicity (parents are from two or more different
groups); and other. American Indian/Native American and other ethnicity were
recoded to missing due to small sample (n=7; n=5; respectively). Socioeconomic
status (SES) was measured with the question: What is the highest grade completed
by your (father or mother)? This question was asked separately about fathers and
mothers. Response options included: 1=not completed elementary school (8
th
grade);
2=not completed high school (12
th
grade); 3=completed high school (received
diploma); 4=some college or job training (1 to 3 years); 5=completed college (4
years); 6=completed graduate school; and 7=don’t know. Don’t know was recoded to
missing. In order to create one SES score, both of these items were averaged, where
a higher score indicates greater education level and SES (O = 0.70). Because greater
levels of acculturation are associated with greater risk for smoking (Landrine et al.,
1994; Unger et al., 2000), alcohol and substance use among Hispanic populations
(Amaro et al., 1990; De la Rosa, 2002; Epstein et al., 1998, 2001; Farabee et al.,
1995; Marsiglia and Waller, 2002; Nielsen and Ford, 2001; Vega et al., 1998b),
73
acculturation was included as a covariate given our large Hispanic sample (69.0%).
Four questions were used to assess linguistic acculturation, including language most
often read and spoken; language most often spoken at home; language spoken with
friends; and language respondents preferred to watch movies, TV, and radio shows.
Responses were on a five-point Likert scale with response categories including: only
English, English more than another language, English and another language equally,
another language more than English, another language only (not English). This scale
has been previously shown to have good validity and reliability (Marin et al., 1987).
Mean scores ranged from 1 to 5, with lower scores indicating English preference, a
marker for higher levels of acculturation (O = 0.84). This variable was reverse-coded
to where a higher mean value indicates greater English proficiency. Acculturation is
a complex process, involving multiple dimensions. However, a major component of
acculturation is language use, accounting for a considerable portion of the variance
in several acculturation measures (Cuellar et al., 1980; Epstein et al., 1996;
Lessenger, 1997). There is sufficient evidence to suggest that language may play an
important role in acculturation and is an efficient proxy for acculturation (Marsiglia
and Waller, 2002; Norris et al., 1996).

Data Analyses
Frequencies, chi-square, t-test analyses, and analyses of variance (ANOVA)
with Duncan grouping were used to analyze differences among ethnic groups. After
checking for regression assumptions, the dependent variables (lifetime alcohol,
74
cigarettes, marijuana, and cocaine) were transformed by their natural log (Frank,
1966). The association between these substances and network composition (drug
network and strength of tie) was then analyzed using three-level linear regression
analyses. Multilevel linear regressions adjusted for age, gender, SES, acculturation,
ethnicity, and other network composition characteristics (network size, network age,
percent male, and network acculturation level). All analyses used individuals as the
unit of analysis, taking into account the random-effects resulting from data collected
in classrooms within schools (Singer, 1998). A one-tailed alpha of 0.05 was used to
determine level of significance using the Statistical Analysis System software (SAS
Institute, 1990).

RESULTS
Demographic characteristics, substance use, and network composition
Table 4.1 illustrates the demographic, substance use and network composition
characteristics of the population overall and by ethnicity. The overall mean age was
16.3 (SD=1.3), with mean SES of 2.8 (SD=1.2; indicating parents completing a high
school diploma). The majority of the sample was male (60.3%) and Latino/Hispanic
(69%). Most participants were in the 11
th
-12
th
grade (58.0%), and predominantly
English only speaking (67.4%), however 27.1% were bilingual. A greater percent of
participants reported lifetime use of alcohol and marijuana more than 30 times
(38.1%; 39.9%; respectively). Over one-third of participants reported never having
smoked a cigarette (34.7%) while 74% never tried cocaine. Average network size
75
was 3.1 (SD=2.0) with a network acculturation level made up mostly of English only
speakers (74.1%). Drug network composition indicates over half of participants had
1-2 friends who use alcohol (58.7%), cigarettes (55.9%), and marijuana (55.4%),
while 74.5% had no cocaine users in their network.
There were significant differences by ethnicity when examining substance use.
A greater proportion of Whites consumed alcohol 31 or more times during their
lifetime (62.6%) versus about one-third of Asian (39.2%), Black (26%), Hispanic
(35.1%), and mixed ethnicity (33.4%) students (p<.001). A greater proportion of
Asian, White, and mixed ethnicity students were more likely to report lifetime
consumption of cigarettes 31 or more times (56.5%; 58.1%, 42.0%; respectively)
whereas Black and Hispanic students were more likely to report never using
cigarettes (58.5%; 35.6%; respectively; p<.001). Black, White, and mixed ethnicity
students were more likely to have lifetime use of marijuana 31 or more times
(48.2%; 62.6%; 46.4%) compared to Asian and Hispanic students (34.8%; 33.7%;
p<.001). About one-third of Asian and Hispanic students reported never having used
marijuana (30.4%; 33.2%; respectively). Although most student reported never used
cocaine, Whites were significantly more likely than the other ethnic groups to report
greater cocaine use (p<.01).
76
Table 4.1. Demographics, substance use characteristics, and network composition stratified by ethnicity
Total Asian Black Hispanic White Mixed
(n=968) (n=23) (n=54) (n=591) (n=108) (n=69)
Age
12-14 103 (10.9) 3 (13.0) 7 (13.2) 63 (10.7) 7 (6.5) 14 (20.3)
15-17 673 (71.4) 16 (69.6) 40 (75.5) 424 (72.2) 78 (72.2) 43 (62.3)
18-21 166 (17.6) 4 (17.4) 6 (11.3) 100 (17.1) 23 (21.3) 12 (17.4)
Mean (SD) 16.3 (1.3) 16.4 (1.2) 16.2 (1.2) 16.3 (1.4) 16.5 (1.2) 16.0 (1.5)
SES (Mean, SD)** 2.8 (1.2) 4.0 (1.2)
a
3.6 (1.0)
a
2.4 (1.1)
b
3.6 (1.1)
a
3.6 (1.2)
a
Gender (n,%)
Male 570 (60.3) 17 (73.9) 35 (66.0) 346 (58.7) 64 (59.3) 41 (60.3)
Female 376 (39.7) 6 (26.1) 18 (34.0) 243 (41.3) 44 (40.7) 27 (39.7)
Grade (n, %)
7-8 41 (4.4) 0 (0.0) 3 (5.8) 27 (4.6) 4 (3.7) 5 (7.2)
9-10 353 (37.6) 7 (30.4) 13 (25.0) 231 (39.6) 33 (30.6) 30 (43.5)
11-12 544 (58.0) 16 (69.6) 36 (69.2) 326 (55.8) 71 (65.7) 34 (49.3)
Acculturation**
English only 660 (67.4) 14 (60.9) 51 (94.4) 298 (50.4) 104 (96.3) 65 (94.2)
Bilingual 266 (27.1) 8 (34.8) 3 (5.6) 242 (41.0) 4 (3.7) 4 (5.8)
Another language only 54 (5.5) 1 (4.3) 0 (0.0) 51 (8.6) 0 (0.0) 0 (0.0)
Mean (SD)** 4.2 (0.8) 4.1 (0.7)
a
4.9 (0.4)
b
3.8 (0.7)
a
4.9 (0.3)
b
4.7 (0.4)
b
Substance use
Alcohol**
Never 165 (18.6) 3 (13.0) 16 (29.6) 121 (20.7) 4 (3.7) 10 (14.5)
1-10 times 223 (25.2) 3 (13.0) 17 (31.4) 154 (26.4) 21 (19.6) 19 (27.5)
11-30 times 160 (18.1) 8 (34.8) 7 (13.0) 104 (17.8) 15 (14.1) 17 (24.6)
31 or more times 337 (38.1) 9 (39.2) 14 (26.0) 205 (35.1) 67 (62.6) 23 (33.4)
77
Table 4.1, Continued
Cigarettes**
Never 304 (34.7) 7 (30.4) 31 (58.5) 206 (35.6) 25 (23.8) 19 (27.6)
1-10 times 230 (26.3) 3 (13.1) 8 (15.1) 189 (32.6) 9 (8.6) 12 (17.4)
11-30 times 106 (12.1) 0 (0.0) 6 (11.3) 74 (12.8) 10 (9.5) 9 (13.0)
31 or more times 235 (26.9) 13 (56.5) 8 (15.1) 110 (19.0) 61 (58.1) 29 (42.0)
Marijuana**
Never 268 (30.3) 7 (30.4) 18 (33.3) 194 (33.2) 19 (17.8) 14 (20.3)
1-10 times 171 (19.3) 5 (21.7) 8 (14.8) 128 (21.9) 11 (10.3) 14 (20.3)
11-30 times 93 (10.5) 3 (13.1) 2 (3.7) 65 (11.2) 10 (9.3) 9 (13.0)
31 or more times 353 (39.9) 8 (34.8) 26 (48.2) 197 (33.7) 67 (62.6) 32 (46.4)
Cocaine*
Never 646 (74.0) 18 (78.3) 51 (94.4) 429 (74.2) 63 (61.2) 52 (76.5)
1-10 times 133 (15.2) 2 (8.7) 2 (3.7) 85 (14.7) 25 (24.3) 10 (14.7)
11-30 times 41 (4.7) 3 (13.0) 0 (0.0) 24 (4.2) 8 (7.7) 4 (5.9)
31 or more times 53 (6.1) 0 (0.0) 1 (1.9) 40 (6.9) 7 (6.8) 2 (2.9)
Network composition
Network size (mean, SD)* 3.1 (2.0) 3.2 (2.1)
ab
2.9 (2.1)
b
3.4 (1.9)
ab
3.7 (1.6)
a
3.4 (1.9)
ab
Network age (mean, SD) 16.4 (1.2) 16.3 (0.8)
a
16.3 (1.2)
a
16.4 (1.2)
a
16.6 (1.0)
a
16.2 (1.3)
a
Number of male network members
0 male 157 (21.8) 4 (21.1) 6 (15.8) 114 (24.1) 10 (10.4) 12 (21.4)
1-2 male 371 (51.5) 10 (52.6) 20 (52.6) 237 (50.1) 56 (58.3) 32 (57.2)
3-5 male 192 (26.7) 5 (26.3) 12 (31.6) 122 (25.8) 30 (31.3) 12 (21.4)
Network acculturation level
English only 726 (74.1) 20 (87.0) 51 (94.4) 383 (64.8) 102 (94.4) 52 (75.4)
Bilingual 228 (23.3) 3 (13.0) 3 (5.6) 184 (31.1) 6 (5.6) 17 (24.6)
Another language only 26 (2.6) 0 (0.0) 0 (0.0) 24 (4.1) 0 (0.0) 0 (0.0)
Mean (SD)** 4.1 (0.7) 4.5 (0.4)
bc
4.6 (0.4)
ab
4.0 (0.7)
d
4.6 (0.4)
a
4.3 (0.5)
c
78
Table 4.1, Continued
Drug network composition
Number of alcohol drinkers in network**
0 friends 101 (14.2) 1 (5.3) 8 (21.6) 72 (15.4) 6 (6.3) 7 (12.5)
1 – 2 friends 417 (58.7) 11 (57.9) 20 (54.1) 287 (61.6) 40 (42.1) 36 (64.3)
3 – 5 friends 193 (27.1) 7 (36.8) 9 (24.3) 107 (23.0) 49 (51.6) 13 (23.2)
Mean (SD)** 1.8 (1.3) 2.4 (1.4)
a
1.6 (1.3)
b
1.7 (1.2)
b
2.6 (1.4)
a
1.8 (1.3)
a
Number of smokers in network**
0 friends 205 (28.9) 1 (5.3) 12 (32.4) 146 (31.5) 15 (15.8) 16 (28.6)
1 – 2 friends 396 (55.9) 12 (63.2) 19 (51.4) 267 (57.5) 48 (50.5) 30 (53.5)
3 – 5 friends 108 (15.2) 6 (31.5) 6 (16.2) 51 (11.0) 22 (33.7) 10 (17.9)
Mean (SD)** 1.3 (1.2) 2.1 (1.1)
a
1.2 (1.1)
b
1.1 (1.0)
b
1.9 (1.4)
a
1.3 (1.2)
b
Number of marijuana users in network**
0 friends 180 (25.3) 2 (10.5) 8 (21.6) 135 (29.0) 9 (9.5) 13 (23.2)
1 – 2 friends 393 (55.4) 11 (57.9) 24 (64.9) 259 (55.7) 45 (47.4) 32 (57.2)
3 – 5 friends 137 (19.3) 6 (31.6) 5 (13.5) 71 (15.3) 41 (43.1) 11 (19.6)
Mean (SD)** 1.5 (1.3) 2.1 (1.5)
a
1.3 (1.0)
b
1.3 (1.1)
b
2.3 (1.5)
a
1.5 (1.2)
b
Number of cocaine users in network*
0 friends 529 (74.5) 13 (68.4) 29 (78.4) 354 (76.1) 59 (62.1) 46 (82.1)
1 – 2 friends 173 (24.4) 6 (31.6) 8 (21.6) 103 (22.2) 36 (37.9) 10 (17.9)
3 – 5 friends 8 (1.1) 0 (0.0) 0(0.0) 8 (1.7) 0 (0.0) 0 (0.0)
Mean (SD) 0.3 (0.6) 0.3 (0.5)
a
0.2 (0.5)
a
0.3 (0.7)
a
0.4 (0.6)
a
0.3 (0.6)
a
Note: Chi-square, t-test, or ANOVA with Duncan grouping used to test for ethnic differences (means with the same letter are not significantly different)
* p<.05; **p<.001
79
Network composition also differed significantly by ethnic group. Blacks
reported the smallest average network size (2.9) while Whites reported the highest
(3.7; p<.05). White and Black students had similarly high network acculturation
level (mean: 4.6; 4.6; respectively). A greater proportion of Hispanics reported less
acculturated (4.1%) and bilingual (31.1%) network members compared to other
ethnic groups (p<.001). Black and Hispanic youth had fewer network members that
use alcohol (mean: 1.6; 1.7; respectively) compared to Asian, White, and mixed
ethnicity students (p<.001). There were significantly greater cigarette and marijuana
users among the network members of Asian and White students, compared to other
ethnic groups (p<.001). There were no significant mean differences in the number of
cocaine using network members among ethnic groups, but 37.9% of White students,
compared to Asian (31.6%), Black (21.6%), Hispanic (22.2%), and mixed ethnicity
(17.9%) reported 1-2 friends who used cocaine (p<.05).

