University of Southern California Dissertations and Theses

Contextualizing social network influences on substance use among high risk adolescents

(USC Thesis Other)

Contextualizing social network influences on substance use among high risk adolescents

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

CONTEXTUALIZING SOCIAL NETWORK INFLUENCES ON
SUBSTANCE USE AMONG HIGH RISK ADOLESCENTS

by
Patchareeya Pumpuang Kwan

___________________________________________________________________

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

May 2010

Copyright 2010 Patchareeya Pumpuang Kwan

ii

DEDICATION
What intrigues me most about the results and implications of my dissertation
work is that social influence and social networks are protective. Oftentimes, we study risk
factors of delinquent behaviors such as substance use and aggression among adolescents.
However, what I find important is that we take an optimistic approach and study the
factors that can be protective against these delinquencies. Among adolescents, one of the
most important protective factors against delinquent behavior is strong family ties and
support. This matter is dear to my heart because I am an example of how strong family
ties and support are protective and beneficial.
I’ve always thought of myself as a very fortunate person. Coming from a working
class immigrant family, I wasn’t raised a “princess” or anything of that sort. However, I
am lucky enough to have very good people in my network – dad, mom, sister, extended
family members, friends, teachers, co-workers and now my own little family. All of these
people whose name I don’t have room to mention in this text have influenced me to do
good things with my life and have provided all types of social support. As their child, my
parents and sister have given me so much love and support that I know I will never get
hurt when I fall. As his wife, my husband has given me more of everything, more than I
can ever imagined and more than anyone is willing to give me. As their mother, my
children have taught me about unconditional love. They’ve all given me the opportunity
to do well and accomplish my goals because I am energized by love and thus I dedicate
my work to them.

iii

ACKNOWLEDGEMENTS
I want to acknowledge so many people who have made this dissertation possible.
Of course, the most important group is my committee members: Drs. Thomas Valente,
Steve Sussman, Ping Sun, Nathaniel Riggs, Ricardo Stanton-Salazar and Eric Rice. At
first it was quit scary to not have the support of a fellow female scholar/mother who I felt
understood about my obligations beyond this academic endeavor. However, after sitting
down with everyone, I was not afraid any longer. Whether directly or indirectly, every
member conveyed to me that they were there to only help me and hence, I am so
fortunate and thankful for all of them. Every comment, suggestions and advice will also
be remembered.
I’d also like to thank my close friends and family who are always there when I
need them. Thank you to my husband for spending countless hours looking up and
printing out research articles and cleaning and entering data. Thank you to my in-laws for
watching my kids when I was working and thank you even more for doing such a great
job raising them. Thank you to my friends for serving as stress relief whenever I needed
to talk to someone about all the crazy things in life.
Last but not least, I want to acknowledge and thank my friends at IPR. We’ve
spent close to a decade together and I have grown and learned from our shared
experiences. Thank you IPR staff members for helping me navigate my way around
throughout the year, thank you Marny for answering all my questions, thank you MPH
staff members for being such wonderful co-workers, and most of all, thank you fellow
doctoral students for all your help and advice. I could never have done it without you.

iv

TABLE OF CONTENTS
Dedication ii
Acknowledgements iii
List of Tables v
List of Figures vi
Abstract vii
Chapter 1. Introduction 1
Chapter 2. Mediation Affects of Intentions on Drug Use 31
Chapter 3. Bilingualism and its Role in Substance Use 49
Chapter 4. Peer Leaders and their Role in Substance Use 73
Chapter 5. Conclusion 92
References 102

v

LIST OF TABLES
Table 1.1 Brief Description of Standard TND Curriculum

22
Table 1.2 Brief Description of Modifications made in TND-
Network Curriculum

23
Table 1.3 Demographic Characteristics of the Study Population

27
Table 1.4 Last 30 Day Drug Use by Time Points

28
Table 2.1 Results of Mediation Analysis

43
Table 3.1 Questions Used to Measure Bilingualism

61
Table 3.2 Number of Monolinguals vs. Bilinguals by Ethnicity

65
Table 3.3 Baseline Drug Use by (Monolinguals vs. Bilinguals)

65
Table 3.4 Cross Tabulation of Best Friends Language Use

66
Table 3.5 Results of Logistic Regression for Both Genders
Combined

67
Table 3.6 Test for Differences Between Male and Female
Language Use

67
Table 3.7 Results of Logistic Regression for Both Genders,
Separately

69
Table 4.1 Demographic Characteristics (All Condition vs. Network
Condition)

85
Table 4.2 Test for Differences Between Peer Leader Substance Use

86
Table 4.3 Results of Logistic Regression for Peer Leader Use

89

vi

LIST OF FIGURES
Figure 1.1 Social Networks and Health

5
Figure 1.2 Study Design and Data collection

25
Figure 2.1 Theoretical/Conceptual Model for Mediation Analysis

38
Figure 3.1 Theoretical/Conceptual Model for Bilingualism

59
Figure 4.1 Theoretical/Conceptual Model for Peer Leader Influence

81

vii

ABSTRACT
Social network influences predict how we act and behave. Most people tend to
comply with social norms and beliefs so that they are accepted by society. Characterized
by physical, emotional and mental changes, adolescents are no different when it comes to
social conformity. Peer influences become a strong predictor of how adolescents act and
behave. Many of these individuals behave in ways that are encouraged and acceptable by
peers despite the delinquent nature of the behavior. The purpose of this dissertation is to
understand adolescent social influence and contextualized social networks as it relates to
drug use.
The purpose of the first analysis, depicted in chapter 2, is to examine intentions as
a mediator of the relationship between friend approval of drug use and drug use. Data on
drug use was examined in roughly 400 alternative high school students. Best friends’
approval of drug use was used as the independent variable with use of alcohol, marijuana,
cigarette, hard drugs and a composite drug use score were used as the dependent
variables. Drug use intentions is the mediator variable of interest. Mixed model logistic
regression was used to conduct mediation analysis as described in Baron and Kenny,
1986. Results of this analysis showed that behavioral intentions fully mediated the
relationship between best friend’s approval and cigarettes and marijuana use. Intentions
partially mediated the relationship between approval and hard drug use. This is consistent
with past research and theoretical models such as the Theory of Reasoned Action.
In chapter 3 we will look at the second set of analysis concerning bilingualism as
a possible protective factor against adolescent drug use. Logistic regression using school

viii

and classroom as the random effect was conducted on approximately 880 male and
female students in alternative high schools. Outcome variables of interest were baseline
use of alcohol, marijuana, cigarette, hard drugs and a composite drug use variable. Age,
gender, ethnicity, SES, acculturation and number of best friends were used as controlled
variables. Individual bilingualism scores and best friends’ bilingualism scores (i.e. the
level of bilingualism of the individual’s five best friends) were the predictor variables of
interest. Results indicated that among males and females, best friends’ bilingualism
decreased the probability of alcohol, marijuana and cigarette use. Sub-group analysis by
sex reveals that among males, bilingualism was not associated with drug use but among
females best friend bilingualism was associated with decreasing the probability of alcohol,
marijuana, cigarettes and all drug use. These findings suggest differences in how male
and female students are influenced by their bilingual peers.
Chapter 4 examines the association between peer leader drug use behaviors and
the behaviors of the individual students. Logistic regression using school, classroom and
network group as the random effects was used to analyze post intervention drug use
behaviors of approximately 525 male and female students. Students were randomized
into three groups – control, standard curriculum or networked curriculum where students
were grouped according to self reported group leader nominations. Peer leader drug
usage behavior was the independent variable of interest. Individual drug use was the
outcome being examined. Results showed that among the combined male and female
group, there was no significant association between peer leader use and post-test drug use.
Among males only, peer leader use at baseline was positively associated with alcohol use

ix

and peer leader use at post-test was positively associated with hard drug use. Both of
these associations were moderated by the network condition (i.e. the interaction term was
significant). Among females only, peer leader use at post-test was negatively associated
with marijuana and cigarette use at post-test. Again, both were moderated by the network
condition. These findings suggest that having peer leaders in the network condition
decreased the odds of marijuana and cigarette use among female students. The opposite
effect was found in males. This lends credence to the idea that males and females interact
differently when placed in peer groups.

1

CHAPTER 1: INTRODUCTION
Developmental theorists such as Jean Piaget and Lev Vygotsky have
acknowledged that social interaction is, to varying extents, an important part of the child
developmental process (Tudge & Rogoff, 1999). These interactions introduce the child to
other individuals whom can set examples and serve as influences. In the transition from
childhood to adulthood, adolescents often look to their peers for support and examples
and, in turn, these peers can be become highly influential (Berndt, 1979).
Causes of adolescent substance abuse have been linked to many factors. Risk
factors for substance abuse can be divided into various categories such as those related to
society and the environment and those related to individual, behavioral and interpersonal
factors such as family, friends, and peer groups (Beman, 1995; Hawkins et al., 1992).
Adolescents have identified peers as among the most influential factors in use of illicit
drugs. Peers influence initiation of marijuana use through encouragement and modeling.
(Kandel, 1985). A one-year prospective study of 173 families revealed that in addition to
parents, peer modeling and peer attitude influenced adolescent alcohol use (Ary et al.,
1993). Fortunately, family and peer groups can also serve to protect and prevent
substance use through positive influence (Liddle et al., 2001).

Specific Aims
Adolescent social networks provide influence and support. These influences can
be negative or positive. Negative influence increases the likelihood of drug use initiation,
experimentation, use and abuse. Positive influences exist to serve as a barrier and

2

potentially discourage use among adolescents. The purpose of this dissertation is to
examine the role of adolescents’ social networks on drug use behaviors. More
specifically, the paper will:
1. Examine the relationship between peer approval of drug use, individual intentions
to use drugs and actual use. Chapter 2 will first examine how the conceptual idea
of best friends’ approval influences one’s intentions to use drugs. Are adolescents
influenced by their perception of friends’ approval? How does this perception
influence their intentions and behavior?
2. Examine the role of bilingualism as a protective factor against adolescent drug use.
Bilingualism has many benefits, one of which is that bilingual individuals have
larger and more diverse social networks which provide more access to resources.
These multiple sources of support serve as protective factors to adolescent drug
use. Do bilingual networks decrease drug use among bilingual individuals? How
does having a bilingual network alter the drug use behaviors of adolescents?
3. Examine the role of peer influence on drug use, in particular, peer leaders and
how they influence others around them to use drugs. Peer leaders are often
imitated by others who strive to be similar and accepted by them. If peer leaders
use drugs, would other students imitate that behavior? Can peer leaders have such
a strong influence on the students they lead?

3

Social Networks
Emotional support, financial assistance, advice giving, advice receiving, influence,
and encouragement…these are some of the products of our social networks. Although
today, we might see social networks in the forms of internet based groups such as
Facebook, MySpace, Friendster or Twitter, in which we find our friends from looking at
our friends’ friends, making friends with our friends’ friends, and looking at who knows
who and how they are related, social networks in the context of our society has been
around since the beginning of time. The spread of faith or religion is a product of our
interaction with others and their influence on what we believe (Stark & Bainbridge, 1980).
The food we eat, the clothes that we wear and the language that we speak are the results
of our family, friends and society’s influence on our behavior. In return, what we do and
how we behave influences our family, friends and society.
The term “social networks” has been used loosely in the earlier parts of the 20
th

century. However, in 1954 while studying a Norwegian village, J.A. Barnes used the term
social networks to describe the relationships between individuals (Barnes, 1954). Social
networks then became a term that refers to the web of individuals and their connections
to each other. It is concerned with the concept of looking at the links or ties between
individuals and how these links are described. In social network terminology, individuals
are called “nodes” and their connections are called “ties”. Egocentric network refers to
the web of individuals whom the person (i.e. ego) has connections. If the people within
the networks are close to each other and know each other very well, then it is said to be a
dense network. If the relationship between two nodes are very strong (e.g. husband and

4

wife, mother and daughter or best friends) then the ties are said to be strong. If the
relationship is not strong (e.g. casual acquaintances or friends of friends) then the ties are
said to be weak ties (Valente, 1995).
Embedded within social networks is the concept of social support and social
influence. People in the same social networks can provide social support and both
positive and negative influence to each other. Social support can be broken down into
four types of support: emotional, instrumental, informational, and appraisal support
(House, 1981). These forms of support can serve as a protective factor when one is faced
with life’s stresses. Unlike social support which is intentional and meant to be helpful,
social influence is a “receiver-initiated social comparison” process (Heaney & Israel,
1997; Taylor et al., 1990). Social influence is usually unintentional and only comes to
fruition when it is received and processed by an individual. The field of social networks
and what it encompasses is very broad so we’ll be focusing only on the contextual side of
social networks and how it is used to understand the dynamics between individuals
within the same networks.
Figure 1.1 depicts the theoretical relationship between social networks and
general health. In this figure, social ties between members of a network provide social
influence and social support in the form of emotional, instrumental, informational and
appraisal support. Together with social influence, social support effect individual
behaviors – either positively thereby resulting in good health outcomes or negatively
thereby resulting in sickness or addiction in the case with substance use. In addition,
social ties also affect psychological state resulting in either despair and an enhanced

5

neuroendocrine responses or elation and a suppressed neuroendocrine response and
improved immunity (Cohen et al., 2004). The combined affect of behavior and biological
responses is what eventually affect health outcomes. This particular paper will emphasize
the pathways between social relationships, social influence, social support and behavior.
Figure 1.1 Social Networks and Health (Adapted from Cohen, 2004)

Research studies across many disciplines have been conducted on the social
networks platform. Between 1971 and 2003, a social network of approximately 12,000
people in the Framingham Heart Study was used to examine the role of social networks
in the spread of obesity. The study found that an individual’s likelihood of becoming
obese increases by 57% if he/she were friends with someone who also became obese in
that time period and by 40% if this other person was a sibling. The study also found that
same sex friends were more likely to have greater influence on weight than opposite sex
Social Relationships/Social Ties/Social Links
Social
Influence
Psychological
States
Social Support
Health Promoting or Deteriorating Behaviors Neuroendocrine
Response
Health Outcomes (i.e. good health or illness, physical and/or psychiatric)

6

friends (Christakis & Fowler, 2007). A study conducted in Stockholm, Sweden found that
people with no close network ties were at an increase risk of developing dementia and
that a large social network was protective against this illness. More importantly,
satisfactory contacts with people within the networks were protective even if such
contacts were rare (Fratiglioni et al., 2000). Social networks of injection drug users have
also been studied. Injection drug users in Baltimore, Maryland whose networks consist of
drug users are at an increase risk for HIV whereas those who reported being in networks
with non-users decreased their risk (Costenbader et al., 2006). Interventions designed to
reduce HIV risk in injection drug users were also more effective when it was
implemented at the social network level rather than solely on the individual level
(Neaigus, 1998). Among networks of homosexual men, HIV prevention intervention
using peer leaders showed a decrease in unprotected sexual intercourse and reduced risk
of contracting HIV (Amirkhanian et al., 2005; Kelly et al., 1992).