Associations between drug network composition and substance use
Table 4.2 reports the unstandardized betas for the association between drug
network composition and substance use, adjusted for age, SES, gender, ethnicity,
acculturation, network composition variables (network size, network age, percent
male, network acculturation level), and random-effects of classrooms nested within
schools. For each increase in one network member who uses alcohol, the risk for
lifetime alcohol use increased by 10% (p<.01). Other significant predictors of
lifetime alcohol use were White ethnicity (vs. mixed ethnicity; b=0.39), acculturation
80
Table 4.2. Associations between drug network composition and substance use

Alcohol Cigarettes Marijuana Cocaine
b b b b
Age 0.01 0.00 0.03 0.03
SES -0.02 0.02 -0.03 0.01
Male 0.02 0.00 0.04 -0.07
Asian 0.03 0.16 -0.19 -0.06
Black -0.20 -0.63** 0.13 -0.27*
Hispanic 0.03 -0.26* -0.03 0.01
White 0.39* 0.18 0.27* 0.10
Acculturation 0.10* 0.05 0.17* -0.05
Network composition
Network size -0.07* -0.07* -0.05 0.03
Network age 0.10* 0.06 0.10* 0.01
Percent male -0.00 -0.00 0.00 -0.00
Network acculturation level 0.02 0.10 0.09 0.09*
Drug network composition 0.10** 0.16** 0.13** 0.01

ICC: 0.02 0.00 0.06 0.02
Note: Multilevel modeling adjusts for random-effects of classrooms nested within schools
ICC: Intraclass correlation
* p<.05; ** p<.01

81
(b=0.10), smaller network size (b=-0.07), and network age (b=0.10). For each
increase in one network member who uses cigarettes, the risk for lifetime cigarette
use increased by 16% (p<.01). Other significant predictors of lifetime cigarette use
include non-Black ethnicity (vs. mixed ethnicity; b=-0.63), non-Hispanic ethnicity
(vs. mixed ethnicity; b=-0.26), and smaller network size (b=-0.07). An increase in
one network member who uses marijuana, the risk for lifetime marijuana use
increased by 13% (p<.01). Lifetime cocaine use was also significantly associated
with White ethnicity (vs. mixed ethnicity; b=0.27), acculturation (b=0.17), and
network age (b=0.10). Drug network composition was not significantly associated
with lifetime cocaine use, but significant predictors include non-Black ethnicity (vs.
mixed ethnicity; b=-0.27) and network acculturation level (b=0.09). We replicated
these analyses with egocentric data and the results indicate drug network
composition was positively and significantly associated with greater alcohol,
cigarette, and other drug use (results not shown).

Associations between strength of tie and substance use
Table 4.3 shows the association between strength of tie and substance use
outcomes. Lifetime alcohol use was not significantly associated with strength of tie,
but was associated with White ethnicity (vs. mixed ethnicity; b=0.47), acculturation
(b=0.12), and network age (b=0.07). Lifetime cigarette use was significantly
associated with having close friends who smoke (first or second friend named vs.
those named third-fifth; b=0.36). Other significant predictors include non-Black
82
Table 4.3. Associations between strength of tie and substance use

Alcohol Cigarettes Marijuana Cocaine
b b b b
Age 0.02 -0.00 0.04 0.02
SES -0.03 0.03 -0.01 0.05
Male -0.03 -0.02 0.03 -0.03
Asian 0.10 0.37 -0.12 -0.32
Black -0.08 -0.46* 0.23 -0.27
Hispanic 0.08 -0.13 0.05 -0.04
White 0.47** 0.35* 0.29* 0.00
Acculturation 0.12* 0.08 0.18* -0.10
Network composition
Network size -0.04 -0.04 -0.03 -0.08
Network age 0.07* 0.02 0.09 -0.03
Percent male -0.00 -0.00 0.00 0.00
Network acculturation level 0.03 0.12 0.19* 0.10
Strength of tie
Third-fifth friend named (reference)
First or second friend named 0.08 0.36* 0.20* 0.19*

ICC: 0.04 0.02 0.11 0.11
Note: Multilevel modeling adjusts for random-effects of classrooms nested within schools.
ICC: Intraclass correlation
* p<.05; ** p<.01
83
ethnicity (vs. mixed ethnicity; b=-0.46) and White ethnicity (vs. mixed ethnicity;
b=0.35). Lifetime marijuana use was significantly associated with having close
friends who use marijuana (first or second friend named vs. those named third-fifth;
b=0.20). Other significant predictors of lifetime marijuana use included White
ethnicity (vs. mixed ethnicity; b=0.29), acculturation (b=0.18), and network
acculturation level (b=0.19). Lifetime cocaine use was significantly associated with
having cocaine-using first or second named friends (vs. those named third-fifth;
b=0.19). We replicated these analyses with egocentric data and the results indicate
strength of tie (being closer to first or second friend named vs. those named third-
fifth) was positively and significantly associated with greater alcohol, cigarette, and
other drug use (results not shown).

DISCUSSION
Researchers have long stated that individuals of different ethnic groups
exhibit varying patterns of substance use. Indeed, the results of this study indicate
significant differences in lifetime alcohol, cigarette, marijuana, and cocaine use by
ethnicity. Whites reported the highest while Blacks reported the lowest use of all
substances. For alcohol use, the next highest users were Asian students, followed by
students of mixed ethnicity. Lifetime cigarette and marijuana use was highest for
Whites, followed by mixed ethnicity and Asian students. After Whites, lifetime
cocaine use was next highest for Hispanic, followed by mixed ethnicity students. In
84
addition, some network composition variables (network size, network acculturation
level, and drug network composition) differed by ethnicity.
Regression analyses further indicated ethnicity as a significant predictor of
lifetime substance use. In Table 4.2, compared to mixed ethnicity, Whites were at
risk for greater alcohol and marijuana use. Blacks, on the other hand, were at a
decreased risk for cigarette and cocaine use. In addition, being Hispanic was
protective for cigarette use. In Table 4.3, we found a similar pattern in that White
students were at a greater risk for lifetime alcohol, cigarette, and marijuana use,
whereas Black students were at lower risk for cigarette use. Although some of the
literature on ethnic differences among regular high school students contradict our
findings, our results are corroborated by others (Johnston et al., 2006; O'Malley et
al., 2006; Wallace et al., 2003).
With regards to network composition, our results indicate that drug network
composition (i.e. greater number of substance using peers) is associated with an
increased risk for lifetime use of alcohol, cigarettes, and marijuana among a sample
of high-risk youth attending alternative high schools in Southern California. After
adjusting for several covariates, we report that for every additional substance using
network member in a participant’s social network, the risk of lifetime alcohol,
cigarette, and marijuana increases by 10%, 16%, and 13%, respectively. Our findings
are consistent with others (Alexander et al., 2001b; Gogineni et al., 2001; Haynie,
2001; Krohn and Thornberry, 1993; Latkin et al., 1999; Schroeder et al., 2001), and
with other risk behaviors including needle sharing (Latkin et al., 1996; Valente and
85
Vlahov, 2001), using shooting galleries (Suh et al., 1997), and drug overdose (Latkin
et al., 2004). Possible explanations for these results could be that substance-using
network members may reinforce substance use, provide more influence and
opportunities to use drugs, or provide more avenues for individuals to increase
interactions with other substance users (Latkin et al., 1999).
Of interest is our result that reporting a smaller network size was inversely
related to lifetime alcohol and cigarette use. Our results are corroborated Valente et
al. (2005) who reported that naming fewer friends was associated with more
susceptibility to smoking. One possible explanation for this finding could be that a
greater number of network members serve a role in providing social support to
participants. Although we did not ask specifically whether network members serve
as a source of social support, nor the type of social support (i.e. instrumental,
emotional, financial, etc.) provided by network members, the stress and coping
literature indicate that if stressors exceed an individual’s coping skills, and if the
individual considers the stressors as uncontrollable, the individual may engage in
rebellion, delinquency and/or drug use (Unger et al., 2004; Vega et al., 2003).
Perhaps a larger network provides individuals a greater number of people with whom
he/she can use as a source of social support, thereby reducing their risk for substance
use.
Or, these findings could support the hypothesis that isolates are associated
with greater substance use (Ennett and Bauman, 1993, 1994). Because classroom
sizes in alternative high schools tend to be smaller compared to regular high school
86
classrooms, and because CHS are comprised of highly mobile students (California
Department of Education, 2006), there is a greater likelihood that participants have
more friends outside of the school. Results of egocentric data analysis indicated that
an increase in one friend outside of school significantly increased the risk of lifetime
alcohol and marijuana use by 4.2% and 4.1%, respectively (results not shown).
Another study reported that individuals with smaller networks have more friends
outside the classroom or outside of school (Valente et al., 2005). It is possible that
within classrooms, those with smaller networks could be considered social isolates
(individuals with little contact with other network members; Wasserman and Faust,
1994). Previous studies have reported higher rates of cigarette use among social
isolates (Ennett and Bauman, 1993, 1994).
Network age, as expected, was a significant predictor of lifetime alcohol and
marijuana use. The importance of this finding derives from the literature indicating
that older adolescents are more likely than younger ones to report substance use.
However, when examining substance use within the context of their peer groups, it is
important to consider the ages of network members. Our results indicate that having
older network members in one’s peer group is associated with greater risk for alcohol
and marijuana. As Kirke (2006: 152) reports, even small differences in age may
expose younger adolescents to the peer influences of older adolescents, who are
more likely to use substances.
We have also shown that the risk for lifetime substance use among high-risk
alternative high school youth was more likely to occur with friends with whom
87
participants have a stronger tie. A study conducted among injection drug users
participating in the Baltimore Needle Exchange Program reported that participants
engaged in more risk taking, i.e. needle sharing, with strong-tie close friends (first or
second friend named) than with weak-tie ones (third to fifth friends named; Valente
and Vlahov, 2001). Although strength of tie was not specifically asked about in the
instrument, these results point to the usefulness of incorporating such variables into
future studies.
Given the importance of acculturation on substance use, our results indicate
that for high-risk youth attending alternative high schools, acculturation is predictive
of lifetime alcohol and marijuana use. Research has reported that greater
acculturation levels are associated with greater cigarette (Landrine et al., 1994;
Unger et al., 2000), alcohol, and substance use (Amaro et al., 1990; De la Rosa,
2002; Farabee et al., 1995; Marsiglia and Waller, 2002; Nielsen and Ford, 2001). A
possible explanation for this could be that since acculturation to the U.S. culture
often involves improved English speaking proficiency, in so doing, youths’
interactions with their peers change (Unger et al., 2000). Acculturating youth may be
able to communicate with peers they once had been unable to, potentially accessing a
larger or more diverse network of peers, one that reaches beyond the immigrant peer
group (Wall et al., 1993). As a result, youth with greater English speaking
proficiency may come across more pro-substance peer influences, including
perceiving substance use (including alcohol and tobacco) as a normative behavior
(Rai et al., 2003; Rice et al., 2003; Unger et al., 2000; Urberg et al., 1997; Windle,
88
2000). Hence, as youth acculturate, their association with U.S.-born peers may
increase, causing them to perceive substance use as normative, and they may perhaps
find themselves in circumstances where peers offer or are using substances (Unger et
al., 2000). Empirical evidence suggests that level of acculturation may affect the
degree to which youth are susceptible to peer influences to use substances (Fosados
et al., under review-a; Fosados et al., under review-b; Wall et al., 1993). Future
studies and school-based substance use prevention programs may be more effective
if they are developed with consideration of the impact peer influences have on
acculturation, especially when working with a multi-ethnic population.