Adolescence
Adolescence is the period between childhood and adulthood. It is difficult to
specify a particular age range in which adolescence occur since the time period varies
across ethnic groups, cultures and countries. What might be considered as a rite of
passage from childhood to adolescences in one culture might not be considered the same
in another. What’s absolute is that adolescence is a time of physical, psychological,
emotional and social changes. Puberty causes physical changes in early adolescence
making boys look more like men and girls more like women. Psychologically,

7

adolescents go through changes in intellectual processes that bring them closer to
adulthood. According to the theory of cognitive development by Jean Piaget, beginning at
age 12 years and onward, boys and girls start to fully develop formal operations which is
a stage when abstract thinking, logical reasoning and problem solving occur (Elkind,
1978; Piaget & Inhelder, 1969). It is at this time that adolescence poses the most threat to
ones behavior since logical thought processes are still forming.
Of interest to this paper is how peer influences affect adolescent life. A study
looking at the influence of best friend’s sexual activity among Caucasian and African
American students in grades 7 through 9 showed that Caucasian females were influenced
by the sexual behaviors of both their male and female best friends. That is, virgin girls
whose best friends were sexually active at time 1 would have had sexual intercourse by
time 2 (Billy & Udry, 1985). Adolescent peer influence was a stronger predictor of body
dissatisfaction among overweight and at risk of being overweight girls in a study
conducted in Florida (Thompson et al., 2007). Adolescent performance and participation
in schools are also influenced by best friends. A study of approximately 300 seventh and
eighth grade students found that students who reported best friends that have good
behaviors improved their behaviors during the school year. Those who reported having
best friends who were disruptive and had bad behavior increased their disruptive
behaviors during the school years. The study also found that girls are more highly
influenced by their best friends than boys (Berndt & Keefe, 1995).
These peer influences seemed more pronounced when it involves risky or
antisocial behaviors such as substance use and abuse. A 2005 study looking at risk taking

8

among 306 individuals divided into three different age groups (i.e. 13-16 years, 18-22
years and 24+ years) found that individuals took more risk and are more likely to think
about the benefits of these risks rather than the consequences when in a peer group
(Gardner & Steinberg, 2005). A study looking at grade point averages and drug use in
ninth thru eleventh grade students showed that friends’ grades and drug use are predictive
of individual grades and drug use but moderated by parenting style (Mounts & Steinberg,
1995). Students who reported high achieving friends also received better grades and those
who reported friends with poor grades and drug use also reported similar behaviors. A
study looking at 446 Caucasian and Hispanic youths found that among substance users,
peer influence is stronger than parental influence. In this group, marijuana use by friends
is the strongest predictor of drug use but a good parent-child relationships lessen the
importance peer influences (Coombs et al., 1991).
Social networks among adolescents have been extensively researched. Aggressive
behavior has been linked to social network interactions among both male and female
adolescents (Coie et al., 1995). Studies of boys and girls’ networks in fourth and seventh
grade found that highly aggressive adolescents were less popular than their peers but
were still identified by peers as central members within the network. In addition,
aggressive adolescents were friends with other aggressive adolescents (Cairns et al.,
1988). Social networks influences have also been observed in adolescent sexual
behaviors (Jorgensen et al., 1980). A study looking at African American girls in low
income neighborhoods find evidence to suggest that a strong parent-child relationship
and having more working adults in the adolescent social network is associated with

9

delayed sexual experience (Moore & Chase-Lansdale, 2001). Social networks and
network influences have also been linked to other behaviors such as food choice and food
intake (Feunekes et al., 1998), physical activity (Voorhees et al., 2005), homeless or
runaway youths (Johnson et al., 2005), language use (Eckert, 1988) and depression
(Olsson et al., 1999).
Within the field of adolescent substance use, social networks have been used to
create prevention interventions and better understand why they use drugs. Social
networks have been employed in creating interventions where students nominate their
peers to serve as leaders who then assist with implementing program curriculum and
serve to influence their peers (Valente et al., 2003; Valente et al., 2007). A school-based
intervention study used student leaders to assist in a tobacco prevention program.
Students were asked to nominate their peers to serve as leaders and those who received
the most nominations were identifying as peer leaders to help distribute program
materials, lead group discussion and organize group activities. Results showed that
students in the peer-led group liked the program more than the comparison groups and
had improved attitudes and self-efficacy toward smoking cessation and decreased
intentions to smoke (Valente et al., 2003). A panel study of roughly 5,100 sixth, seventh
and eighth grade students found that adolescents with less dense networks and networks
with high smoking prevalence were more likely to smoke and use marijuana (Ennett et al.,
2006). Another study looking at over 1, 000 junior and high school students found that
friendship groups predicted the transition from initiation to current use of cigarettes and
that close friends predicted the transition to current alcohol use (Urberg et al., 1997).

10

Social network analysis of ninth grade students revealed that in four out of five schools
being studied, students with the least number of best friends were more likely to be
current smokers (Ennett & Bauman, 1993).
This long list of studies on adolescent drug use and social networks reminds us of
the importance of network influences on adolescents. How much and how often, if any,
adolescents use drugs is largely determined by peer influences especially among high risk
adolescents.

High-Risk Adolescents
“High-Risk” Defined
This study’s population of adolescents is considered “high risk”. The definition of
“high risk” varies according to the variables of interest (Dryfoos, 1990). It depends on the
subjects that are being studied, their characteristics such as age, sex and ethnicity,
predisposing factors such as social status and environment and causal factors, all of
which are determining factors of behaviors (Dryfoos, 1990). A review of 29 programs
targeting high risk youth found 9 different definitions of high risk (Sussman et al., 2004).
These youths are often referred to being “at-risk” of certain behaviors such as smoking,
drinking, substance use, violence, and risky sexual behavior which may eventually lead
to unwanted pregnancies or sexually transmitted diseases. In terms of this paper,
adolescents considered as “high risk” for substance abuse are defined by the proportion
of drug users within their surroundings (Johnson et al., 1990; Pentz, 1994). That is, if the
proportion of drug users within a school is high, the adolescents who attend those schools

11

are considered high risk or at risk for substance use. These youths are often the target of
many prevention programs because they are more prone to risky behaviors than other
youths.

Characteristics of High-Risk Adolescents
Generally high risk youths are often those who come from disadvantaged
socioeconomic groups, children of substance-users, risk-takers, those suffering academic
problems, or persons who are targets of drug promotions by the tobacco and alcohol
industries (Sussman et al., 2004). As defined earlier, risk increases with the percentage of
users within a social environment (Johnson et al., 1990). It serves as a rationale for
developing and implementing prevention programs that specifically targets high risk
youths. Such programs should include motivation, skills, and decision making (Sussman
et al., 2004), in addition to skills for making changes (i.e. effective listening,
communication, and self control) (Watson & Tharp, 2002).
High risk adolescents are at an increase risk of being substance users. Of the
many possible delinquent behaviors exhibited by high risk adolescents such as physical
violence (Broidy et al., 2003; Orpinas et al., 1995) and risky sexual behaviors like
unprotected sex (Biglan et al., 1990; Luster & Small, 1994; Turner et al., 1998),
substance abuse ranks as one of the most prevalent and serious. It is one of the major
causes of morbidity and mortality among adolescents in the United States (Sussman et al.,
2008). These substances can range from easily accessible drugs such as alcohol and
cigarettes to drugs that were more difficult to obtain like cocaine, hallucinogens and

12

heroine. Due to the nature of their social networks and environmental surroundings, high
risk adolescents are at an increase risk of being substance abusers and suffering the
consequences of substance use.

Substance Use
Substance use has been extensively mentioned in the previous sections because it
poses such a risk to adolescents and thus deserves to be reviewed. Substance abuse
among high risk adolescents is a major concern of our society. Section 321(g)(1) of the
Federal Food, Drug and Cosmetic Act defines drug as “articles intended for use in the
diagnosis, cure, mitigation, treatment, or prevention of disease in man or other
animals…intended to affect the structure or any function of the body of man or other
animals” (FDA, 2004). The Controlled Substances Act further defines controlled
substances as any substance “which has a stimulant, depressant, or hallucinogenic effect
on the central nervous system”. Specific definitions of drugs such as marijuana and
opiates and drug-related terms such “distribute”, “dispense” and “addict” are also given
in this document produced by the United States Food and Drug Administration (FDA,
2004).
According to the National Survey on Drug Use and Health published by the
Office of Applied Studies at the Substance Abuse and Mental Health Services
Administration (SAMHSA) in 2008, 9.3 percent of youths aged 12-17 currently use illicit
drugs – 6.7 percent marijuana, 2.9 percent prescription-type drugs, 1.1 percent inhalants,
1.0 percent hallucinogens and 0.4 percent cocaine (SAMHSA, 2009). As expected more

13

males age 12 years and older used illicit drug than females of the same age group (i.e. 9.9
percent versus 6.3 percent). However, the rate of drug use increased in females between
2007 and 2008 (i.e. from 5.8 percent to 6.3 percent) while that of males remained
relatively the same. In terms of alcohol use, more than half of all Americans age 12 years
and older are current users and more than 20 percent of these individuals binge drink.
What’s more alarming is that 13.1 percent of youths aged 14-15 and 26.2 percent of
youths age 16-17 currently use alcohol. Among smokers, youths age 18-25 had the
highest rate of tobacco use at 41.4 percent. Past month tobacco use data showed that 35.7
percent of these uses cigarettes, 11.3 percent were cigars and 5.4 percent were smokeless
tobacco (SAMHSA, 2009).
It is well understood that substance abuse is linked to physiological, psychological,
social and emotional problems among adolescents. Among the most serious physiological
problems resulting from drug abuse is the impairment of the central nervous system
functions. A study published in 2006 found significant slowing down of simple reaction
time task (i.e. a basic measure used to test the function of the central nervous system)
among male and female heroin ex-users between months 1-3 after withdrawal (Liu et al.,
2006). Another study looking at substance abuse patients and matched controls at the
same health maintenance organizations found that more than one third of the conditions
they examined (i.e. medical and psychiatric) were more common in substance abuse
patients and that pain-related diagnoses such as arthritis and headaches were more
common in the substance abuse group as well (Mertens et al., 2003). In addition to the
high rates of transmitted infections such as HIV and hepatitis, rates of physical violence,

14

mortality and morbidity were also high among injection drug users (Marshall et al., 2008).
Social and emotional impacts of drug use as an adolescent on later life as a young adult
was studied in a sample of approximately 650 subjects. Results showed that adolescent
drug use was associated with early marriage, unhappiness during the marriage and later
divorce (Newcomb & Bentler, 1988). The same study also showed that adolescent drug
use was associated with drug crimes, increase stealing, and reduced college involvement.
Depression, developmental delays, withdrawal and apathy are some of the psychological
problems that result from substance use. Adolescents who are substance users are at a
higher risk for depression, conduct disorders, personality disorders, and suicide (Crowe &
Bilchik, 1998). A 1992 article, cited decrease commitment to education and increase
truancy rates as being associated with substance use among adolescents (Hawkins et al.,
1992). The Survey of Youth in Custody, 1987 interviewed close to 3,000 adolescents in
50 long-term, State-operated juvenile correctional institutions. More than 80 percent of
the respondents use illicit drugs and 47.6 percent were under the influence of drugs or
alcohol at the time of their offense (Beck et al., 1988).

Reasons for adolescent substance abuse
Adolescents use drugs for many reasons. Studies looking at psychological health
among adolescents showed that drug abuse is related to personal and social
maladjustment (Shedler & Block, 1990). Some studies indicate that male adolescents
tend to use drugs for pleasure and a sense of belonging while female adolescents use
drugs to relieve stress and cope with the problems they face (Mooney et al., 1987;

15

Newcomb et al., 1988; Novacek et al., 1991). However, there are also evidence to suggest
that both males and females adolescents use drugs to cope with stress (Kilpatrick et al.,
2000). A national survey of approximately 4,000 adolescents aged 12-17 years found that
adolescents who were physically and sexually assaulted, had witness others being
assaulted and those who had family members who were substance abusers, are at an
increase risk for substance abuse themselves (Kilpatrick et al., 2000). The same study
also showed that posttraumatic stress disorder also increases the risk of marijuana and
hard drug abuse among adolescents. Data from the National Women’s Study found that
alcohol abuse as adults was associated with childhood rape (Epstein et al., 1998). Another
study among 285 women also found that avoidance coping was associated with increase
substance use and psychological distress (Min et al., 2007). Thus, drug prevention
programs that introduce positive coping behaviors have been found successful (Fok &
Wong, 2005). The reasons why adolescent use drugs may best be depicted in the
following list taken directly from Hawkins, Catalano and Miller (1992):
laws and norms favorable toward drug use; availability of
drugs; extreme economic deprivation; neighborhood
disorganization; certain psychological characteristics; early
and persistent behavior problems including aggressive
behaviors in males; other conduct problems; hyperactivity
in childhood and adolescence; a family history of
alcoholism and parental use of illegal drugs; poor family
management practices; family conflict; low bonding to
family; academic failure; lack of commitment to school;
early peer rejection; social influences to use drug;
alienation and rebelliousness; attitudes favorable to drug
use; and early initiation of drug use (Hawkins et al., 1992)

16

Hispanic Adolescents
The Office of Management and Budget defines Hispanics as “a person of Cuban,
Mexican, Puerto Rican, South or Central American, or other Spanish culture or origin
regardless of race” (OMB, 1997). As of July 1, 2008 the U.S. Census Bureau reports a
total of 46.9 million people living in the United States of Hispanic or Latino origin
(Census, 2008). That is 15 percent of the nation’s population, making Hispanics the
largest minority group. Hispanics are also the largest growing population in the U.S. with
projections of around 132.8 million people or 30 percent of the nations’ population by
2050 (Census, 2008). Among youths ages 10-14, approximately 3, 986, 000 are
Hispanics (51.1 percent male), among those 15-19 years of age, 3,799,000 are Hispanics
(51.1 percent male) and among those ages 20-24, there are 3,617,000 Hispanics (52.5
percent are male) (Census, 2008).
Like every ethnic minority, Hispanic adolescents face many unique challenges. A
study from the University of California, Berkeley found that among Hispanic boys, U.S.
born Hispanics who were not well acculturated had more stress and self-esteem problems
than bicultural U.S.-born Hispanic adolescents (Gil et al., 2006). Acculturation strains
such as language conflicts, acculturation conflicts, perceived discrimination, and
perception of a closed society were studied in a sample of 2, 360 Hispanic adolescents in
Miami, Florida. Results indicate that acculturation strains in the school setting may affect
performance and lower educational aspirations among U.S. born and foreign born
Hispanic adolescents (Vega et al., 1995). Cultural identification, or identifying oneself
with a specific culture and cultural norms, may also guide Hispanic adolescent behavior.

17

High levels of cultural identification may lead to adolescents behaving in a way which is
deemed acceptable in that culture. If substance use was frowned upon in one’s culture,
someone with high cultural identification will not use drugs (Strait, 1999).
Among Hispanic youths, the rates of drug use are also alarming. About 6 percent
of Hispanics age 12 years and older user illicit drugs in the past month. About 25 percent
of Hispanics reported binge alcohol use and among those aged 12-17 years, 14.8 percent
of Hispanics reported current use. Over 20 percent of Hispanics aged 12 years and older
are current tobacco users. Among youths aged 12-17 years, 7.9 percent are cigarette
smokers and amongst those 18-25 years, 30.0 percent currently smoke. In 2008,
approximately 9.5 percent of Hispanics reported substance dependence or abuse
(SAMHSA, 2009).
What seems to be unique about Hispanic adolescents and substance use is their
strong ties to family and family values. Family influence and support are negatively
associated to substance use among Hispanic adolescents (Coombs et al., 1991; Pabon,
1998; Rodriguez & Weisburd, 1991). A study looking at family support on gang
involvement, alcohol, tobacco and marijuana use, among a predominantly Hispanic
sample found that perceived family support reduced the influence of deviant peers on
tobacco and marijuana use (Frauenglass et al., 1997). One rationale for this unique
association is that Hispanic adolescents rely more on family support than Caucasians
(Booth et al., 1990; De La Rosa, 1988). In turn, families have a strong influence on the
intentions and behaviors of Hispanic adolescents.

18

Regardless of race or ethnicity, the growing number of substance users in the
United States reminds us if the importance of drug prevention programs. Although drug
prevention interventions do exist outside the school system, the most efficient types of
programs have been conducted in the school setting.

Substance Abuse Prevention Programs
A variety of school-based prevention programs currently exist. A quick internet
search reveals hundreds of prevention programs targeting more than 15 different fields
including substance use, youth violence, suicide, obesity, diabetes, teenage pregnancy,
skin cancer, hearing loss, depression, eating disorders, school drop-outs, and dental caries.
Such programs are implemented at all levels from elementary to high schools. The
(S)Partners for Heart Health program is an intervention involving partnerships with
physical education teachers in elementary schools to improve the cardiovascular health of
fifth grade students through bi-monthly lessons on heart healthy nutrition and heart
healthy physical activities (Carlson et al., 2008). A group in Germany designed a school-
based program to prevent the increase of depressive symptoms in eighth grade students
known as LISA-T which is based on cognitive-behavioral therapy concepts. LISA-T was
effective in reducing symptoms in students with minimal to mild depressive symptoms
(Possel et al., 2004). New Moves is a multi-component school-based physical education
class designed to prevent obesity among adolescent girls. Findings showed that girls who
took New Moves perceived positive program impact on their physical activity, eating
behaviors and self-image (Neumark-Sztainer et al., 2003). The examples given are only a

19

few of the many school-based prevention programs that exist. However, pertinent to this
paper are substance abuse prevention programs.
Substance abuse prevention curriculum in schools has become a priority in most
Western countries. In fact, in some countries, schools are required by law to run specific
programs or messages about drug use and adopt a “reference” in the national curriculum
on substance abuse (Cuijpers, 2002). Because of the level of importance that schools and
society, in general, have placed on substance abuse prevention in youths, the median
number of different prevention activities currently taking place in the typical school is
fourteen (Gottfredson & Gottfredson, 2001). Nine out of ten schools provide information
about tobacco, alcohol, drugs, violence, accidents, health or mental health, and risky
sexual behaviors (Gottfredson & Wilson, 2003). Three out of four schools provide
prevention lessons in the form of curriculum instructions. The number of activities per
school range from 0 in some schools to about 66 in others. Overall, there are over 200
school-based substance abuse prevention programs available (Gottfredson & Wilson,
2003).
With so many school-based substance abuse prevention programs available,
selecting the right program is very important. Systematic reviews of effective school-
based drug prevention programs found several important characteristics (Cuijpers, 2002;
Gottfredson & Wilson, 2003). Effective programs (1) are evidence-based and proven to
be effective (White & Pitts, 1998); (2) provide resistance-skills training or training for
adolescents to resist social-influence to engage in substance abuse (Botvin, 1990); (3) are
aimed at changing normative beliefs about drug use (Gottfredson, 1997; Hansen, 1992);

20

(4) are environmentally focused (Wilson et al., 2001); (5) provide interactive learning
amongst the participants rather than a one-way learning environment (Black et al., 1998;
Tobler & Stratton, 1997); (6) consider the training and background of the leaders or
educators (Hansen, 1992) (7) utilize peer-leaders or peer group components (Black et al.,
1998; Rooney & Murray, 1996); and (8) are longer in duration and provide “booster”
sessions (Durlak, 1995; Gottfredson & Wilson, 2003). An example of an intervention
program with some of these fundamental components is an intervention program
conducted in the Transciplinary Prevention Research Center.

Transciplinary Prevention Research Center
Data for the three studies will be derived from datasets collected at the
Transciplinary Prevention Research Center (TPRC) at the Institute of Prevention
Research housed within the University of Southern California’s Keck School of Medicine
between 2002 and 2007. TPRC was funded by the National Institute on Drug Abuse to
conduct two main research projects – Project One which looked at memory and implicit
cognition and Project Two which looked at social networks. Both projects utilized Project
Towards No Drug Abuse (TND), an empirically proven intervention previously used in
schools, as the medium for the studies. One-year follow-up study of TND has shown
decreased use of cigarettes, alcohol, marijuana, hard drugs, and weapons carrying
(Sussman et al., 2002). Effects of TND on cigarette smoking were also maintained at
two-year follow-up (Sussman et al., 2003).

21

Schools
The intervention was a randomized controlled trial conducted in Southern
California continuation high schools. Continuation high schools are schools for high-
school-age adolescents who cannot continue in the traditional high school setting due to
emotional, behavioral or function problems (Sussman et al., 1997). Out of 25
continuation high schools districts which were recruited for the study, a total of 8 school
districts agreed to participate. One school district was used for the pilot study and the
remaining 7 provided classrooms for the main study. A total of 14 continuation high
schools from these 7 districts provided 75 classrooms for the intervention. Twenty-eight
classrooms were used as the control, 22 received the standard curriculum and 25 received
a revised social-network based curriculum.

Intervention
Schools were randomized to one of three study conditions – control, standard
TND or TND-Network. Project TND is a 12-session school-based curriculum designed to
motivate youths to change their perspectives and perceptions of drug use and teach learn
social skills, life skills and decision making techniques to help them plan good solutions
to difficult problems and situations (Sussman, 1996). Students learn to role play and defy
myths regarding drug use among adolescents like themselves. In the TND-Network
curriculum, the standard TND curriculum was revised to involve the use of peer-leaders
and more group interaction amongst the students (Valente et al., 2007). Table 1.1

22

provides a brief description of standard TND and table 1.2 describe ways in which TND-
Network made modifications to the standard curriculum.
Table 1.1 Brief Description of Standard TND Curriculum (Adapted from Skara et
al., 2005)
Session Description
1. Communication and Active
Listening
Students are introduced to TND and learn how to be active
listeners.
2. Stereotyping Students learn that believing in stereotypes can lead to self-
fulfilling prophecies and puts them at risk for substance
abuse.
3. Myths and Denial Students learn to identify myths involving drug use, how to
distinguish fact from fiction and how people use myths to
deny or justify their use.
4. Chemical Dependency Students learn about the consequences of drug use and
dependency.
5. Talk Show Students role play and learn about the physical, emotional
and social consequences of drug use.
6. Marijuana Panel Students learn about the consequences of marijuana through
group activities.
7. Tobacco Basketball and Use
Cessation
Students play a questionnaire game and learn about the
consequences of tobacco in addition to cessation.
8. Stress, Health and Goals Students learn about ways to cope with stress and the
importance of health.
9. Self-control Students learn to identify their own self-control.
10. Positive and negative
Thought and Behavior Loops
Students learn about their thoughts, both positive and
negative, affect their choices and behaviors.
11. Perspectives Students present their views on various topics such as
smoking and drug use.
12. Decision-making and
Commitment
Students learn they have many choices when it comes to
drug use and make commitments to themselves about drug
use.

23

Table 1.2 Brief Description of Modifications made in TND-Network Curriculum
(Adapted from Valente et al., 2007)
Session Modifications
A. Peer Leader Training Added to specifically train peer leaders and discuss what it
means to be a good leader.
1. Communication and Active
Listening
Process questions were made into group discussion formats
instead of facilitated by the health educators.
2. Stereotyping Students complete activities in a group format.
3. Myths and Denial Peer leaders serve as classroom assistants and students
discuss myths in their group instead of the classroom. Peer
leaders summarize the discussion and share it with the class.
4. Chemical Dependency Peer leaders review information with group. As a group,
students select items from toolkit and share it with family
and friends.
5. Talk Show --
6. Marijuana Panel --
7. Tobacco Basketball and Use
Cessation
Changed into teams playing a game similar to “horse”.
8. Stress, Health and Goals Added group discussions on ways to handle stress and
decide which suggestions are most helpful.
9. Self-control Peer leaders model behaviors and students discuss ways
they can be supportive.
10. Positive and negative
Thought and Behavior Loops
Added group discussions on how violence can be avoided.
11. Perspectives --
12. Decision-making and
Commitment
Students discuss ways that drug use can impact their lives
and peer leaders support positive commitments.

Peer leaders were identified prior to the start of the curriculum by student
nominations. The students were asked to nomination people in the class who they felt
made good leaders. Data from these surveys were then used to identify peer leaders in
each classroom and group students into individual groups led by the peer leaders. Peer
leaders were trained on facilitating group discussions, program materials the week prior
to each session and assisted health educators with delivering program messages. In the
standard condition, the classroom was divided into two teams and a game is played at the
end of each session to review materials and concepts taught during class. In the network

24

condition, network-based groups served as the team and stayed together throughout the
12 session curriculum. Peer leaders facilitated these small group discussions.
Only one session lasting approximately 45-60 minutes is delivered in the
classroom each day for a period covering 3-4 weeks. A total of 16 trained health
educators were used to deliver the curriculum. In the network classrooms, peer-leaders
lead group discussions and assisted the health educators with program delivery. In all, 47
classrooms received either standard TND or TND Network over a course of 9 months.

Data Collection
Baseline data were collected approximately one week prior to implementation of
the study. Baseline data collected information on basic demographic characteristics, main
effects variables of interest for the study and network nominations. Pre-tests were
collected prior to the start of the first curriculum and contained similar questions as those
on the baseline survey but with the addition of knowledge questions pertaining to the
curriculum. Post-tests were collected after the 12
th
session was implemented. Follow-up
data was collected approximately 12 months after the completion of the study. Post-test
and follow-up surveys collected information such as drug use and program effects.
Except for the follow-up data, all observations were collected in person on the school
grounds. Follow-up data were either collected in person by research staff or via telephone
surveys conducted by trained staff members. See figure 1.2.

25

Figure 1.2 Study Design and Data Collection (Adapted from Valente, 2007)

Overall Demographic Characteristics
Table 1.3 provides demographic characteristics of the study population at three
points in time – baseline, post-test and follow-up. Approximately 887 students
participated in the baseline survey consisting of 255 (29 percent) students in the control
condition, 296 (33 percent) in the standard TND condition and 336 (38 percent) in the
network condition. Out of 671 students who took part in the post-test survey, 178 (27
percent) students were in the control condition, 225 (34 percent) in the standard TND
condition and 268 (40 percent) in the network condition. Among the follow-up group, a
total of 552 students participated in the 1- year follow-up with 143 (26 percent) students
75 Classrooms from 14 continuation high schools
Administered baseline survey
28 Control classes 22 Standard TND 25 TND Network
Administered Post-test (after 12
th
session)
Administered one year follow-up survey
Pre-test survey

26

in the control condition, 186 (34 percent) in the standard TND condition and 223 (40
percent) in the network condition. Information on ethnicity was collected at baseline and
follow-up but not at post-test. Mean age of the students was about 16 years old with
grade 10 as the mean grade. Males made up approximately 60 percent of the student
population. In terms of ethnic groups, Hispanics/Latinos made up the largest ethnic group
amongst the student population with approximately 60 percent at baseline and 70 percent
at follow-up. Caucasians made up the second largest group followed by Blacks/African
Americans and those of mixed ethnicities.

27

Table 1.3 Demographic Characteristics of the Study Population
Baseline Post-test Follow-up
Total Control TND Network Total Control TND Network Total Control TND Network
N 887 255 296 336 671 178 225 268 552 143 186 223
Mean Age 16.40 16.41 16.42 16.38 16.35 16.51 16.36 16.23 16.84 17.61 16.94 16.46
Mean Grade 10.61 10.80 10.66 10.41 10.58 10.86 10.60 10.37 10.81 11.57 10.86 10.49
Males (%) 60.04 61.48 57.19 61.49 58.59 60.33 56.58 59.11 61.24 51.85 60.00 66.13
Ethnicity
(%)
Asian/Asian
American
2.67 2.39 4.14 1.56 1.27 2.10 1.08 0.90
Black/Af.
American
6.27 3.98 5.86 8.44 6.70 3.50 6.99 8.52
Hispanic/
Latino
68.99 71.71 67.93 67.81 72.83 73.43 74.73 70.85
White/
Caucasian
12.66 13.15 11.38 13.44 9.60 12.59 6.99 9.87
American
Indian/Nat.
Indian
0.81 1.20 0.69 0.63 0.36 0.70 0 0.45
Mixed 8.01 7.57 9.31 7.19

Ethnicity not measured at post-test
5.62 6.29 5.38 5.38

28

Table 1.4 provides information on substance use behaviors of the adolescents at
baseline, post-test and follow-up. This information provides a quick glance at the usage
behaviors of the current student population. Students were asked how many times they
used the various drugs in the past 30 days and were given answers choices that ranged
from 1 for zero times, 2 for 1-10 times, 3 for 11-20 times, 4 for 21-30 times, 5 for 31-40
times, all the way up to 11 for 91 or more times in the past 30 days. Mean use ranged
from 1.042 to 2.768 over the course of the intervention and one-year follow-up. Alcohol,
marijuana, and cigarettes were most often used by students at all points in time. On
average, students reported that they used marijuana the most at baseline and post-test and
used cigarettes the most at one-year follow-up.

Table 1.4 Last 30 Day Drug Use by Time Points*
Drug Item Baseline (mean) Post-test (mean) Follow-up (mean)
Alcohol 2.354 2.508 1.931
Marijuana 2.622 2.768 1.949
Cocaine/Crack 1.241 1.324 1.104
Cigarettes 2.382 2.452 2.334
Ecstasy 1.113 1.197 1.051
Hallucinogens 1.111 1.233 1.072
Stimulants/Amphetamines 1.345 1.411 1.147
Tranquilizers 1.110 1.195 1.057
Opiates 1.126 1.182 1.088
Inhalants/Vapors 1.187 1.292 1.081
Other club/party drugs 1.107 1.231 1.042
* 1 = 0 times, 2 = 1-10 times, 3 = 11-20 times, 4 = 21-30 times, 5 = 31-40 times, 6 = 41-50 times,
7 = 51-60 times, 8 = 61 - 70 times, 9 = 71 -80 times, 10 = 81-90 times, 11 = 91 or more times in
the past 30 days

Outcome Measures
The use of alcohol, cigarettes, marijuana, hard drugs and all drugs at various
points in time will be used as the outcome variable for the three papers. Initially based on

29

a total of 11 drug items including alcohol, marijuana, cocaine/crack, cigarettes, ecstasy,
hallucinogens, stimulants/amphetamines, tranquilizers, opiates, inhalants/vapors, and
other club/party drugs (see table 1.4), the current analysis will look at alcohol, cigarettes,
and marijuana separately and take the average of the remaining drug items to represent
“hard drugs” and the average of all the drug items to create a composite drug index
variable – “all drugs”. Descriptive statistics showed that data on the drug use variables
are not normally distributed and skewed. To best fix the problem, the drug use variables
were recoded and changed into dichotomous variables – 0 for no use and 1 for use (i.e. at
least once in the past 30 days).