Limitations and Implications
This study suggests that social network characteristics, in particular those
related to peer influence such as drug network composition and strength of tie, are
significant predictors of lifetime substance use. These results highlight the
importance of incorporating social network indicators when conducting research on
youth substance use. However, there are a number of notable limitations with this
study.
First, our results are based on youths’ self-report of substance use, which may
have been biased. However, because our results were confidential, the
underreporting bias tends to diminish. In fact, previous research has reported the
validity of self-reported substance use in survey research (Graham et al., 1984;
89
Harrell, 1997). In addition, our measure has been reliably used in other studies with
similar populations (Sussman et al., 2000; Sussman et al., 1995b).
Second, because acculturation is a multi-dimensional process involving other
attitudes and behaviors (Berry, 2005), our acculturation measure is limited in that it
is based on one single aspect, language preference. Nonetheless, there is sufficient
evidence to suggest that language use may be an efficient proxy for acculturation
(Epstein et al., 1996; Marsiglia and Waller, 2002).
Third, these results are cross-sectional therefore no causality can be inferred.
It is possible that drug network composition is a result of one’s substance use, where
participants purposefully seek out others with whom to use substance with (i.e.,
selection). Or, non-using participants could have been influenced by their drug
network (i.e. influence), whose composition was initially based on some other factor
other than substance use. Although peers may reinforce each other’s substance use
even if peers were not involved in each other’s initiation (Kirke, 2006: 130), future
research is needed to understand the formation of peer networks among high-risk
youth attending alternative high schools.
Finally, our results may be generalizable only to other high-risk youth,
especially those attending alternative/continuation high schools in Southern
California. Because the majority of the sample was Hispanic/Latino, and because
most Hispanics from the Southern California area are of Mexican descent, these
results may not generalize to Hispanics from different countries of origin (e.g. Cuba,
90
Guatemala, and Puerto Rico). Future research should sample high-risk youth from
other parts of the country with diverse ethnic groups.
Despite these limitations, these findings have potential implications for
alternative high school-based substance use prevention programs. Although
prevention programs have reported significant behavioral effects on substance use
among a high-risk alternative/continuation school population (Sussman et al., 1998;
Sussman et al., 2003), promoting the development of ties with non-users can assist
with preventing initiation or maintaining cessation from substances in the long-term.
Doing so may decrease drug availability, the number and frequency of substance
using cues, and the pressure by network members to use substances. However, these
options may be difficult to achieve among this type of setting. Prevention efforts
may benefit by focusing on changing the pro-substance use environment observed
among these types of schools to one that promotes anti-substance using norms.
91
CHAPTER 5: CONCLUSION

These three studies, encompassing two different datasets, shed light on the
importance of peer social influence in substance use among a high-risk sample of
alternative/continuation high school students in Southern California. The first study,
conducted among a sample of regular and alternative/continuation high school
(CHS) students, indicated that linguistic acculturation (i.e., greater English speaking
proficiency) was associated with greater substance use, and peer social influence
mediated the relationship. Specifically, among Hispanic CHS students, acculturation
was associated with greater risk for lifetime use of cigarettes, marijuana, and hard
drugs, and current (past 30-day) use of cigarettes and marijuana. Among Hispanic
regular high school (RHS) students, linguistic acculturation was associated with
greater use of all lifetime and current use of alcohol, cigarettes, marijuana, and hard
drugs. Acculturation was associated with peer social influence, and peer social
influence was associated with substance use. According to the criteria set by Baron
and Kenney (1986), peer social influence met the conditions for being a mediator for
the relationship between acculturation and some substance use outcomes.
The second study, conducted among a Hispanic sample of CHS students in
Southern California, indicated that greater English speaking proficiency (linguistic
acculturation) is associated with increased risk of lifetime alcohol and marijuana, and
current (past 30-day) use of alcohol, cigarettes, marijuana, and hard drugs.
Acculturation was found to have a strong impact on peer social influence, and peer
92
social influence had a strong impact on all of the substance use outcomes. Mediation
analyses indicated that the significant associations between linguistic acculturation
and substance use became non-significant when peer social influence was added to
the model. Our results suggest that as Hispanic youth become more proficient in
speaking English and acculturate to the American youth culture, they may become
more susceptible to peer influences to use substances. And this susceptibility may
increase their risk for substance use.
The third study, using the same dataset as study two but including several
ethnic groups, was valuable because it allowed for the examination of peer social
influence with social network methods. Due to the sociometric nature of the dataset,
all data on participant and peer substance use were self-reported by them,
minimizing the potential exaggeration of perceived substance use whenever
participants are asked about their peers’ use (Kirke, 2006). Focusing on sociometric
friendship networks, this study found that drug network composition (i.e. greater
number of substance using peers) was associated with greater risk for lifetime
alcohol, cigarette, and marijuana use. Further, the risk for lifetime substance use was
more likely to occur with friends with whom participants have a stronger tie.
The results of these separate studies concerning different populations at high-
risk for substance use suggest that peer social influence, including the composition of
peer networks, is important in substance use. Also, one’s acculturation level appears
to affect susceptibility to peer influences and subsequent substance use. In one part,
these results are probably indicative of how being able to speak English may allow
93
students to access previously inaccessible peer groups. Once accessed, ensuing
interactions with peers may expose previously non-English speakers to new
substance using norms and influences that may lead to substance use. This is
reflected in the fact that our results show that Spanish speakers use substances less
than half the rate English speakers do. However, because acculturation is learned
within the context of peer groups, it is also important to understand how the make-up
of peer groups may accelerate or deter subsequent substance use.
Several limitations to these studies must be noted. First, our results are based
on cross-sectional data therefore no causality can be inferred. It is possible that
acculturation mediates the relationship between peer social influence and substance
use. It is also possible that substance use can lead to peer social influence, as in the
case of selection where individuals purposefully select peers who use substances.
Second, our results are based on self-report. Although substance use may have been
under-reported, this bias tends to diminish because the data were confidential. In
addition, other studies have reported the validity of self-reported substance use, and
all of our measures have been reliably used in other studies (Graham et al., 1984;
Harrell, 1997; Sussman, Dent and Leu, 2000; Sussman et al., 1995b). Third,
acculturation was based on one aspect, language preference. Acculturation is
complex, involving several dimensions; language preference assesses one. Yet,
enough evidence exists to suggest that language is an important part of the
acculturation process and therefore is an efficient proxy for acculturation (Epstein et
al., 1996; Marsiglia and Waller, 2002). Finally, our results have limited
94
generalizability, in particular to youth attending alternative/continuation high schools
in Southern California. Because the majority of Hispanics in Southern California are
of Mexican descent, these results may not generalize to Hispanics from different
countries of origin (e.g. Puerto Rico, Cuba, Guatemala, etc.).
Despite these limitations, these findings can prompt future research. Future
research should conduct longitudinal studies examining how acculturation and other
cultural factors influence susceptibility to peer influence among diverse ethnic
groups, geographic regions, and school types. Considering research with populations
at high-risk for substance use remains somewhat limited, future research should also
explore the role of network position, network structural characteristics (e.g. network
density, centrality, etc.), and differences in homophily and peer nomination patterns
between users and nonusers. The literature indicates that network-level indicators
may help explain how interpersonal interactions among peer groups, especially the
communication of substance use, occur among network members. Incorporating such
aspects into future studies will help improve our understanding of the influence of
peer networks on substance use behavior.
The primary aim of this project was to understand how peer influences,
acculturation, and social networks influence substance use. The hope is that these
three studies, each with their own focus, provides valuable information to researchers
interested in curtailing adolescent substance use. A positive contribution of this
research is that it features a high-risk youth population, and the results enhance our
95
understanding of the role peer influence, culture, ethnicity, and social networks have
on substance use among an infrequently studied population.
96
REFERENCES

Alaniz, M. L. (2002) Migration, acculturation, displacement: migratory workers and
"substance abuse". Subst Use Misuse 37, 1253-7.

Alexander, C., Piazza, M., Mekos, D. and Valente, T. W. (2001a) Peer networks and
adolescent cigarette smoking: An analysis of the national longitudinal study
of adolescent health. Journal of School Health 29, 22-30.

Alexander, C., Piazza, M., Mekos, D. and Valente, T. W. (2001b) Peers, schools, and
adolescent cigarette smoking. Journal of Adolescent Health 29, 22-30.

Amaro, H., Whitaker, R., Coffman, G. and Heeren, T. (1990) Acculturation and
marijuana and cocaine use: findings from HHANES 1982-84. Am J Public
Health 80, 54-60.

Bandura, A. (1977) Social learning theory. Prentice Hall, Englewood Cliffs, NJ.

Bandura, A. (1986) Social foundations of thought and action: A social congnitive
theory. Prentice Hall, New Jersey.

Baron, R. M. and Kenny, D. A. (1986) The moderator-mediator variable distinction
in social psychological research: Conceptual, strategic, and statistical
considerations. Journal of Personality and Social Psychology 51, 1173-1182.

Barsevick, A. M., Dudley, W. N. and Beck, S. L. (2006) Cancer-related fatigue,
depressive symptoms, and functional status: A mediation model. Nursing
Research 55, 366-372.

Bauman, K. E., Fisher, L. A., Bryan, E. S. and Chenoweth, R. L. (1984)
Antecedents, subjective expected utility and behavior: A study of adolescent
cigarette smoking. Addictive Behaviors 48, 121-136.

Beauvais, F. and Oetting, E. R. (1986) Drug use in an alternative high school.
Journal of Drug Education 16, 43-50.