Covariates
Information on age, sex, socioeconomic status (SES) and ethnicity at baseline will
be used as covariates for this analysis. Sex will be coded as a dummy variable with males
= 1 and females = 0. Information on the highest grade completed by the subject’s father
and mother will be used as a proxy for socioeconomic status with higher values
representing higher SES. Response on father’s and mother’s education ranged from 1
“not completed elementary school” to 6 “completed graduate school”. The sum of both
parents’ education level will be used for SES. Descriptive analysis of the variable on SES
showed approximately 40 percent missing data and thus the mean of SES was imputed
(mean
SES
= 5.75). Since SES is a sum of both parents’ education level, a mean of 5.75
depicts a low level of education or in this particular sample, low SES. Hispanics make up
more than 60 percent of the study population and thus, a “Hispanic” dummy variable was

30

used to represent ethnicity (i.e. Hispanic = 1 for all Hispanic students, Hispanic = 0 for all
other ethnicities).

Analysis
STATA version 10 will be used for the analysis of this paper. Since the outcome
variables are dichotomous and the intervention was randomized at the classroom level,
mixed model logistic regression using the school and classroom as the random effects
will be used in order to control for classroom and school effects. In chapter 4, peer
network group was added as a third random effect.

31

CHAPTER 2: MEDIATION AFFECTS OF INTENTIONS ON DRUG USE

CHAPTER 2 ABSTRACT
It is well understood that intentions is an important determinant of behavior and that
intentions itself, is governed by attitudes and subjective norms. Drug use behavior is no
different from any other behavior because it is highly governed by behavioral intentions.
OBJECTIVE: The purpose of this analysis is to examine intentions as a mediator of the
relationship between friend approval of drug use and drug use. METHOD: 12-month
follow-up data on drug use was examined in roughly 400 alternative high school students.
Best friends’ approval of drug use was used as the independent variable with use of
alcohol, marijuana, cigarette, hard drugs and a composite drug use score at follow-up as
the dependent variable. Drug use intentions as measured at post-test is the mediator
variable of interest. ANALYSIS: Mixed model logistic regression was used to conduct
mediation analysis as described in Baron and Kenny, 1986. RESULTS: Behavioral
intentions fully mediated the relationship between best friend’s approval of cigarettes and
marijuana. Partial mediation existed for hard drug use at follow-up. Intentions did not
mediate alcohol use. CONCLUSION: Drug use intentions mediated the relationship
between friends’ approval and drug use. This is consistent with past research studies on
intentions and theoretical models such as the Theory of Reasoned Action.

32

INTRODUCTION
What we do and why we do something is usually determined by how we feel and
our motivations. Our behavior is based on our readiness to perform the behavior which is
governed by a variety of interpersonal and intrapersonal factors. As children, we often do
as are told in order to appease our parents. As adolescents and young adults, we want to
be socially accepted so we do things to appease our friends. As adults, our desire to
appease others and feel accepted is still present but we tend to shift our focus to our
spouse or children. The people who are important to us influence our intentions to behave
in a certain way. Our intentions then guide our behavior.

Behavioral Intentions
Behavioral intentions are defined as a person’s perceived readiness or likelihood
to perform a behavior. It is a construct that is part of the Theory of Reasoned Action
which postulates that attitudes and subjective norms determine behavioral intentions
which in turn governs the behavior (Ajzen & Fishbein, 1980; Montano et al., 1997).
Behavioral intention is also part of the extended version of the Theory of Reasoned
Action – the Theory of Planned Behavior (Ajzen, 1991). The same construct is present in
other theories such as the Theory of Interpersonal Behavior (Triandis, 1977) and
Protection Motivation Theory (Rogers, 1983), all of which agree that intentions are the
most important determinant of a behavior (Sheeran, 2002).
A 2002 meta-analysis of meta-analysis from the University of Sheffield in the
United Kingdom studied intentions in detail to determine how well intentions predict

33

behavior, what factors determine this ability and why there is a gap between intentions
and the actual behaviors. With a total sample size of 82, 107 people, the study found that
intentions explain approximately 28 percent of the variance in behavior which the author
suggests is good. The author also breaks down intentions into 4 groups – (1) people who
are inclined to act on the behavior and then act, (2) people who inclined to act on the
behavior and then don’t act, (3) people who are not inclined to act on the behavior and
then act and (4) people who are not inclined to act on the behavior and then don’t act.
The meta-analysis showed that the biggest group to cause the gap between intentions and
behavior are those who intend to act but end up not acting on the intentions. They
account for 47% of the failure from intentions to behavior (Sheeran, 2002).
More contextually, behavioral intentions have been extensively studied in a
variety of fields, both related and unrelated to health. It has been studied in the field of
information technology (Mathieson, 1991), advertisements (Grube & Wallack, 1994),
weight loss (Schifter & Ajzen, 1985), breast self-examination (Lierman et al., 1990),
marketing (Baker et al., 2002; Patterson & Spreng, 1997), condom use (Fisher, 2006;
Sheeran & Taylor, 1999) and many more. A study looking at blood donation behavior
among 157 students, faculty and staff found that attitudes influence behavior but only
through intentions (Bagozzi, 1981). In addition, the study found that the more the
behavior (i.e. donating blood) is performed the less it is influenced by intentions. A study
of roughly 550 African American men looked that the men’s intentions to get tested for
prostate cancer. The showed that belief in the efficacy of the screening and desire for

34

treatment, if actually found to have prostate cancer, were positively associated with
intentions (Myers et al., 2000).
In the field of substance use, intentions have been the subject of many research
studies. Chassin et al. (1981) tested the notion that adolescents’ intentions to smoke
cigarettes are predicted by two theories – the Theory of Reasoned Action (Ajzen &
Fishbein, 1980) and the Problem Behavior Theory (Jessor & Jessor, 1977). In the Theory
of Reasoned Action, intentions are shaped by proximal variables such as attitudes and
normative beliefs. According to Chassin and colleagues, distal variables such as
personality and perceived environment in the Problem Behavior Theory also shape
intentions. Results of this study indicate that both the proximal and distal variables
significantly predicts smoking status among middle and high school students but that the
distal variables accounts for less variance in intentions (Chassin et al., 1981). More recent
works have also looked at the role of intentions on drug use. Alcohol use among
approximately 4, 000 fifth through eight grade students in Michigan was studied using
constructs of the Theory of Reasoned Action and Theory of Planned Behavior. Analysis
found that intentions to use alcohol explained 26 percent of the variance in use and 38
percent of the variance in frequency of use (Marcoux & Shope, 1997). The same analysis
also showed that attitudes, subjective norms and perceived behavioral control explained
about 76 percent of the variance in intentions to use alcohol. A similar study looked at
alcohol use among fifth and sixth grade students. Results of this study showed that family
strongly affected intentions for alcohol use among fifth graders and peers strongly
affected intentions for alcohol use among sixth graders (Webb et al., 1995). An

35

Australian study looked at 225 high school students who were surveyed about their
attitudes, norms, behavioral control, intentions and actual cigarette use behavior. Results
indicated that attitudes, previous smoking behavior, and perception of what his/her
girlfriend/boyfriend thought of smoking were significantly associated with intentions and
that intentions and past behavior predicted actual use (O'Callaghan et al., 1999).

Friend Approval
Social approval is desired by everyone because it brings about a sense of
belonging and relatedness, especially among children and adolescents (Juvonen, 1996). A
sense of belonging and relatedness may then encourage children and adolescents to be
involved in pro-social behavior or the behavior that is perceived as been acceptable
among a particular social group. Like adults, adolescents tend to seek approval from
people within their network such as parents, teachers and peers.
Peers have been documented as a major influence in adolescent drug use. Peers
are a source of knowledge, attitudes, behavioral norms, and values for adolescents
(Deutsch & Gerard, 1995). Several studies have documented that peer influence plays a
major role in adolescent smoking (Botvin & Baker, 1993; Fritz, 2000; Headen et al.,
1991). Having a friend who uses drugs is strongly associated with drug use in
adolescents (Hawkins et al., 1985; Needle et al., 1986; Sussman et al., 1999; Zogg et al.,
2004). Data from a questionnaire completed by approximately 2, 500 high school
students also indicate that peer influence was associated with adolescent alcohol use
(Nash et al., 2005). Among high school students on the US/Mexico border, academic

36

performances, peer influence, parental attitudes and age were important predictors of
alcohol consumption (Almodovar et al., 2006). Social anxiety disorder and marijuana use
disorders were also shown to be moderated by peer use of alcohol and marijuana
(Buckner et al., 2006). In a cross-national study of adolescents in six European countries,
Kokkevi et al. (2007) found that there is a strong association between smoking, going out
in the evening and having friends who smoke. Marijuana and illegal drug use was also
strongly associated with having friends or older siblings who were also users (Kokkevi et
al., 2007). Adolescent alcohol use is related to social activity among peers (Varlinskaya
et al., 2001). Peer influence can sway alcohol use by peer pressure or directly soliciting
use (Graham et al., 1991). Peers can also influence drug use by social modeling (Zogg et
al., 2004) or by providing opportunities where drugs can be obtained (Sussman et al.,
1999). In addition, friends serve to provide positive confirmation that drugs use is
socially acceptable.
Peer approval is considered an important risk factor in adolescent drug use
(Newcomb et al., 1987; Perry et al., 2002). Jenkins, et al. studied the associations of peer
gateway drug use, perceptions of friends’ drug use attitude and family structure in eighth,
tenth and twelve grade students from seventeen school districts in Ohio. Data showed that
adolescents who were at risk of experimenting with drugs were those who had friends
that are using drugs or who perceive that the friend approves of the behavior (Jenkins &
Zunguze, 1998). Leung and Lau (1988) evaluated 1, 668 seventh, eighth and ninth grade
students’ response to questionnaires inquiring about self-esteem, self-concept of physical
ability, social ability, physical appearance, academic ability and delinquent acts. Results

37

showed that perceived approval by peers was associated with delinquent acts (Leung &
Lau, 1988). Another study of 468 fifth and sixth grade students in northern California
found that perceived peer approval of drinking alcohol was associated with less perceived
consequences of the behavior (Grube & Wallack, 1994). A study of older adolescents
found that peer approval was significantly related to alcohol use (Akers et al., 1979).

Conceptual/Theoretical Model
A systematic review of school-based drug prevention programs by Cuijpers
(2002) found several significant mediators of substance use among adolescents.
Important mediators of the Life-Skills Training program against tobacco use were social
acceptability knowledge, normative expectation of peers and adults, and prevalence
knowledge (Botvin et al., 1992). Important mediators in the Adolescent Alcohol
Prevention Trial include prevalence estimates and beliefs about acceptability (Donaldson
et al., 1994). The Midwestern Project Prevention program against alcohol, tobacco and
other drug use listed friends reaction to drug use and intentions not to use drugs as an
important mediator (MacKinnon et al., 1993). In the review, Cuijpers argues that one of
the most important mediators that need to be considered in prevention programs is
intentions (Cuijpers, 2002).
Given that past research on intentions and peer approval have found it to be
associated with substance use in adolescents, the aim of this paper is to examine how
intentions and peer approval are related to substance use. More specifically, do intentions
mediate the relationship between peer approval and substance use? It seems relatively

38

simple to explain that best friends’ approval of drug use would directly affect individual
drug use. In addition, it would also seem that best friends’ approval would affect
individual intentions to use drugs. If friends approve use then intentions to use drugs
would increase which will then direct the behavior. If friends did not approve use then
intentions might decrease and behavior would not take place. Thus it is hypothesize that
best friend approval would affect intentions which then affects drug use, that intentions
would mediate the relationship between friend approval and use. Figure 2.1 depicts this
relationship.
Figure 2.1 Proposed Theoretical Model for Mediation Analysis

METHODS

Chapter 2 will cover a longitudinal study with three time points – baseline which
represents the first time point, post-test which is approximately 3-4 weeks after baseline
and follow-up which is 12 months after post-test.

Best friends’ approval
of drug use at baseline
Intentions to
use drugs at post-test
Drug use
at follow-up
a b
c
c`

39

Measures
Predictor Variable: Best Friends’ Approval of Drug Use
At baseline, best friends’ approval of drug use was measured with the question
“Think of your five best friends: How many of them think it’s OK for someone to use
drugs?” Possible answer choices were 1 for none of them, 2 for 1 or 2 friends, 3 for 3 or 4
friends, and 4 for all 5 friends. Answers to this question were changed into a binary
variable with answers of 1 (none of the best friends think its okay for someone to use
drugs) being recoded to 0 (no best friends think its okay for someone to use drugs) and all
other answers (i.e. at least one best friend thinks its okay for someone to use drugs) was
recoded to 1 (i.e. best friends think it is okay to use drugs).

Proposed Mediating Variable: Drug Use Intentions
Adolescent drug use intention was assessed with the question “How likely is it
that you will use this drug in the next year (12 months)?” with inquiries about the 11 drug
items that have been used earlier. However, unlike previous questions which allowed
students to answer in forms of frequency of use, answers on intentions were limited to 1
for “definitely not”, 2 for “probably not”, 3 for “a little likely”, 4 for “somewhat likely”
and 5 for “very likely”. Questions probing about alcohol, marijuana and cigarette use
intentions were left as is while all other drug items were averaged to represent hard drug
use intentions and a composite drug use item. Data for this item was collected at post-test.
In the analysis, intentions was changed to a binary variable when it was used as the

40

outcome variable to test for its mediation effects (see table 2.1, model 2). At all other
times, intentions was a ranged from 1 to 5.

Outcome Variable: Drug Use
Past 30 day drug use at 1-year follow-up was assessed with the same question that
assessed past 30 day drug use at baseline and post-test. That is, students were asked how
many times they used each drug item in the past 30 days and allowed to provide answers
that range from none to as much as 91 or more times in the past 30 days. A detailed
description of this variable was outlined in Chapter 1.

Covariates
Age, sex, SES, ethnicity (Hispanic), intervention conditions and past 30 day drug
use at baseline will be used as covariates for this study.

Analysis
Mixed model logistic regression will be used to analyze the mediation effects of
intentions on friend approval and drug use. This mediation analysis will be following
Barron and Kenny’s (1986) four step approach:
Let: Y = substance use
X = best friend approval
Z = intentions (proposed mediator)
Step 1: Y = b
0
+ b
1
X

+ error

41

Step 2: Z = b
0
+ b
1
X

+ error
Step 3: Y = b
0
+ b
1
Z

+ error
Step 4: Y = b
0
+ b
1
X + b
2
Z + error
If the effect of Z (i.e. intentions) is significant in step 4 after controlling for X (i.e.
approval), then it can be concluded that some form of mediation exists. If both X and Z is
significant then we will have partial mediation. If X is not significant after controlling for
Z then it’s a full mediation model (Baron & Kenny, 1986; Newsom, 2008).

RESULTS
Descriptive
A total of 380 subjects provided data for this analysis. It is important to note that
follow-up surveys were collected on roughly 540 subjects. However, there were only
about 380 subjects who provided complete data on friend approval at baseline, intentions
at post-test and drug use at follow-up which made this mediation analysis possible.
Demographic characteristics are similar to the information outlined in Chapter 1, table
1.3.

Mediation Analysis
Table 2.1 shows the results of the mediation analysis using odds ratios. Model 1 is
testing the effect of best friend approval as the independent variable on use at follow-up,
the outcome variable. Model 1 shows that best friend approval was significantly
associated increasing the odds of marijuana (OR = 1.43, 95% CI 1.14 – 1.78), cigarettes

42

(OR = 1.31, 95% CI 1.06 – 1.61), hard drugs (OR = 1.49, 95% CI 1.16 – 1.91) and all
drug (OR = 1.35, 95% CI 1.08 – 1.69) use at follow-up. Model 2 looks at the relationship
between best friend approval and the proposed mediating variable, intentions. Results of
this analysis shows that best friend approval increased the odds of having intentions to
use marijuana (OR = 1.43, 95% CI 1.19 – 1.72), cigarettes (OR = 1.19, 95% CI 1.01 –
1.40) and hard drugs (OR = 1.34, 95% CI 1.16 – 1.56). Model 3 then looked at the affect
of the proposed mediator on drug use and found that intentions to use the specific drugs
was significantly associated with alcohol (OR = 1.70, 95% CI 1.45 – 2.00), marijuana
(OR = 1.90, 95% CI 1.58 – 2.29), cigarettes (OR = 1.70, 95% CI 1.41 – 2.05), hard drugs
(OR = 2.97, 95% CI 1.90 – 4.64) and all drugs (OR = 6.65, 95% 3.60 – 12.27) use at
follow-up. Lastly, model 4 takes into account the predictor variable and proposed
mediator. Results show that intentions was significantly related to alcohol (OR = 1.60,
95% CI 1.35 – 1.90), marijuana (OR = 1.88, 95% CI 1.54 – 2.30), cigarettes (OR = 1.73,
95% CI 1.41 – 2.14), hard drugs (OR = 2.69, 95% CI 1.68 – 4.32) and all drugs (OR =
5.64, 95% CI 2.78 – 11.44) use at follow-up. The 4 models tell us that use of marijuana
and cigarettes are fully mediated by intentions. That is, best friend approval on drug use
is mediated by individual behavioral intentions to use marijuana and cigarettes. For hard
drugs, friend approval was partially mediated by intentions.

43

Table 2.1 Results of Mediation Analysis (Odds Ratios are shown)
Use at follow-up
N = 380
Alcohol Marijuana Cigarettes Hard drugs All drugs
Model 1: Pathway a - Independent Variable to Outcome (Use at follow-up)
Age 1.05 0.74 ** 0.98 0.70 ** 1.10
Male 1.20 1.39 0.99 0.87 1.34
SES 0.98 0.99 1.07 0.89 1.07
Hispanic 0.77 0.80 0.79 1.10 0.77
TND 0.97 0.89 0.89 1.28 0.88
Network 0.94 1.26 1.16 1.67 0.85
Use at BL 1.82 *** 1.37 *** 1.62 *** 1.17 3.16 ***
BF Approval 1.19 1.43 ** 1.31 * 1.49 ** 1.35 **
Model 2: Pathway b - Independent Variable to Mediator (Intentions)
Age 1.14 1.00 1.03 1.03 1.06
Male 0.81 0.77 0.81 0.88 1.06
SES 0.92 0.99 0.99 1.07 0.87
Hispanic 0.98 0.80 1.00 0.98 0.90
TND 0.73 0.59 0.71 0.73 0.75
Network 0.96 0.89 0.80 0.85 0.86
Use at BL 2.01** 1.78 *** 1.72 *** 5.60 *** 10.91 ***
BF Approval 1.13 1.43 *** 1.19 * 1.34 *** 1.24
Model 3: Pathway c - Mediator to Independent Variable (Use at follow-up)
Age 1.04 0.78 * 1.06 0.76 * 1.24 *
Male 1.10 1.45 0.96 0.82 1.25
SES 1.04 0.96 1.10 0.85 1.13
Hispanic 1.29 1.14 0.87 1.10 1.21
TND 1.18 0.93 1.24 1.11 1.04
Network 1.09 1.29 1.64 1.08 1.04
Use at BL 1.28 * 1.14 * 1.24 ** 0.57 * 0.92
Intentions 1.70 *** 1.90 *** 1.70 *** 2.97 *** 6.65 ***
* p < 0.05, ** p < 0.01, *** p < 0.001

44

Table 2.1 Results of Mediation Analysis (continued)
Use at follow-up
N = 380
Alcohol Marijuana Cigarettes Hard drugs All drugs
Model 4: Pathway c` - Independent Variable & Mediator to Outcome
Age 1.03 0.76 * 1.05 0.70 * 1.19
Male 1.11 1.57 1.07 0.90 1.28
SES 1.02 0.98 1.08 0.83 * 1.10
Hispanic 1.11 1.00 0.80 1.28 1.06
TND 1.05 0.91 1.18 1.40 0.85
Network 1.01 1.19 1.52 1.48 0.90
Use at BL 1.42 ** 1.19 ** 1.26 ** 0.63 1.81
BF Approval 1.03 0.99 1.25 1.38 * 1.08
Intentions 1.60 *** 1.88 *** 1.73 *** 2.69 *** 5.64 ***
* p < 0.05, ** p < 0.01, *** p < 0.001

45

It is important to note the other significant associations presented in table 2.1. As
expected, baseline drug use was positively associated with follow-up drug use and
intentions in most instances. Age was negatively associated with follow-up drug use of
marijuana and hard drugs (p<0.01) but not intentions. This means that as an adolescent
grows older, his odds of using marijuana or hard drugs decreases.

DISCUSSION
This paper establishes that intentions to use drugs are a mediator of peer approval
at baseline and drug use at follow-up in this sample of continuation high school students.
Findings from this paper are consistent with previous research on the relationships
between approval, intentions and substance use (Cuijpers, 2002; MacKinnon et al., 1993).
Although an association exists between friend approval and drug use, intentions help to
clarify this relationship by telling us that friend approval is indirectly associated with
drug use. Friend approval is related to use via intentions – best friend influence intentions
which influence behavior.
Studies on mediation effects on substance use are extensive. Detailed analysis of
Project Northland, a randomized study of a multilevel community-wide alcohol use
prevention program among adolescents, found several mediators of alcohol use: (1) peer
influence, (2) functional meanings of use, (3) attitudes and behaviors, and (4) parent-
child communication relating to alcohol (Komro et al., 2001). A 20-year longitudinal
study on childhood aggression and late adolescent drug use found that use and delinquent
behavior as children and young adolescents mediated the relationship between childhood

46

aggression and late adolescent drug use (Brooks et al., 1996). Another study showed that
depression partially mediated the association between education and cigarette use and
income and cigarette use. The same study found that among Caucasian adolescents only,
depression partially mediated the relationship between parental education and cocaine use
(Goodman & Huang, 2002).
Although the list of mediators of drug use is extensive, surprisingly, many studies
have documented possible mediators which did not show any significant effect. Self-
efficacy and self-esteem were not significant mediators in the Life-Skills Training
program which looked at 47 schools undergoing a tobacco prevention program (Botvin et
al., 1992). Normative beliefs, decision making skills, self-esteem and resistance skills all
were not significant mediators of drug use in the Drug Abuse Resistance Education
program also known as DARE (Hansen & McNeal, 1997). The Midwestern Prevention
Project documented that resistance skills, perceive peer norms and negative consequences
were all not significant mediators of drug use (MacKinnon et al., 1993). It would seem
that resistance skills would serve as a mediator between peer pressure and drug use in the
Midwestern Prevention Project but it didn’t.

Limitations and Conclusion
This study was conducted among continuation high school students which are
considered high risk adolescents. The results of this analysis might not be generalizable
to the larger population or even in a nearby high school. Another limitation to the results
is that only a subset of the sample population is present in the analysis. As described

47

earlier, we started off with roughly 800 subjects at baseline and had about 540 follow-up
surveys. Since the three main variables of interest were at different time points (i.e.
baseline, post-test and follow-up), we lost more data and thus only had a sample size of
4380 subjects for this mediation analysis.
It is important to note that this analysis only looked at intentions as the mediating
variable. In other instances, peer approval might be the mediator between the relationship
between intentions and drug use. That is, an individual’s intentions to use drugs might be
affected by what his friends feel about drug use or at least, what he perceives are his
friends approval of drug use. A positive intention to use drugs might change if peer
approval was low or might multiply if peer approval was high. In this particular dataset,
we did not measure intentions at baseline and approval at post-test and thus, we can only
look at intentions at post-test as the mediator between approval at baseline and drug use
at follow-up.
Future research in the role of intentions as a mediating variable in the relationship
between friend approval and drug use should look at drug specific approval and
intentions. This study did not question students on specific drug items but rather used one
question to ask about friends’ approval. Friends’ approval of cigarette smoking might not
be the same as friends’ approval of hard drugs like cocaine or hallucinogens. Friends who
are already hard drug users might approve of cocaine use while friends those who are
smoking or occasionally drinking alcohol might not approve of cocaine use. Similarly,
friends’ approval might also differ between low risk and high risk adolescent groups.
Among low risk adolescents where the proportion of drug users is low, friends might

48

approve alcohol use but not approve ecstasy use. Among high risk adolescents where the
proportion of users are high, friends might approve both alcohol and ecstasy use. The
differences between friends’ approval might affect intentions so it is important to look at
drug specific and even group specific approval since norms vary across different groups.

49

CHAPTER 3: BILINGUALISM AND ITS ROLE IN SUBSTANCE USE

CHAPTER 3 ABSTRACT
Studies have shown that bilinguals have greater cognitive skills, more developed
inhibitory control and multiple social networks in which vast resources and support can
be drawn. OBJECTIVE: The purpose of this analysis is to study bilingualism as a
possible protective factor against adolescent drug use. METHOD: Mixed model logistic
regression using school and classroom as the random effect was conducted on
approximately 880 male and female students in alternative high schools. Outcome
variables of interest were baseline use of alcohol, marijuana, cigarette, hard drugs and a
composite drug use variable. Age, gender, ethnicity, SES, acculturation and number of
best friends were used as controlled variables. Individual bilingualism scores and best
friends’ bilingualism scores were the predictor variables of interest. RESULTS: Among
males and females, best friends’ bilingualism decreased the probability of alcohol,
marijuana and cigarette use. Sub-group analysis by sex reveals that among males,
bilingualism was not associated with drug use but among females best friend bilingualism
was associated with decreasing the probability of alcohol, marijuana, cigarettes and
composite drug use. CONCLUSION: Although individual bilingualism is not associated
with drug use, best friends’ bilingualism decreases drug use among females. This finding
suggests the bilingualism levels of close peers are a protective factor among females and
that males and females have different network structures. That is, a difference exists in
how male and female students are influenced by their bilingual peers.

50

INTRODUCTION
In writing about bilingualism, it is easy to jump to the conclusion that
bilingualism is just the ability to speak two languages. However, a further look at
bilingualism makes you wonder about what this ability entails. Can someone who knows
a few words in Spanish claim that they are bilingual? Would being able to find your way
around Tokyo with basic Japanese classify you as being bilingual? To what extent is this
ability and how well do you really have to speak the two languages?
The Merriam-Webster dictionary defines bilingualism as the ability to speak two
languages (Merriam-Webster, 2009). However, several other definitions and
characterizations of bilingualism have also been used. Earlier works in the mid 1900s
have defined bilingualism in a variety of ways. In 1933, bilingualism was described as
the “native-like control of two languages” (Bloomfield, 1933). It was also sometimes
referred to as the identical mastery of two languages (Marouzeau, 1951) or the “complete
meaningful utterances in the other language” (Haugen, 1953) that also involved “passive-
knowledge” of the languages (Diebold, 1961). In looking at bilingualism, we must also
be aware of the differences between early and late bilinguals. Early bilinguals are those
who learned the second language before age 12 while late bilinguals are those who
learned it afterwards (Lenneberg, 1967). For late bilinguals, learning of the second
language is mediated through the first language. More recent works described
bilingualism as the alternate use of two or more languages (Mackey, 2000).
The degree of bilingualism varies and should be viewed based on a variety of
factors such as age of the individual, how and when the language was acquired, language

51

used in school, personal and social attitudes toward the languages, individual ability to
acquire these languages and the topic or subject in which the language is used (i.e. some
might use English to describe a Shakespearian play and use Spanish to describe musical
styles like mariachis) (Ardila, 1998). Some have argued that bilingualism is not strictly
dichotomous where you can easily identify someone as being bilingual or not bilingual
(Ardila, 1998). Bilingualism should be seen as a continuum where the left side represents
monolingualism and the far left represents high bilingualism.
William Mackey identifies bilingualism as a characteristic of the language usage
rather than an incident of the language itself (Mackey, 2000). According to Mackey, the
definition of bilingualism is somewhat ambiguous and it is very hard to tell exactly when
someone becomes bilingual. He suggests a series of points to classify bilinguals which
include degree, function, alternation and interference. The degree of an individual’s
bilingualism involves how bilingual someone is and often involves skills such as listening,
reading, speaking and writing. Function involves the condition under which the
languages are used (i.e. use at home, school, work, mass media, etc). Alternation deals
with how the two languages alternate from one to the other. Interference involves the
interface between the two languages and how well it is keep apart as well as used
together simultaneously.
Are there benefits to being bilingual? Yes, definitely but it is not always
beneficial. Although we would think that bilinguals have an easier time securing
employment and earn higher wages than monolinguals, research suggests otherwise. Data
from the National Adult Literacy Survey showed that bilinguals do have higher wages