Beauvais, F. and Oetting, E. R. (2002) Variances in the etiology of drug use among
ethnic groups of adolescents. Public Health Reports 117, S8-S14.

97
Beck, S. L., Dudley, W. N. and Barsevick, A. M. (2005) Pain, sleep disturbance, and
fatigue in patients with cancer: Using a mediation model to test a symptom
cluster. Oncology Nursing Forum 32, 48-55.

Bernard, H. R., Johnsen, E. C., Kilworth, P. D., McCarty, C., Shelly, G. A. and
Robinson, D. (1990) Comparing four different methods for measuring
personal social networks. Social Networks 12, 179-215.

Berry, J. (1998) Acculturative stress. In Readings in ethnic psychology, Organista, P.
B., Chun, K. M. and Marin, G. eds, pp. 113-117, New York: Routledge.

Berry, J. (2005) Acculturation: Living successfully in two cultures. International
Journal of Intercultural relations 29, 697-712.

Black, S. A. and Markides, K. S. (1993) Acculturation and alcohol consumption in
Puerto Rican, Cuban-American, and Mexican-American women in the
United States. Am J Public Health 83, 890-3.

Blake, S. M., Ledsky, R., Goodenow, C. and O'Donnell, L. (2001) Recency of
immigration, substance use, and sexual behavior among Massachusetts
adolescents. American Journal of Public Health 91, 794-8.

Botvin, G. J., Schinke, S. P., Epstein, J. A. and Diaz, T. (1994) Effectiveness of
culturally focused and generic skills training approaches to alcohol and drug
abuse among minority youths. Psychology of Addictive Behaviors 8, 116-127.

Botvin, G. J., Schinke, S. P., Epstein, J. A., Diaz, T. and Botvin, E. M. (1995)
Effectiveness of culturally focused and generic skills training approaches to
alcohol and drug abuse prevention among minority adolescents: Two-year
follow-up results. Psychology of Addictive Behaviors 9, 183-194.

Brener, N. D. and Wilson, T. W. (2001) Substance use on school property among
students attending alternative high schools in the United States. J Drug Educ
31, 329-342.

Brooks, A. J., Stuewig, J. and LecRoy, C. W. (1998) A family based model of
Hispanic adolescent substance use. Journal of Drug Education 28, 65-86.

Bullers, S., Cooper, M. L. and Russell, M. (2001) Social network drinking and adult
alcohol involvement: A longitudinal exploratioin of the direction of
influence. Addictive Behaviors 26, 181-199.

98
Burnam, M., Hough, R., Karno, M., Escobar, J. I. M. D. and Telles, C. (1987)
Acculturation and lifetime prevalence of psychiatric disorders among
Mexican Americans in Los Angeles. Journal of Health and Social Behavior
28, 89-102.

Burr, J. A. and Mutchler, J. E. (2003) English language skills, ethnic composition,
and household composition: Older Mexican immigrants. The Journals of
Gerentology 58, S83-S92.

Caetano, R. (1987) Acculturation and drinking patterns among U.S. Hispanics.
British J Addict 82, 789-99.

Caetano, R. (2003) Alcohol-related health disparities and treatment-related
epidemiological findings among Whites, Blacks, and Hispanics in the United
States. Alcoholism: Clinical & Experimental Research 27.
Caetano, R., Clark, C. and Tam, T. (1998) Alcohol consumption among racial/ethnic
minorities. Alcohol Health Res World 22, 233-41.

Caetano, R. and Mora, M. E. (1988) Acculturation and drinking among people of
Mexican descent in Mexico and the United States. Journal of Studies on
Alcohol 49, 462-71.

California Department of Education (2006) California Department of Education:
Data & Statistics. Accessed on May 19, 2007 from http://www.cde.ca.gov/ds

Canino, G., De la Rosa, M., Turner, J., Valdez, A., Vega, W. A. and Warner, L. A.
(2003) Epidemiology of drug abuse among Hispanics in the United States. In
National strategic plan on Hispanic drug abuse research, Amaro, H. and
Cortes, D. E. eds, pp. 25-42. National Hispanic Science Network on Drug
Abuse and Northeastern University, Boston.

Centers for Disease Control and Prevention (2004) Surveillance Summaries: Youth
Risk Behavior Surveillance-United States, 2003. MMWR; 53 (No. SS-2).

Cherpitel, C. J. (1992) Acculturation, alcohol consumption, and casualties among
United States Hispanics in the emergency room. International Journal of the
Addictions 27, 1067-77.

Cherpitel, C. J. and Borges, G. (2001) A Comparison of Substance Use and Injury
Among Mexican American Emergency Room Patients in the United States
and Mexicans in Mexico. Alcohol Clin Exp Res 25, 1174-1180.

99
Cherpitel, C. J. and Borges, G. (2002) Substance use among emergency room
patients: an exploratory analysis by ethnicity and acculturation. Am J Drug
Alcohol Abuse 28, 287-305.

Collins, L. M., Sussman, S., Rauch, J. M., Dent, C. W., Johnson, C. A., Hansen, W.
B. and Flay, B. R. (1987) Psychosocial predictors of young adolescent
cigarettes smoking: A sixteen-month, three-wave longitudinal study. Journal
of Applied Social Psychology 17, 554-573.

Cuellar, I., Harris, L. and Jasso, R. (1980) An acculturation scale for Mexican
American normal and clinical populations. Hispanic Journal of Behavioral
Sciences 2, 199-217.

De la Rosa, M. (2002) Acculturation and Latino adolescents' substance use: a
research agenda for the future. Subst Use Misuse 37, 429-56.

De la Rosa, M., Khalsa, J. H. and Rouse, B. A. (1990) Hispanic and illicit drug use:
A review of recent findings. International Journal of Addiction 25, 665-691.

Delgado-Gaitan, C. (1994) Socializing young children in Mexican-American
families: An intergenerational perspective. In Cross-cultural roots of minority
child development, Greenfield, P. and Cocking, R. eds, pp. 55-86. Erlbaum,
Mahway, NJ.

Deosaransingh, K., Moreno, C., Woodruff, S. I. and Sallis, J. F. (1995) Acculturation
and smoking in Latino youth. Health Values 19, 43-52.

Dornelas, E., Patten, C., Fischer, E., Decker, P. A., Offord, K., Barbagallo, J.,
Pingree, S., Croghan, I. and Ahluwalia, J. S. (2005) Ethnic variation in
socioenvironmental factors that influence adolescent smoking. Journal of
Adolescent Health 36, 170-177.

Dudley, W. N., Benuzillo, J. G. and Carrico, M. S. (2004) SPSS and SAS
programming for the testing of mediation models. Nursing Research 53,
59-62.

Ebin, V. J., Sneed, C. D., Morisky, D. E., Rotheram-Borus, M. J., Magnusson, A. M.
and Malotte, C. K. (2001) Acculturation and interrelationships between
problem and health-promoting behaviors among Latino adolescents. Journal
of Adolescent Health 28, 62-72.

100
Ennett, S. T., Bailey, S. L. and Federman, E. B. (1999) Social network characteristics
associated with risky behaviors among runaway and homeless youth. Journal
of Health and Social Behavior 40, 63-78.

Ennett, S. T. and Bauman, K. E. (1993) Peer group structure and adolescent cigarette
smoking: A social network analysis. Journal of Health and Social Behavior,
3.

Ennett, S. T. and Bauman, K. E. (1994) The contribution of influence and selection
to adolescent peer group homogeneity: The case of adolescent cigarette
smoking. Journal of Personality and Social Psychology 67, 653-663.

Ennett, S. T. and Bauman, K. E. (2000) Adolescent social networks: Friendship
cliques, social isolates, and drug use risk. In Improving prevention
effectiveness, Hansen, W. B., Giles, S. M. and Fearnow-Kenney, M. eds.
Tanglewood Research Inc., Greensboro, NC.

Ennett, S. T., Bauman, K. E. and Koch, G. G. (1994) Variability in cigarette smoking
within and between adolescent friendship cliques. Addictive Behaviors 19,
295-305.

Epstein, J. A., Botvin, G. J., Baker, E. and Diaz, T. (1999a) Impact of social
influences and problem behavior on alcohol use among inner-city Hispanic
and Black adolescents. Journal of Studies on Alcohol 60, 595-604.

Epstein, J. A., Botvin, G. J. and Diaz, T. (1998) Linguistic acculturation and gender
effects on smoking among Hispanic youth. Preventive Medicine 27, 583-89.

Epstein, J. A., Botvin, G. J. and Diaz, T. (1999b) Etiology of alcohol use among
Hispanic adolescents. Archives of Pediatric and Adolescent Medicine 153,
1077-1084.

Epstein, J. A., Botvin, G. J. and Diaz, T. (2001) Linguistic acculturation associated
with higher marijuana and polydrug use among Hispanic adolescents.
Substance Use & Misuse 2001, 4.