52

than monolinguals but this relationship disappears once variables such as education were
controlled because bilinguals in this study tended to have higher educational levels than
monolinguals (Fry & Lowell, 2003). Analysis of the National Longitudinal Study of 1988
showed that school programs supporting bilingual education for students with limited
English proficiency did not improve the students’ educational attainment or earnings
(Lopez, 2002).
Recent movements in the study of the benefits of bilingualism have drawn interest
from many scholars. Although studies in the early parts of the 20
th
century have found
negative associations between bilingualism and intelligence, more recent studies have
found the opposite associations (Barik & Swain, 1976). According to these later studies,
bilingualism has been found to be associated with cognitive functions such as flexibility,
divergent thinking and creativity (Barik & Swain, 1976). One of the first studies looking
at bilingualism and intelligence was published in 1962. Studying 10-year old children
fluent in English and French at McGill University, Peal and Lambert showed that their
bilingual sample performed better on measures of verbal intelligence and non-verbal tests
(Peal & Lambert, 1962). Possible reasons for these differences in intelligence were
described by various researchers. Leopold (1961) suggested that bilinguals learn from
early in life that word sounds and meanings are separate entities. This separation of sound
and meaning leads to an early understanding of words and the subjectivity of languages
which promotes advanced abstract thinking (Leopold, 1961). Ben-Zeev (1977)
hypothesized that the bilingual child’s two languages causes him to develop a strategy
within the brain to help interface and separate the two different languages and therefore

53

develop advanced cognitive skills. She studied a sample of Hebrew-English bilingual
children in four different groups: Hebrew-English speakers in the United States, Hebrew-
English speakers in the Israel, English only speakers in the United States, and Hebrew
only speakers in Israel. Of interest were their performances on intelligence scores using
four subtests of the Wechsler Intelligence Scale for Children – similarities, digit span,
picture completion, and picture arrangements. Results indicate that although bilingual
children had lower vocabulary levels (i.e. knew less vocabulary words), they were more
advanced in processing verbal materials, better at differentiating perceptual distinctions,
better at searching for structure in perceptual situations and more able to reorganize their
thoughts in response to a feedback (Ben-Zeev, 1977).
Bilingualism has also been linked to inhibitory control. According to Bialystok
(2001), bilinguals have enhanced inhibitory controls or the ability to suppress responses
to outside stimulus. When these individuals communicate in one language, they have to
mentally suppress everything involving the other language. Doing this overtime enhances
their inhibitory controls involving language use which spills into other parts of their
cognitive processes (Bialystok, 2001; Bialystok et al., 2004). Colzato et al. (2008) further
studied this association in a study comparing monolinguals and bilinguals. Both groups
were given inhibitory control tests such as stop signal performance (i.e. subjects are given
a cue telling them to execute a specific response followed by another cue to stop the
response), inhibition of return (i.e. cuing the location of a target with some random
stimulus like a flash of light), and the attentional blink (i.e. providing one stimulus to be
identified and quickly followed that by another stimulus). Results showed that

54

monolinguals and bilinguals performed similarly on stop signal reactions but bilinguals
did better on inhibition of return effect and attentional blink (Colzato et al., 2008). In
terms of the link between bilingualism, inhibitory control and substance abuse, some
scholars have suggested that inhibitory control may play a role in suppressing substance
use. Researchers studying cocaine users found that the user’s impairment in inhibitory
control may somehow have contributed to their cocaine abuse (Fillmore & Rush, 2002).
Adolescents with developed inhibitory control may be more likely to suppress their
response to stimuli such as peer pressure or their own impulsivity to experiment with
drugs. Their inhibitory control might somehow prevent them from experimenting with
something they know to be delinquent.
More concrete benefits of bilingualism involves academic performance and more
importantly, social support networks. Data from the United States’ 1990 Public Use
Microdata Samples was used to analyze the association between language use and school
drop-out rates among ethnic minorities such as Vietnamese, Chinese, Koreans, Mexicans,
and Puerto Ricans. Results showed that bilingual students were less likely to drop out of
school than those who were English-only speakers, English dominant speakers or those in
English-limited households (Feliciano, 2001). Another study looking at the 1988 National
Education Longitudinal Study found that bilingualism did not directly affect academic
achievement but that there is a temporary positive affect of bilingualism on academic
achievement among students who speak the native language with their monolingual,
immigrant parents (Mouw & Xie, 1999). Another study looking at approximately 3,000
children of immigrants in south Florida suggests that bilingualism was positively

55

significantly related to academic performance and future educational and occupational
aspirations (Portes & Schauffler, 1994).
In terms of social support networks, being able to speak more than one language
increases social capital which in turn can be used by bilinguals to obtain different types
of support that are necessary for social and academic success (Stanton-Salazar &
Dornbusch, 1995). Stanton-Salazar and Dornbusch define social capital as social
relationships or interactions where one can obtain benefits such as help and resources.
Having a bilingual social network means that adolescents are immersed in cultural norms
and traditions of both their parents and the host country (i.e. Mexican heritage and
American lifestyles). They can draw valuable resources from their Hispanic parents and
grandparents that instill cultural pride, faith and teachings that discourage drug use. In
general, more Caucasian adolescents report drug use as compared to those of Hispanic,
African-American or Asian descent (Bachman et al., 1991). One factor that might
possibly explain the relationship between ethnicity and drug use is that African, Asian
and Hispanic adolescents report stronger social bonds than that of Caucasians (Ellickson
et al., 1999). More specifically, Mexican American adolescents report stronger bonds
with their family members, more cohesive families and more mutual support (Freeberg &
Stein, 1996; Markides, 1986; Mindel, 1980). In turn, these bonds can serve as protective
barriers against drug use. A study looking at illicit drug use among a sample of
adolescents from the RAND Adolescent Panel Study found that Mexican Americans
were more affected by their family than any other ethnic groups (Ellickson et al., 1999).

56

In addition, an adolescent who is immersed in multiple social networks has a lot
of social capital which they can use to gain emotional support, appraisal support and
instrumental support. Having a grandparent who is able to provide hugs, an uncle who
gives compliments or a best friend who provides car rides to the local library can greatly
enhance an adolescent’s self-worth and desire to succeed. Whereas an adolescent who
grows up without these multiple networks and lack social capital may feel less desire to
succeed and develop more anti-social behaviors such as taking drugs and hanging out
with delinquent friends. Studies among minorities have found that competence with the
“home language” or mother tongue is advantageous. Competence with the home
language has a positive affect on social interactions such as the interaction between other
speakers of the home language. People who can speak the home language as well as
English have stronger ethnic identity and greater understanding of the values and beliefs
of the home culture thus making them more pro-social (Cho, 2000).

Acculturation
We cannot talk about bilingualism without mentioning acculturation.
Acculturation is a process by which immigrants adapt to the culture and ways of the host
society or country (Rogler et al., 1991). Adoption of the host language, food, music,
ideals, religions and way of life are some examples of ways immigrants become
acculturated. There are varying levels of acculturation in that some groups may only
adopt some of the host society’s way of life and ideals but keep the language and food of
their native society or country. Higher degrees of acculturation may include adopting

57

language, food, religion, lifestyle and normative beliefs of the host society and not
retaining any aspect of the native country.
Acculturation and bilingualism are not the same and cannot be used in the same
context. Acculturation is concerned with the degree to which individuals adopt the host
society’s culture while bilingualism is concerned with the degree to which individuals
speak the host language and their home language. High levels of acculturation refers to
an individual who has adopted many aspects of the host society while high levels of
bilingualism means that the individual speaks and communicates in both the host
language and their home language equally. High levels of bilingualism can infer lower
levels of acculturation since someone who is highly acculturated has already adopted the
host language as their own and thus tend to not speak the home language well. Vice versa,
high levels of acculturation can infer that the individual is not bilingual.
Although no concrete evidence exists to show that acculturation does in fact
increase substance use, extensive research has been conducted to show a possible link
between acculturation and substance use. Acculturation has shown negative effects on
certain health outcomes in Hispanics and Hispanic Americans, particularly substance use,
dietary habits, and birth outcomes (Gamboa et al., 2005). Random digit dialing surveys of
Hispanics living in San Francisco showed that more acculturated individuals smoked
more cigarettes per day than the less acculturated individuals (Marin et al., 1989). A
study examining the lifetime risk of mental illness and substance abuse disorders among
Hispanics living in the U.S. showed that acculturation was predictive of mental illness
among Mexican-Americans and other Hispanics and predictive of substance abuse

58

disorders among Puerto Ricans (Ortega et al., 2000). A 2004 survey of 63,000 middle and
high school students participating in the Florida Youth Substance Abuse Survey suggests
that acculturation was a strong predictor of substance use among adolescents and that this
affect was mediated by family and peers (Saint-Jean et al., 2008).
Is bilingualism independent of acculturation? While high degrees of bilingualism
seems to serve as a protective factor against substance use, high degrees of acculturation
serves as a risk factor. Although high degrees of acculturation can usually infer low
bilingualism, these two concepts are not measuring the same concepts. Acculturation is
concerned with all aspects of life and culture while bilingualism is only concerned with
language. Thus the role that each concept plays on substance abuse should be
independent. However, it may be safe to say that bilingualism is a concept embedded in
acculturation and thus bilingualism is not independent of acculturation and is somehow
affected by it.
In general, bilingualism is beneficial and protective against problematic behaviors
such as substance use in adolescents in basically three main ways: (1) bilingual
adolescents have greater cognitive skills than their monolingual counterparts, (2)
bilinguals have more developed inhibitory control making them more likely to be able to
suppress their responses to outside stimulus and (3) bilingual adolescents tend to be
immersed in more than one culture and have more connections to social networks and
social support systems which can be protective. Particularly among Hispanic adolescents,
strong ties with family, friends and cultural norms can serve as protective barriers against
anti-social behaviors such as drug use.

59

Theoretical/Conceptual Model
The model being studied will test the association between bilingualism and
substance use. More specifically, bilingualism and best friends’ bilingualism will be used
as the independent variable while alcohol, marijuana, cigarettes, hard drug and all drug
use will be used as the dependent variable with sex, age, socioeconomic status, ethnicity
and acculturation as the covariates in the regression equation. Best friends’ bilingualism
is included in the model because language involves communication between 2 or more
individuals. If an individual’s bilingualism was being evaluated, their friends’
bilingualism cannot be ignored. Figure 3.1 below depicts this relationship.

Figure 3.1 Theoretical/Conceptual Model for Bilingualism

The purpose of this analysis is to examine the role of bilingualism as a protective
factor against drug use in adolescents. It is hypothesized that individual bilingualism

Age

Sex

SES

Ethnicity

Acculturation

# of Friends

Covariates
Individual
bilingualism
Substance use
(alcohol, marijuana,
cigarettes, hard drugs and
all drugs)
Best friends’
bilingualism

60

along with best friends’ bilingualism will be protective against alcohol, marijuana,
cigarette, hard drugs and all drug use.

METHODS
Data collected at baseline will be used for this analysis which will look at the
association between individual bilingualism, best friend bilingualism, and current
substance use. In looking at individual bilingualism, we cannot neglect best friends’
bilingualism since it may also predict drug use among these adolescents. As explained in
previous paragraphs, peer influence is an important factor in adolescent drug use either in
the negative or positive direction as this paper will try to examine.

Measures
Predictor Variable: Bilingualism
Data on adolescent language use was collected with four baseline survey
questions adapted from a 12-item acculturation scale shown to be a valid and reliable
measure of acculturation among Hispanics (Marin et al., 1987). Part of Marin’s
acculturation scale questionnaires are used in this study as a proxy for bilingualism.
Although several definitions have been provided for bilingualism, this particular study
will define bilingualism as the equal use of two languages in everyday life. Table 3.1 lists
the four bilingualism questions with the possible answer choices.

61

Table 3.1 Questions Used to Measure Bilingualism
Question Answer choices
1. In general, what language(s) do you most often
read and speak?

2. What language(s) do you usually speak at
home?

3. What language(s) do you usually speak with
your friends?

4. In what language(s) are the movies, TV, and
radio shows you like to watch and listen to?

1 = only English
2 = English more than another language
3 = English and another language equally
4 = another language more than English
5 = only another language (not English)

To measure bilingualism all responses with answer choice “3” for “English and another
language equally” was coded as 1 and every other response was coded as 0. All of the 1’s
and 0’s were then summed to get a variable for bilingualism which ranges from 0 to 4.
Thus, a subject whose bilingualism score is 4 will be considered as highly bilingual
because they’ve responded that they used both English and another language equally for
all four survey questions. A subject whose bilingualism score is 0 will be considered as
monolingual because they either mostly used English or another language.

Predictor Variable: Best Friend Bilingualism
Baseline surveys asked participants to list their five best friends in the class.
Students were given a roster with a list of names for every student in their class in
addition to each student’s network identification number. This number ranges from 1 to
the number of students in their class (i.e. a class of 25 students will have network
identification numbers from 1 to 25). Students were asked to list the names and network
identification number of each of their best friends. In order to obtain the peer

62

bilingualism variable, five new variables were created to represent the bilingualism score
for each of the five best friends. Best friend network identification number was then
matched to their subject identification number which was unique for each student in the
study. The subject identification number was then matched to the individual bilingualism
scores. The individual bilingualism scores where then transferred into the five new
variables for best friend bilingualism score. To obtain the best friends’ bilingualism score
variable the individual bilingualism scores of the five best friends were then summed.
The summation of the five best friends’ bilingualism score was chosen over the average
of the scores because the sum of these scores better represents the contribution of each
friend’s language use than the average. The summed variable shows a combined
influenced of bilingualism by the five best friends rather than the average influence. If
students did not list all five friends then the best friends’ bilingualism score is the sum of
the total number of friends that they had listed. The best friend bilingualism variable
ranges from 0 (i.e. none of the best friends are bilingual) to 20 (i.e. all five best friends
are highly bilingual) although the highest best friend bilingualism score for this sample is
11 (i.e. the five best friends are bilingual in varying degrees or a few best friends are
bilingual and a few are not).

Outcome Variable: Substance Use
The use of alcohol, cigarettes, marijuana, hard drugs and all drugs at baseline will
be used for this analysis. Detailed description of this variable is outlined in Chapter 1.

63

Control Variable: Acculturation
The acculturation variable for this study was established using the same
questionnaires as that of bilingualism. Unlike the bilingualism variable which measures
the frequency of answering one particular response (i.e. giving the response 3 = “English
and another language equally”), the acculturation variable was created by first reverse
coding the response so that higher response values represent higher levels of
acculturation (see table 5) and then taking the average of the responses to represent the
individual level of acculturation which ranges from 1 (i.e. not acculturated) to 5 (i.e.
highly acculturated). Correlation analysis between bilingualism and acculturation showed
that the two variables are significantly correlated (r = 0.70, p < 0.001).

Control Variable: Number of friends
Although students were asked to nominate their five best friends, not every
student nominated all five friends. Approximately 50 percent of the students nominated
five friends and an even spread nominated four, three, two or one friend. Since some
students listed more friends than others, a friend variable accounting for the total number
of friends listed was added as a control in this analysis. Higher value for the friend
variable indicates more friends being nominated and thus ranges from 0 to 5.

Covariates
Age, sex, SES, ethnicity (Hispanic) and acculturation will be used as covariates
for this study. Details have been described in Chapter 1.

64

Analysis
STATA 10.0 will be the statistical software used for this analysis. Mixed model
logistic regression approach will be used to analyze the data using school and classroom
as the random effect. We will be looking at one model with subgroup analysis. Data from
all participants will be analyzed first and then followed by analysis of male and female
subjects separately. Since it has been previously documented that males and females use
drugs for different reasons and that they interact with people differently, it would
appropriate to analyze the groups separately as well to see if results will vary according
to gender.

RESULTS
Descriptive
Approximately 880 subjects provided baseline data for this study. Demographic
characteristics have already been described in Chapter 1. Table 3.2 below depicts the
number of monolinguals versus bilinguals by ethnicity. Among monolinguals, 53 percent
were Hispanic, 20 percent Caucasian, 13 percent of mixed ethnicities and 10 percent
African American. Among bilinguals, Hispanics made up 92 percent followed by Asians
at 3 percent and subjects of mixed ethnicities at 2.4 percent. This information tells us that
the majority of the bilinguals spoke English and Spanish.

65

Table 3.2 Number of Monolinguals versus Bilinguals by Ethnicity
Ethnicity
Number (percent of row total)
Asian Af.
Am.
Hisp. Cau. N.
Am.
Mixed Other Total
Monolinguals 12
(2%)
49
(10%)
272
(53%)
105
(20%)
6
(1%)
67
(13%)
4
(0.8%)
515
(100%)
Bilinguals 11
(3%)
5
(1%)
338
(92%)
8
(2%)
1
(0.2%)
9
(2.4%)
1
(0.2%)
373
(100%)

It is important to note that a separate analysis was conducted to identify how
many of the monolingual subjects spoke mainly English and how many spoke mainly
another language. Results of this analysis showed that almost 99 percent of the
monolinguals spoke English and that about 1 percent (i.e. 9 subjects) spoke mainly
Spanish. Since it can be assumed that the 9 Hispanic subjects can speak at least a little
bit of English, they were grouped into the bilingual speakers group. Thus when we refer
to monolinguals, we are referring only to those who speak mainly English.
Table 3.3 provides a quick comparison of mean baseline drug use between
monolinguals and bilinguals. Based on the table, we can see that bilinguals seemed to use
drugs less often than monolingual speakers. Interestingly, mean marijuana and cigarette
use between monolinguals and bilinguals were significant differently. The mean use for
the all drugs variable is not significantly different although it is moving towards that
direction.
Table 3.3 Baseline Drug Use (Monolinguals vs. Bilinguals)
Monolinguals
Mean
Bilingual
Mean
Significance
Alcohol 2.43 2.256 0.27
Marijuana 2.90 2.10 0.00
Cigarette 2.64 1.96 0.00
Hard drugs 1.27 1.21 0.50
All drugs 1.68 1.49 0.06

66

Since we are looking at friends’ bilingualism and their influence on individual
drug use, it is important to examine how many monolinguals have bilingual friends and
vice versa. Table 3.4 shows the distribution of monolingual and bilingual speakers in the
friendship network. About 45 percent of monolingual speakers have best friends who are
also monolingual and 55 percent of monolinguals speakers have best friends who are
bilingual. About 39 percent of bilingual speakers have best friends who are monolinguals
and 61 percent of bilinguals have best friends who are bilinguals. There seems to be a
good distribution of monolinguals and bilinguals in the friendship network. Monolinguals
are not only best friends with other monolinguals and bilinguals are not only friends with
the bilinguals. This is pertinent to the analysis because it shows that friends’ bilingualism
can influence drug use regardless of individual bilingualism.
Table 3.4 Cross Tabulation of Best Friends Language Use
Best friends are
monolingual
Best friends are
bilingual
Total
Monolingual
speakers
232 (45%) 279 (55%) 511 (100%)
Bilingual speakers 148 (39%) 231 (61%) 379 (100%)

Mixed Model Logistic Regression
Males and Females
Results of the logistic regression analysis for males and females combined are
shown in table 3.5. Age was associated with an increase odds of alcohol use (OR = 1.15,
95% CI 1.01 – 1.30) and all drug use (OR = 1.19, 95% CI 1.03 – 1.36). Males were at a
lower odds of using hard drugs (OR = 0.70, 95% CI 0.50 – 0.98). Being of Hispanic
ethnicity was negatively associated with marijuana use (OR = 0.63, 95% CI 0.43 – 0.93).

67

Number of best friends nominated was associated with marijuana use (OR = 1.10, 95%
CI 1.00 – 1.20). That is, the more best friends you have or the larger your social networks,
the odds of marijuana use increases. Best friend bilingualism decreased the odds of
alcohol (OR = 0.92, 95% CI 0.85 – 1.00), marijuana (OR = 0.91, 95% CI 0.84 – 0.99)
and cigarette (OR = 0.90, 95% CI 0.83 – 0.98) use at baseline.
Table 3.5 Results of Logistic Regression for Both Genders, Combined (Odds Ratios)
Alcohol Marijuana Cigarettes Hard
drugs
All drugs
Males & Females (n~ 880)
Age at baseline 1.15 * 0.98 0.99 0.99 1.19 *
Gender (male) 0.77 1.14 0.78 0.70 * 0.97
SES 0.94 0.99 1.04 1.08 0.93
Ethnicity
(Hispanic)
0.89 0.63 * 0.67 1.08 0.77
Acculturation 1.02 1.24 1.01 1.06 1.19
Number of friends 1.01 1.10 * 1.05 1.07 1.04
Individual
bilingualism
1.00 0.95 0.99 1.00 1.00
Best friend
bilingualism
0.92 *

0.91 *

0.90 *

0.95 0.92
p<0.05, **p<0.01, ***p<0.001
Table 3.6 depicts results of the test for differences between mean bilingualism of
males and females indicating that bilingualism scores for males and females were
significantly different with females being more bilingual than males (i.e.
mean_bilingualism
female
= 0.868, mean_bilingualism
male
= 0.694). Best friends’
bilingualism scores of males and females were also significantly different with friends of
females being more bilingual than friends of males (mean_bf_bilingualism
female_bf
= 2.057,
mean_bf_bilingualism
male_bf
= 1.578).
Table 3.6 Test for Differences Between Male and Female Language Use
Male Female Significance
Individual Bilingualism 0.69 0.87 0.02
Best friends’ bilingualism 1.58 2.06 0.00

68

Due to the significant differences between the predictor variables among the two
genders mentioned earlier, a subgroup analysis was conducted to look at the genders
separately and is shown in table 3.7. Among males only, age was positively associated
with alcohol use (OR= 1.22, 95% CI 1.06 – 1.4) and all drug use (OR = 1.27, 95% CI
1.07 – 1.49). SES was negatively associated with all drug use (OR = 0.89, 95% CI 0.80 –
1.00). Like the combined gender analysis, being Hispanic was negatively associated with
marijuana (OR = 0.59, 95% CI 0.37 – 0.94). Number of best friends was positively
associated with cigarette use among males only (OR = 1.13, 95% CI 1.00 – 1.26).
More significant results are seen with female students. Amongst female students
only, best friend bilingualism decreased the odds of alcohol (OR = 0.87, 95% CI 0.77 –
0.98), marijuana (OR = 0.86, 95% CI 0.76 – 0.97), cigarettes (OR = 0.89, 95% CI 0.79 –
1.00) and all drug use (OR = 0.87, 95% CI 0.76 – 0.99). When looking at males and
females combined, we see a significant affect of best friends’ bilingualism on alcohol,
marijuana and cigarette use at baseline. However, when looking at the genders separately,
we see the results repeated among female students only and not among male students.
Since females make up less than half of the sample size but account for the significance
in the combined gender analysis, it can be suggested that the effects of best friends’
bilingualism on drug use in female students are very strong.

69

Table 3.7 Results of Logistic Regression for Both Genders, Separately (Odds Ratios)
Alcohol Marijuana Cigarettes Hard
drugs
All drugs
Males Only (n~520)
Age at baseline 1.22 ** 0.99 0.98 0.99 1.27 **
SES 0.92 0.95 0.98 1.04 0.89 *
Ethnicity
(Hispanic)
0.73 0.59 * 0.63 1.22 0.65
Acculturation 1.02 1.38 0.97 1.24 1.26
Number of friends 1.05 1.08 1.13 * 1.07 1.11
Individual
bilingualism
1.01 1.05 1.08 1.20 1.05
Best friend
bilingualism
0.93 0.95 0.91 0.96 0.92
Females Only (n~360)
Age at baseline 1.06 0.93 1.03 1.01 1.06
SES 0.96 1.09 1.11 1.12 1.01
Ethnicity
(Hispanic)
1.42 0.74 0.74 1.07 1.22
Acculturation 1.00 0.95 0.99 0.77 0.97
Number of friends 0.95 1.09 0.90 1.10 0.90
Individual
bilingualism
0.90 0.77 0.84 0.75 0.83
Best friend
bilingualism
0.87 *
0.77-0.98
0.86 *
0.76-0.97
0.89 *
0.79-1.00
0.91 0.87 *
0.76-0.99
p<0.05, **p<0.01, ***p<0.001

DISCUSSION
Bilingualism was hypothesized as being protective against drug use for a variety
of reasons such as those related to cognitive functions and social influence and support
via the form of social capital. In this particular analysis, we did not find evidence for
support of the cognitive effects of bilingualism on drug use but found evidence to suggest
that bilingualism is protective against use in the form of friendship influence. Female
students’ alcohol, marijuana, cigarette and all drug use was significantly associated with
their best friends’ bilingualism. That is, having bilingual best friends decreased the odds

70

of drug use among female students. This can be interpreted as the more bilingual their
best friends are, the odds of females using drugs decreases. It can also be interpreted as
the more bilingual best friends females have, the chances of them using drugs decreases.
No significantly associations between male students drug use and bilingualism scores
were found in this study. Overall, females were more bilingual than males and best
friends of females were more bilingual than best friends of males.
Why are the results for males and female students so different? A possible reason
may be that when it comes to language use, males and females are different. Studies on
gender and sociolinguistics break down gender and language use into three frameworks:
deficit, dominance and difference (Pavlenko & Piller, 2001). In the deficit framework,
women speak a “powerless language” and tend to be inferior speakers. In the dominance
framework, males dominant and females are linguistically oppressed. In the difference
framework, different genders have different linguistic attributes – women use more
prominent forms of a language and are better learners and users of languages (Pavlenko
& Piller, 2001). Although there was a small difference, a meta-analysis looking at verbal
abilities among the two genders found that females were more superior in language
performance than males (Hyde & Linn, 1988). Females tend to speak more eloquently
and use less ambiguous words than males of the same socioeconomic class. For example,
more women would say “I don’t want anything” while men would say “I don’t want
nothing” (Cheshire & Gardner-Chloros, 1998). In addition, more females are found to be
users of the minority language (Lieberson, 1970; Woolard, 1997).

71

Males and females might also differ in how they interact within a social group,
particularly around language use. The social psychological field has found girl peer
network groups to be smaller, denser and with stronger ties than that of boys. Girl groups
tend to value cohesion and homogeneity with a more intense best friendship than boys
(Woolard, 1997). It may be that since girls communicate better through language, both in
delivery and receipt of information, they exchange more social influence and support via
their network groups. While boys tend to rely on larger networks with weaker ties, girls
rely on their intimate, strong ties with other people in their networks making influence
more prominent. When bilingual girls group together, they not only share high values for
their ties but there is an added sense of relatedness from having similar language
preferences. Even among monolingual girls who have bilingual friends, the sense of
relatedness is still strong. This strong sense of togetherness, particularly among Hispanic
girls, then creates an environment that is protective. With the added effect of Hispanic’s
strong ties to family and family values, talks about deviant behavior and drug use are not
significant.

Limitations and Conclusions
Studies on the benefits of bilingualisms have shown many promising results. This
particular study found that bilingualism was more related to decrease drug use among
female students than male students. It is important to note that this sample is unique in
that they are high risk adolescents with different experiences and background than the
general population to which the previous studies on bilingualism have been conducted.

72

Many of the adolescents attending continuation high schools are those who cannot
function in the traditional high school system due to behavior problems and drug use.
Their interaction amongst their peer groups are related to their experience and antisocial
behavior. In addition, many of the studies on bilingualism have been conducted among
younger children and not too many have been conducted with Spanish speakers.
Another issue to point out is that although there are standardized tools used to
measure bilingualism such as the Language Facility Test, the Home Bilingual Usage
Estimate and the Teacher Judgment Questionnaire, this study measured bilingualism with
an acculturation scale. This scale might not be a good measure of bilingualism since it
was not intended to measure how well someone spoke two languages but rather, how
assimilated someone is to the host culture. If a correct measure of bilingualism was used
in this study, there could have been more significant findings.
The idea of bilingualism and its protective nature against drug use is fascinating.
Future studies looking at the association of bilingualism and drug use should first
measure bilingualism with an appropriate scale. Differences between male and females
should be taken into account as well as socioeconomic backgrounds, family structure and
other risk factors of drug use such as deviancy. It would also be interesting to see how the
associations vary among different levels of bilingualism (i.e. low versus highly bilingual).
This study mainly looked at Hispanic adolescent but it would be interesting to review the
same associations among bilingual Asian Americans, European Americans and African
Americans.

73

CHAPTER 4: PEER LEADERS AND THEIR ROLE IN SUBSTANCE USE

CHAPTER 4 ABSTRACT
Adolescents have a strong level of influence on each other. Peers serve as role models
and encouragement for both negative and positive behavior. Of most concern is when
adolescent imitate negative behaviors of their friends especially those they admire such as
peer leaders. OBJECTIVE: The purpose of this analysis is to examine the association
between peer leader drug use behaviors and the behaviors of the individual students.
METHOD: Mixed model logistic regression was used to analyze drug use behaviors of
approximately 525 male and female students. Students were randomized into three arms
– control, standard TND and TND networked. Within the network condition, students
were asked to nominate people in their classroom who would make good peer leaders.
Information from this survey was then used to assign students to groups of 3-5 with one
peer leader selected by the students. ANALYSIS: Peer leader drug usage behavior was
the independent variable of interest. To analyze how this association might exist in
students randomized to the control and standard TND group, peer leader nomination from
these students were used to assign students to peer led groups, post intervention.
Interaction terms between the network condition and peer leader use was created to
examine the possible moderation affect of the network condition. Analysis was conducted
for males and females combined and then separately. RESULTS: Among the combined
male and female group, there was no significant association between peer leader use and
post-test drug use. Among males only, peer leader use at baseline was positively

74

associated with alcohol use at post-test and peer leader use at post-test was positively
associated with hard drug use at post-test. Both of these associations were moderated by
the network condition. Among females only, peer leader use at post-test was negatively
associated with marijuana and cigarette use. Again, both were moderated by the network
condition. CONCLUSION: Having peer leaders in the network condition decreased the
odds of marijuana and cigarette use among female students. The opposite effect was
found in males. This lends credence to the idea that males and females interact differently
when placed in peer groups.

INTRODUCTION
Adolescence is a time of change and transition. During this time, adolescents look
to their peers for support and are strongly influenced by these peers (Berndt, 1979).
Various types of behaviors can be influenced by peers during adolescences such as using
peers as a reference group when it comes to physical activity (Anderssen & Wold, 1992),
deviant behavior (Aseltine, 1995), consumer decisions (Childers & Rao, 1992), sexual
behavior (Whitaker & Miller, 2000), educational and occupational aspirations (Duncan et
al., 1968), and body image issues (Vincent & McCAbe, 2000).