Epstein, J. A., Botvin, G. J., Dusenbery, L., Diaz, T. and Kerner, J. (1996) Validation
of an acculturation measure for Hispanic adolescents. Psychological Reports
79, 1075-1079.

Epstein, J. A., Doyle, M. and Botvin, G. J. (2003) A mediational model of the
relationship between linguistic acculturation and polydrug use among
Hispanic adolescents. Psychological Reports 93, 859-66.
101
Escobar, J. (1998) Immigration and mental health: Why are immigrants better off?
Archives of General Psychiatry 55, 781-782.

Farabee, D., Wallisch, L. and Maxwell, J. C. (1995) Substance use among Texas
Hispanics and Non-Hispanics: who's using, who's not and why. Hispanic
Journal of Behavioral Sciences 17, 523-536.

Felix-Ortiz, M. and Newcomb, M. D. (1995) Cultural identity and drug use among
Latino and Latina adolescents. In Drug use prevention with multiethnic
youth, Botvin, G. J., Schinke, S. and Orlandi, M. A. eds. Sage Publications,
Thousand Oaks, CA.

Fosados, R., Chou, C., Sussman, S., Baezconde-Garbanati, L., Pachon, H. and
Valente, T. W. (under review-a) Exploring the association between linguistic
acculturation and substance use among Hispanic continuation high school
youth: Evidence for social influence as a mediator. J Health Soc Behav.
Fosados, R., Sussman, S., Baezconde-Garbanati, L., Chou, C.-P., Pachon, H. and
Valente, T. W. (under review-b) Linguistic acculturation and substance use
among high and low-risk youth: Peer social influence as a potential mediator.
Addiction.
Frank, R. E. (1966) Use of transformations. Journal of Marketing Research 3,
247-253.

Gaughan, M. (2003) Predisposition and pressure: Mutual influence and adolescent
drunkenness. Connections 25, 17-31.

Gogineni, A., Stein, M. D. and Friedman, P. D. (2001) Social relationships and
intravenous drug use among methadone maintenance patients. Drug &
Alcohol Dependence 64, 47-53.

Graham, J. W., Flay, B. R., Johnson, C. A., Hansen, W. B., Grossman, L. M. and
Sobel, J. L. (1984) Reliability of self-reported measures of drug use in
prevention research: evaluation of the Project SMART questionnaire via test-
retest reliability matrix. Journal of Drug Education 14.

Granovetter, M. (1973) The strength of weak ties. American Journal of Sociology 78,
1360-1380.

Granovetter, M. S. (1982) The strength of weak ties: A network theory revisited. In
Social structure and network analysis, Marsden, P. V. and Lin, N. eds, pp.
105-130. Sage, Beverly Hills, CA.
102
Gritz, E. R., Prokhorov, A. V., Hudmon, K. S., Jones, M. M., Rosenblum, C., Chang,
C. C., Chamberlain, R. M., Taylor, W. C., Johnston, D. and De Moor, C.
(2003) Predictors of susceptibility to smoking and ever smoking: A
longitudinal study in a triethnic sample of adolescents. Nicotine & Tobacco
Research 5, 493-506.

Grunbaum, J. A. and Basen-Engquist, K. (1993) Comparison of health risk behaviors
between students in a regular high school and students in an alternative high
school. Journal of School Health 63, 421-425.

Grunbaum, J. A., Kann, L., Kinchen, S., Ross, J., Hawkins, J., Lowry, R., Harris, W.
A., McManus, T., Chyen, D. and Collins, J. (2004) Youth Risk Behavior
Surveillance - United States, 2003. MMWR 53, 1-100.

Grunbaum, J. A., Kann, L., Kincher, S. A., Ross, J. G., Gowda, V. R., Collins, J. L.
and Kolbe, L. J. (1999) Youth risk behavior surveillance -- National
alternative high school youth risk behavior survey, United States, 1998.
Morbidity Mortality Weekly Report 48, 1-44.

Grunbaum, J. A., Lowry, R. and Kann, L. (2001) Prevalence of health-related
behaviors among alternative high school students as compared with students
attending regular high schools. Journal of Adolescent Health 29, 337-343.

Grunbaum, J. A., Tortolero, S. R., Weller, N. F. and Gingiss, P. (2000) Cultural,
social, and intrapersonal factors associated with substance use among
alternative high school students. Addictive Behaviors 25, 145-151.

Hall, J. A. and Valente, T. W. (2007) Adolescent smoking networks: The effects of
influence and selection on future smoking. Addictive Behaviors,
doi:10.1016/j.addbeh.2007.04.008.

Harrell, A. V. (1997) The validity of self-report drug use data: The accuracy of
responses on confidential self-administered answered sheets. In The validity
of self-reported drug use: Improving the accuracy of survey estimates,
Harrison, L. and Hughes, A. eds, NIDA Research Monograph 167. U.S.
Department of Health and Human Services, National Institutes of Health,
Rockville, MD.

Harris, J. R. (1995) Where is the child's environment? A group socialization theory
of development. Psychological Review 102, 458-489.

103
Havassy, B. E., Wasserman, D. A. and Hall, S. M. (1995) Social relationships and
abstinence from cocaine in an American treatment sample. Addiction 90,
699-710.

Haynie, D. L. (2001) Delinquent peers revisited: Does network structure matter? AJS
106, 1013-57.

Hoffman, B. R., Monge, P. R., Chou, C-P. and Valente, T. W. (2007) Perceived peer
influence and peer selection on adolescent smoking. Addictive Behaviors 32,
1546-1554.

Hoffman, B. R., Sussman, S., Unger, J. and Valente, T. W. (2006) Peer influences on
adolescent cigarette smoking: A theoretical review of the literature.
Substance Use & Misuse 41, 103-155.

Hsu, T. K., Grant, A. E. and Huang, W. (1993) The influence of social networks on
the acculturation behavior of foreign students. Connections 16, 23-36.

INEGI (2006) Porcentaje de población hablante de lengua indígena de 5 y más años
por entidad federativa, 2000 y 2005 Instituto Nacional de Estadistica
Geografia e Informatica, Mexico, D.F. Accessed on January 21, 2006 from
http://www.inegi.gob.mx/est/contenidos/espanol/rutinas/ept.asp?t=mlen02&c
=3327.

Jessor, R. and Jessor, J. L. (1977) Problem behavior and psychological development:
A longitudinal study of youth. Academic Press, San Diego, CA.

Johnston, L. D., O'Malley, P. M. and Bachman, J. G. (2001) The Monitoring the
Future National survey results on adolescent drug use: Overview of key
findings, 2000 National Institute on Drug Abuse, (NIH Publication No. 01-
4923). Bethesda, MD.

Johnston, L. D., O'Malley, P. M., Bachman, J. G. and Schulenberg, J. (2005)
Monitoring the future national results on adolescent drug use: Overview of
key findings, 2005. Bethesda, MD: National Institute on Drug Abuse.
Accessed June 4, 2006 from http://monitoringthefuture.org/pubs/
monographs/overview2005.pdf.

Johnston, L. D., O'Malley, P. M., Bachman, J. G. and Schulenberg, J. (2006)
Monitoring the Future national survey results on drug use, 1975-2005.
Volume I: Secondary school students. Bethesda, MD: National Institute on
Drug Abuse. Accessed September 17, 2006 from
http://monitoringthefuture.org/pubs/monographs/vol1_2005.pdf.
104
Kandel, D. B. (1978) Homophily, selection, and socialization in adolescent
friendship. American Journal of Sociology 84, 427-436.

Kandel, D. B. (1985) On processes of peer influences in adolescent drug use: A
developmental perspective. Advances in Alcohol and Substance Abuse 4,
139-163.

Kandel, D. B. and Davis, M. (1991) Friendship networks, intimacy and illicit drug
use in young adulthood: A comparison of two competing theories.
Criminology 29, 441-470.

Keefe, K. and Newcomb, M. D. (1996) Demographic and psychosocial risk for
alcohol use: Ethnic differences. Journal of Studies on Alcohol 57, 521-530.

Kim, Y. Y. (2005) Inquiry in intercultural and development communication. Journal
of Communication 55, 554-577.

Kirke, D. M. (2004) Chain reactions in adolescents' cigarette, alcohol, and drug use:
Similarity through peer influence or the patterning of ties in peer networks?
Social Networks 26, 3-28.

Kirke, D. M. (2006) Teenagers and substance use: Social networks and peer
influence. Palgrave Macmillan, New York.

Kobus, K. (2003) Peers and adolescent smoking. Addiction 98, 37-55.

Krackhardt, D. (1992) The strength of strong ties: The importance of philos in
organizations. In Networks and organizations: Structure, form, and action,
Nohria, N. and Eccles, R. eds, pp. 216-239. Harvard Business School Press,
Boston, MA.
Krohn, M., Lizotte, A., Thornberry, T., Smith, C. and McDowall, D. (1996)
Reciprocal causal relationship among drug use, peers, and beliefs. Journal of
Drug Issues 26, 405-428.

Krohn, M. and Thornberry, T. P. (1993) Network theory: A model for understanding
drug abuse among African-American and Hispanic youth. NIDA Research
Monograph 130, 102-128.

Landrine, H. and Klonoff, E. A. (2004) Culture change and ethnic-minority health
behavior: an operant theory of acculturation. Journal of Behavioral Medicine
27, 527-55.

105
Landrine, H., Richardson, J. L., Klonoff, E. A. and Flay, B. (1994) Cultural diversity
in the predictors of adolescent cigarette smoking: The relative influence of
peers. Journal of Behavioral Medicine 17, 331-346.

Latkin, C. A., Hua, W. and Tobin, K. E. (2004) Social network correlates of self-
reported non-fatal overdose. Drug & Alcohol Dependence 73, 61-67.

Latkin, C. A., Knowlton, A. R., Hoover, D. R. and Mandell, W. (1999) Drug
network characteristics as a predictor of cessation of drug use among adult
injection drug users: A prospective study. American Journal of Drug &
Alcohol Abuse 25, 463-473.

Latkin, C. A., Mandell, W., Vlahov, D., Oziemkowska, M. and Celentano, D. A.
(1996) People and places: Behavioral settings and personal network
characteristics as correlates of needle sharing. Journal of Acquired Immune
Deficiency Syndrome 13, 273-280.

Lessenger, L. H. (1997) Use of acculturation rating scale for Mexican Americans II
with substance abuse patients. Hispanic Journal of Behavioral Sciences 19,
187-198.

MacKinnon, D. P., Fairchild, A. J. and Fritz, M. S. (2007) Mediation Analysis. Annu
Rev Psychology 58, 593-614.

Marin, G. and Marin, B. (1991) Research with Hispanic populations. Sage, Newbury
Park, CA.

Marin, G., Sabogal, F., Marin, B. V., Otero-Sabogal, R. and Perez-Stable, E. J.
(1987) Development of a short acculturation scale for Hispanics. Hisp J
Behav Sci 9, 183-205.

Marsiglia, F. F., Kulis, S., Hecht, M. L. and Sills, S. (2004) Ethnicity and ethnic
identity as predictors of drug norms and drug use among preadolescents in
the US Southwest. Subst Use Misuse 39, 1061-94.

Marsiglia, F. F. and Waller, M. (2002) Language preference and drug use among
Southwestern Mexican American middle school students. Children and
Schools 24, 145-158.

McCuller, W. J., Sussman, S., Dent, C. W. and Teran, L. (2001) Concurrent
prediction of drug use among high-risk youth. Addictive Behaviors 26,
137-42.

106
McQueen, A., Getz, J. G. and Bray, J. H. (2003) Acculturation, substance use, and
deviant behavior: examining separation and family conflict as mediators.
Child Development 74, 1737-50.

National Institute of Drug Abuse and University of Michigan (2004) Monitoring the
Future National Survey Results on Drug Use, 1975-2003, Volume I:
Secondary School Students.

National Institute on Drug Abuse (2004) NIDA's strategic plan on reducing health
disparities. Accessed on December 27, 2006 from http://www.drugabuse.gov/
pdf/healthdispplan.pdf.

Netter, J., Wasserman, W. and Kuthner, M. H. (1990) Applied linear statistical
models. Richard D. Irwin Inc., Illinois.

Nielsen, A. L. and Ford, J. A. (2001) Drinking patterns among Hispanic adolescents:
results from a national household survey. J Stud Alcohol 62, 448-56.

Norris, A., Ford, K. and Bova, C. (1996) Psychometrics of a brief acculturation scale
for Hispanics in a probability sample of urban Hispanic adolescents and
young adults. Hispanic Journal of Behavioral Sciences 18, 29-38.

O'Malley, P. M., Johnston, L. D., Bachman, J. G., Schulenber, J. E. and Kumar, R.
(2006) How substance use differs among American secondary schools.
Prevention Science 10.

Oetting, E. and Donnermeyer, J. (1998) Primary socialization theory: The etiology of
drug use and deviance. Substance Use & Misuse 33, 995-1026.

Perry, C. and Jessor, R. (1985) The concept of health promotion and the prevention
of adolescent drug abuse. Health Education Quarterly 12, 169-184.

Portes, A. (1990) Immigrant America: A portrait. University of California Press,
Berkeley, CA.

Portes, A. and Rumbaut, R. (2001) Legacies: the story of the immigrant second
generation. University of California Press, Berkeley, CA.

Rai, A. A., Stanton, B., Wu, Y., Li, X., Galbraith, J., Cottrell, L., Pack, R., Harris, C.,
D'Alessandri, D. and Burns, J. (2003) Relative influences of perceived
parental monitoring and perceived peer involvement on adolescent risk
behaviors: An analysis of six cross-sectional data sets. Journal of Adolescent
Health 33, 108-118.
107
Rice, R. E., Donohey, L. and Clayton, R. (2003) Peer network, sensation seeking,
and drug use among junior and senior high school students. Connections 25,
32-58.

Sabogal, F., Marin, G., Otero-Sabogal, R., Marin, B. and Perez-Stable, E. J. (1987)
Hispanic familism and acculturation: What changes and what doesn't?
Hispanic Journal of Behavioral Sciences 9, 397-412.

Samaniego, R. Y. and Gonzales, N. A. (1999) Multiple mediators of the effects of
acculturation status on delinquency for Mexican American adolescents.
American Journal of Community Psychology 27, 189-210.

SAS Institute (1990) SAS user's guide: Basics. Version 9 edition. author, Cary, NC.

Schroeder, J. R., Latkin, C. A., Hoover, D. R., Curry, A. D., Knowlton, A. R. and
Celentano, D. D. (2001) Illicit drug use in one's social network and in one's
neighborhood predicts individual heroin and cocaine use. AEP 11, 389-394.

Schweizer, T., Schnegg, M. and Berzborn (1998) Personal networks and social
support in a multiethnic community of southern California. Social Networks
20, 1-21.

Shin, H. B. and Bruno, R. (2003) Language use and English-speaking ability: 2000,
Washington, D.C.: U.S. Department of Commerce, Economics and Statistics
Administration, U.S. Census Bureau. Accessed on January 17, 2007 from
http://www.census.gov/prod/2003pubs/c2kbr-29.pdf.

Sieving, R. E., Perry, C. L. and Williams, C. L. (2000) Do friendships change
behaviors, or do behaviors change friendships? Examining paths of influence
in young adolescents' alcohol use. Journal of Adolescent Health 26, 27-35.

Simon, T. R., Stacy, A. W., Sussman, S. and Dent, C. W. (1994) Sensation seeking
and drug use among high risk Latino and Anglo adolescents. Person Indiv
Diff 17, 665-672.

Singer, J. D. (1998) Using SAS PROC MIXED to fit multilevel models, hierarchical
models, and individual growth models. Journal of Educational and
Behavioral Statistics 24, 323-355.

Skara, S., Rohrbach, L. A., Sun, P. and Sussman, S. (2005) An evaluation of the
fidelity of implementation of a school-based drug abuse prevention program:
Project Towards No Drug Abuse (TND). Journal of Drug Education 35,
305-329.
108
Sobel, M. E. (1982) Asymptotoc confidence intervals for indirect effects in structural
equation models. In Sociological methodology, Leinhart, S. ed, pp. 290-312.
Jossey-Bass, San Francisco.

Suh, T., Mandell, W., Latkin, C. A. and Kim, J. (1997) Social network
characteristics and injection HIV-risk behaviors among street injection drug
users. Drug and Alcohol Dependence 47, 137-143.

Sussman, S., Dent, C. W., Burton, D., Stacy, A. W. and Flay, B. R. (1995a)
Developing school-based tobacco use prevention and cessation programs.
Sage, Thousand Oaks (CA).

Sussman, S., Dent, C. W. and Leu, L. (2000) The one-year prospective prediction of
substance abuse and dependence among high-risk adolescents. Journal of
Substance Abuse 12, 373-386.

Sussman, S., Dent, C. W., Simon, T. R., Stacy, A. W., Galaif, E. R., Moss, M. A.,
Craig, S. and Johnson, C. A. (1995b) Immediate impact of social influence-
oriented substance abuse prevention curricula in traditional and continuation
high schools. Drugs & Society 8, 65-81.

Sussman, S., Dent, C. W., Stacy, A. W., Burciaga, C., Raynor, A., Turner, G. E.,
Charlin, V., Craig, S., Hansen, W. B., Burton, D. and Flay, B. R. (1990) Peer-
group association and adolescent tobacco use. Journal of Abnormal
Psychology 99, 349-352.

Sussman, S., Dent, C. W., Stacy, A. W. and Craig, S. (1998) One-year outcomes of
Project Towards No Drug Abuse. Prev Med 27, 632-642.