Peer Influences on Substance Use
Peers are an important influence when it comes to substance use and abuse
(Chassin et al., 1986; Kaplan et al., 1984). It seems that peer influences in the form of
peer norms and peer modeling are strongly related to adolescent alcohol use during

75

middle adolescences (mean age is 15.2 years) (Biddle et al., 1980). Early (mean age of
12.9 years) and late (mean age of 18.4 years) adolescent alcohol use are influenced by
both parents and peers. Studies looking at smoking also showed that middle adolescence
is when peer influences seemed to be strongly related to use (Chassin et al., 1981; Levitt,
1971). Latent growth modeling was used to analyze longitudinal data for substance use
among adolescents between the ages of 11-18 years and family and peer influences.
Results showed that peer and family influence was associated with initial use of alcohol,
cigarettes and marijuana (Duncan et al., 1995).
In addition, studies have shown that adolescent substance use is strongly
influenced by peers who are substance users (Hawkins et al., 1985b; Kandel, 1985;
Valente et al., 2007). A study of 465 Caucasian and Hispanic youths age 9-17 years
showed that among adolescents who were already using drugs, peer influence was
strongly associated with their use (Coombs et al., 1991). The same study also showed that
the strongest predictor of drug use among these adolescents was their friend’s level of
marijuana use. Another study of 294 young adults found that peer use was predictive of
cigarette use, binge drinking, and problem use (Andrews et al., 2002).

Peer Leaders
Like peers, peer leaders have substantial influence over adolescent behaviors.
“Popular” students, who are seen as opinion leaders or peer leaders within the school
environment, are often imitated by other students. Peer leaders display the behaviors that
are common within the school. If a high proportion of the students in the school are

76

smokers, peer leaders will most likely smoke in order to uphold their image of being part
of the school crowd.
In many instances peer leaders within the school setting are passively selected
from the student population through friendship networks and school activities such as
sports or cheerleading (i.e. school athletes and cheerleaders are the popular kids).
However in the research setting, peer leaders are identified via active student nominations.
Using the sociometric method of asking students to nominate classmates who they felt
will be good leaders Valente (2007) was able to identify students who were peer leaders.
These peer leaders were then used to assist health educators in delivering drug prevention
materials and led student groups (Valente et al., 2007).

Peer-Led Intervention Programs
Peer-led programs have been advocated as an effective means for substance abuse
prevention among adolescents. A peer-led program is defined as “an educational program
that is delivered to students by other students of comparable age or slightly older”
(Mellanby et al., 2000). These peer leaders usually share similar backgrounds such as age,
ethnicity, and socioeconomic backgrounds with they people they lead. Three different
uses of peer leaders are often seen in current interventions: (1) “peer leaders”, where
peers are leading group discussions or delivering the program curriculum; (2) “peer
counselors” where peers are used as counselors to advise other adolescents about specific
health behaviors; and (3) “peer mediators” where peers are used to bring different
individuals or groups together (e.g. delinquent student and teacher). These peer leaders,

77

peer mediators and peer counselors are used to enhance program curriculum and increase
program effectiveness.
As of the mid-90s, there were about 90 peer-led studies with over 130 different
interventions (Rooney & Murray, 1996). Interactive programs involving peer leaders or
peer involvement have been shown to be more effective than programs that did not
involve this interactive component (Black et al., 1998). The most common rationale for
using peer-led programs is that peer influences have a larger impact on adolescents’
attitudes towards substance use and abuse than adults because adolescents are especially
influenced by the expectations, attitudes, and behavior of the group to which they belong
(Linsey, 1997). Thus using peer leaders to help teach or lead a substance abuse
prevention program would most likely have the most impact on the behaviors of other
adolescents.

Peer Leader Selection
Peer-leaders are usually same age peers or older teens from nearby neighborhoods
who usually share similar backgrounds and characteristics with the subjects they lead.
Leaders may be selected through a variety of selection methods such as teacher
nominations, self-selection (e.g. volunteers), convenience sampling, or peer-nominations,
a sociometric method (Valente & Pumpuang, 2007). Peer-nomination is a sociometric
method which requires students to nominate or select one or more persons amongst their
group to serve as a leader. Data from these nominations are entered and a network matrix
is created. Persons receiving a specified threshold of nominations are then identified as a

78

peer leader. For example, Valente (2003) asked students to nominate five people in their
class who they think will be good leaders – usually with the question “Name 5 people in
this class who you think would be a good leader.” Data from these nominations were
used to assign students to peer leadership roles and later help deliver program curriculum
to prevent tobacco use. Study results showed that students in peer led groups liked the
program more than the comparison groups (i.e. control and teacher-led) and had
improved attitudes, improved self-efficacy, and decreased intention to smoke (Valente et
al., 2003).

Pros and Cons
Peer based programs where same-age peers work and learn together is effective
because each individual simultaneously influence the behaviors and attitudes of others.
Peer-led programs are simply just a type of peer-based programs which employ peers to
serve as leaders of the group. As previously discussed, adolescents are highly influenced
by the attitudes and behaviors of their peers and thus using peers to encourage healthy
behaviors such as saying “no” to alcohol and discourage negative behaviors such as drug
use is very important. Peer leaders can better relate to adolescents than many adult
counterparts. Peer leaders are usually easy to identify and assessable when done in the
school system. Since they are already participating in the study, there is no need to make
special arrangements to accommodate them as you would with a health educator and their
help is usually free of charge to the research project.

79

Some studies have found positive effects of being a peer leader. Process
evaluation of the Teens Eating for Energy and Nutrition at School Study showed that peer
leaders enjoyed being peer leaders, reported that their friends thought it was cool to be a
peer leader, and indicated that they learned more about healthy eating as a peer leader
(Story et al., 2002). Peer leaders seem to perform better in the intervention than subjects
who were not peer leaders. In Baltimore, Maryland peer leaders were trained to promote
HIV prevention among networks of injection drug users. Results showed that peer leaders
reported significant increase in the use of condoms during sexual intercourse and bleach
in cleaning their needles (Latkin, 1998). Twenty-one students selected as opinion or peer
leaders by their peers to assist in leading the Students Together Against Negative
Decisions project were also studied to test the effectiveness of being a peer leader. These
students reported significantly greater increase in AIDS Risk Behavior Knowledge,
frequency of having talks with peers about birth control and sexually transmitted diseases,
condom use self-efficacy and consistency in use (Smith et al., 2000). Allowing students,
especially high risk students, to serve as peer leaders seems to motivate them to perform
favorably. Oftentimes, these students are not perceived as being responsible and
trustworthy and thus when nominated to serve as leaders, they seem to develop a positive
identity with being a leader and start to develop pro-social and positive behaviors.
On the contrary, studies have also shown that putting high risk students into peer
groups and/or using high risk students as peer leaders can have negative effects. Dishion
and colleagues tested the hypothesis that high risk youths escalate their problem
behaviors when grouped into peer groups (Dishion et al., 1999). Defined as “the process

80

of contingent positive reactions to rule-breaking discussions”, deviancy training is an
iatrogenic affect of placing high risk or deviant youths into peer groups (Dishion et al.,
1999). Conversations between 186 boys and their friends were videotaped, analyzed, and
coded into by topics (i.e. normative vs. rule-breaking) and reaction of the listener (i.e.
laugh vs. pause). Analyses showed a linear relationship between rule-breaking talk and
affirmative behavior (Dishion et al., 1996). Longitudinal analyses revealed that the
deviancy training among these boys was predictive of delinquent behavior.
Based on what was briefly described, is it safe to put high risk youths together?
Although works from researchers like Dishion have found evidence to suggest negative
effects of peer groups, especially among boys, a few recent research studies have
provided evidence not in support of deviancy training (Mager et al., 2005; Weiss et al.,
2005).

Theoretical/Conceptual Model
Figure 4.1 depicts the hypothesized association between peer substance use and
individual substance use at post-test. Individual use at baseline, peer leader use at
baseline and post-test are considered to be associated with individual use at post-test. Age,
sex, SES and ethnicity are used as covariates in this model. The aim of chapter 4 is to
assess the association between peer leader drug use and individual drug use behaviors. It
is hypothesized that students’ drug use will be associated with the drug use behaviors of
their peer leaders. Peer leader drug use will be associated with the students drug use in

81

that leaders who increase or decrease their drug use will cause the students they led to
change their drug use accordingly.
Figure 4.1 Theoretical/Conceptual Model for Peer Leader Influence

METHODS
Baseline and post-test data will be used for the analysis.

Measures
Predictor Variable: Peer Leader Use
During the baseline survey, adolescents were asked to nominate people in their
classroom who they felt made good leaders. Information from this set of questionnaires
was used to assign students to groups with one student peer leader per group who was
nominated by others in the class. In order to access how peer leader use affect individual
use, peer leader alcohol, cigarettes, marijuana, hard drug and all drug use were matched

Age

Sex

SES

Ethnicity

Covariates
Individual drug use
at baseline
Drug use of peer
leader at baseline

Drug use of peer
leader at post-test

Individual drug use
at post-test

82

to that of the students in their group. A total of ten new variables were created to
represent peer leader use of alcohol, cigarettes, marijuana, hard drugs and all drugs at
baseline and post-test. Data on study subjects plus their corresponding peer leaders
identified via unique subject IDs and peer leader IDs were merged and cross-referenced
so that each individual had data for the ten variables that was based on their peer leader’s
corresponding alcohol, cigarette, marijuana, hard drug and all drug use at both points in
time. This means that there will be about 3-5 subjects who have the same values for all
ten variables since it is based on one particular peer leader. As expected, the ten peer
leader-related variables for subjects who served as peer leaders themselves will have the
same values as their own drug use at baseline and post-test.
As mentioned earlier, only students in the network condition were assigned into
groups led by a peer leader. Students in the control and standard TND condition were not
assigned into groups and did not have designated peer leaders. However, we are still
interested in how students in these two conditions will behave if they had peer leaders.
Thus in order to answer this question, peer leader nomination data for students in the
control and standard TND conditions were analyzed and students were assigned to
network groups with peer leaders, post intervention. Peer leader drug use data for these
students were handled in the same way as that of network students. All students were
assigned to groups with a designated peer leader regardless of condition. Students who
received the most nominations were identified as peer leaders.

83

Predictor Variable: Interaction term between Network Condition & Peer Leader Use
To access moderating effects of the network condition, interaction terms were
created between network condition and peer leader use at baseline (i.e. network * PL use
at baseline) and network condition and peer leader use at post-test (i.e. network * PL use
at post-test). These two interaction terms were included in all model runs.

Outcome Variable: Drug Use at Post-test
Alcohol, cigarettes, marijuana, hard drugs and all drug use behavior at post-test
are the outcomes of interest are this particular study. Due to non-normal distribution, all
outcome variables have been dichotomized. Detailed description of the outcome variables
has been presented in Chapter 1.

Covariates
Age, sex, SES, ethnicity (Hispanic), intervention conditions and past 30 day drug
use at baseline will be used as covariates for this study.
Analysis
All analysis for this study will be conducted in STATA version 10. Logistic
mixed model regression will be conducted since the outcome of interest is binary with
school, classroom and the addition of peer network groups as the random effects. This
three level model was used because students were grouped into peer groups and thus
there may be significant differences between these groups that must be accounted for in
the analysis.

84

RESULTS
Descriptive
Although there is complete data on roughly 880 subjects, about 1,000 subjects
were initially randomized to the three conditions. About 27 percent were randomized to
the control condition, 33 percent to the standard TND condition and 40 percent were
randomized to the network condition. Ideally 33 percent of students should have been
randomized to each of the three conditions but classroom and school infrastructure as
well as student attendance disabled us from the evenly distributing students into each
group.
Approximately 660 subjects provided complete responses to post-test surveys.
From this group, data on peer leader drug use was gathered from 550 students which
were used for this analysis. As mentioned earlier, peer leader drug use data was collected
from students in the network condition as well as generated from students in the control
and standard TND condition. Since there was no protocol to handle missing network data
on students in the control and standard curriculum group, missing network data from
these groups remained missing. Of the data that was generated, roughly 100 students
have peer leaders who do not exist in the study. That is, students nominated people at the
baseline survey who later moved out of the school, out of the class, drop out of the study
or those who did not attend the school.
Table 4.1 below shows that demographic data of the entire sample and network
condition student only since those are the actual groups that were networked. A total of
73 students in the network condition served as peer leaders and 311 were members. A

85

total of 114 students in the control and standard TND were identified as peer leaders with
about 500 members. Mean age for all groups was around 16 years of age. About 49
percent of the peer leaders were male students and about 60 percent of the member
population was male students. It is interesting to note that male students make up about
60-65 percent of the sample population but among peer leaders, they make up only 50
percent which means that more females were chosen as peer leaders (i.e. female students
make up about 40 percent of the sample population but over 50 percent of the peer
leaders). Hispanic students were proportionately selected as peer leaders (i.e. Hispanics
made up close to 70 percent of the sample population as well as 70 percent of the peer
leader group).
Table 4.1 Demographic Characteristics (All Condition vs. Network Condition)
All Conditions Network Condition Only
Peer
Leaders
N=187
Member
N=798
Peer Leaders
N=73
Member
N=311
Mean Age (years) 16.42 16.39 16.21 16.42
Male (%) 48.85 62.76 48.53 64.79
Mean Grade 10.80 10.57 10.61 10.37
Ethnicity (%)
Asian American 3.05 2.58 0.00 1.95
African American 7.93 5.88 9.38 8.20
Hispanic/Latino 69.51 68.87 68.75 67.58
White/Caucasian 9.15 13.49 10.94 14.06
Amer. Indian/Native Amer. 1.83 0.57 1.56 0.39
Mixed Ethnicities 7.93 8.03 7.81 7.03
Other 0.61 0.57 1.56 0.78

Table 4.2 shows a side-by-side comparison of drug use behavior among the all
conditions and the network condition. There are no significant differences in drug use
behavior amongst the entire group versus the network condition. Students of all the
conditions behaved similarly in terms of drug use at both baseline and post-test.

86

Table 4.2 Test for Differences between Peer Leader Substance Use
All Conditions Network Condition Only
Peer
Leaders
Mean
(Std.
Error)
Non-Peer
Leaders
Mean
(Std.
Error)
Sig. Peer
Leaders
Mean
(Std.
Error)
Non-Peer
Leaders
Mean
(Std.
Error)
Sig.
Use at Baseline
Alcohol 2.58
(0.19)
2.30
(0.08)
0.13 2.40
(0.26)
2.20
(0.12)
0.46
Marijuana 2.77
(0.24)
2.59
(0.10)
0.43 2.53
(0.33)
2.55
(0.17)
0.97
Cigarette 2.51
(0.21)
2.35
(0.10)
0.49 2.49
(0.30)
2.42
(0.17)
0.86
Hard drug 1.