Sussman, S., Stacy, A. W., Dent, C. W., Simon, T. R., Galaif, E. R., Moss, M. A.,
Craig, S. and Johnson, C. A. (1995c) Continuation high schools: Youth at
risk for drug abuse. Journal of Drug Education 25, 191.
Sussman, S., Sun, P., McCuller, W. J. and Dent, C. W. (2003) Project Towards No
Drug Abuse: Two-year outcomes of a trial that compares health educator
delivery to self-instruction. Preventive Medicine 37, 155-162.

Triandis, H. C. (1995) Individualism and collectivism. Westview Press, Boulder, CO.

U.S. Census Bureau (2001) The Hispanic population: Census 2000 brief,
Washington, DC: U.S. Department of Commerce, Economics and Statistics
Administration. Accessed January 8, 2007 from http://www.census.gov/
prod/2001pubs/c2kbr01-3.pdf.
109
Unger, J., Baezconde-Garbanati, L., Shakib, S., Palmer, P., Nezami, E. and Mora, J.
(2004) A cultural psychology approach to "drug abuse" prevention.
Substance Use & Misuse 39, 1779-1820.

Unger, J., Cruz, T. B., Rohrbach, L. A., Ribisl, K. M., Baezconde-Garbanati, L.,
Chen, X. and et al. (2000) English language use as a risk factor for smoking
initiation among Hispanic and Asian American adolescents: Evidence for
mediation by tobacco-related beliefs and social norms. Health Psychology 19,
403-410.

Unger, J., Gallaher, P., Shakib, S., Ritt-Olson, A., Palmer, P. and Johnson, C. A.
(2002) The AHIMSA acculturation scale: A new measure of acculturation for
adolescents in a multicultural society. Journal of Early Adolescence 22,
225-51.

Unger, J., Rohrback, L. A., Cruz, T. B., Baezconde-Garbanati, L., Howard, K. A.,
Palmer, P. H. and Johnson, C. A. (2001) Ethnic variation in peer influences
on adolescent smoking. Nicotine & Tobacco Research 3, 167-176.

Urberg, K. A., Cheng, C. H. and Shyu, S. J. (1991) Grade changes in peer influence
on adolescent cigarette smoking: A comparison of two measures. Addictive
Behaviors 16.
Urberg, K. A., Degirmencioglu, S. M. and Pilgrim, C. (1997) Close friend and group
influence on adolescent cigarette smoking and alcohol use. Developmental
Psychology 33, 834-844.

Urberg, K. A., Luo, G., Pilgrim, C. and Degirmencioglu, S. M. (2003) A two-stage
model of peer influence in adolescent substance use: Individual and
relationship-specific differences in susceptibility to influence. Addictive
Behaviors 28, 1234-1256.

Valente, T. W. (2003) Social network influences on adolescent substance use: An
introduction. Connections 25, 11-16.

Valente, T. W., Gallaher, P. and Mouttapa, M. (2004) Using social networks to
understand and prevent substance use: A transdisciplinary perspective.
Substance Use & Misuse 39, 1685-1712.

Valente, T. W., Okamoto, J., Pumpuang, P. and Sussman, S. (in press) Differences in
perceived implementation of a standard versus a peer-led interactive
substance-abuse prevention program. American Journal of Health Behavior.
110
Valente, T. W., Sussman, S., Unger, J., Ritt-Olson, A., Okamoto, J. and Stacy, A. W.
(submitted) Peer influence and deviancy training: Effects of a network
tailored substance abuse prevention program among high risk adolescents.
Addiction.
Valente, T. W., Unger, J. B. and Johnson, C. A. (2005) Do popular students smoke?
The association between popularity and smoking among middle school
students. Journal of Adolescent Health 37, 323-329.

Valente, T. W. and Vlahov, D. (2001) Selective risk taking among needle exchange
participants: Implicaitons for supplemental interventions. Am J Public Health
91, 406-411.

Vega, W., Alderete, E., Kolody, B. and Aguilar-Gaxiola, S. (1998a) Illicit drug use
among Mexicans and Mexican Americans in California: the effects of gender
and acculturation. Addiction 93, 1839-1850.

Vega, W., Gil, A. and Wagner, E. (1998b) Cultural adjustments and Hispanic
adolescent drug use. In Drug use and ethnicity in early adolescents, Vega, W.
A. and Gil, A. eds, pp. 125-148. Plenum Press, New York.

Vega, W., Zimmerman, R., Warheit, G. and Gil, A. (2003) Acculturation, stress and
Latino adolescent drug use. In Socioeconomic conditions, stress and mental
disorders: Toward a new synthesis of research and public policy, National
Institute of Mental Health ed, Bethesda, MD.

Vega, W. A., Gil, A. G. and Zimmerman, R. S. (1993) Patterns of drug use among
Cuban-American, African-American, and white non-Hispanic boys. Am J
Public Health 83, 257-9.

Wall, J. A., Power, T. G. and Arbona, C. (1993) Susceptibility to antisocial peer
pressure and its relation to acculturation in Mexican-American adolescents.
Journal of Adolescent Research 8, 403-148.

Wallace, J. M., Bachman, J. G., O'Malley, P. M., Schulenberg, J., Cooper, S. M. and
Johnston, L. D. (2003) Gender and ethnic differences in smoking, drinking,
and illicit drug use a mong American 8th, 10th, and 12th grade students,
1976-2000. Addiction 98, 225-234.

Warner, L. A., Valdez, A., Vega, W. A., De La Rosa, M., turner, R. J. and Canino,
G. (2006) Hispanic drug abuse in an evolving cultural context: An agenda for
research. Drug and Alcohol Dependence 84S, S8-S16.

111
Wasserman, D. A., Weinstein, M. G., Havassy, B. E. and Hall, S. M. (1998) Factors
associated with lapses to heroin use during methadone maintenance. Drug &
Alcohol Dependence 58, 183-192. In: M.D. Stein, A. Charuvastra, B.J.
Anderson (2002). Social support and zero sharing risk among hazardously
drinking injection drug users. Journal of Substance Abuse Treatment. 23,
225-230.

Wasserman, S. and Faust, K. (1994) Social network analysis: Methods and
applications. Cambridge University Press, New York, NY.

Wills, T., Resko, J. A., Ainette, M. G. and Mendoza, D. (2004) Role of parent
support and peer support in adolescent substance use: A test of mediated
effects. Psychology of Addictive Behavaviors 19, 122-134.

Windle, M. (2000) Parental, sibling, and peer influences on adolescent substance use
and alcohol problems. Applied Developmental Science 4, 98-110.

Abstract (if available)

Abstract Students attending alternative/continuation high schools report significantly higher substance use rates than regular high school students. Peer influence is often implicated as a significant correlate of substance use, with peer relationships being one of the primary factors involved in whether or not youth decide to engage in and maintain these behaviors. This dissertation argues for the need to examine peer social influence as a potential mediator for the association between acculturation on substance use, and the need to apply social network methods when investigating peer influence and substance use.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Contextualizing social network influences on substance use among high risk adolescents

PDF

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

PDF

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

PDF

Social self-control and adolescent substance use

PDF

Factors and correlates of sexual behaviors among young adults from continuation high schools

PDF

Sociocultural stress, coping and substance use among Hispanic/Latino adolescents

PDF

Investigating factors that influence peer relationships and obesity during middle childhood

PDF

Energy drink consumption, substance use and attention-deficit/hyperactivity disorder among adolescents

PDF

Role transitions, past life events, and their associations with multiple categories of substance use among emerging adults

PDF

The role of acculturation in the sun-safe behaviors of US Latino adults in the United States

PDF

Social network influences on depressive symptoms among Chinese adolescents

PDF

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

PDF

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

PDF

The dynamic relationship of emerging adulthood and substance use

PDF

Individual, family, and peer influences on substance use, delinquency, and problem behaviors among at-risk Asian American and Pacific Islander adolescents

PDF

Examining the built and online environments to understand cannabis and e-cigarette availability, marketing, and product use in Southern California

PDF

Adolescent social networks, smoking, and loneliness

PDF

The influence of contextual factors on the processes of adoption and implementation of evidence-based substance use prevention and tobacco cessation programs in schools

PDF

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

PDF

Normative and network influences on electronic cigarette use among adolescents

Asset Metadata

Creator Fosados, Raquel (author)

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

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Preventive Medicine (Health Behavior)

Publication Date 06/21/2007

Defense Date 06/06/2007

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

Tag acculturation,alternative/continuation high school,Hispanic,OAI-PMH Harvest,peer influence,social network,substance use

Language English

Advisor Valente, Thomas W. (committee chair), Baezconde-Garbanati, Lourdes (committee member), Chou, Chih-Ping (committee member), Pachon, Harry (committee member), Sussman, Steven (committee member)

Creator Email fosados@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m545

Unique identifier UC1474213

Identifier etd-Fosados-20070621 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-506746 (legacy record id),usctheses-m545 (legacy record id)

Legacy Identifier etd-Fosados-20070621.pdf

Dmrecord 506746

Document Type Dissertation

Rights Fosados, Raquel

Type texts

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

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu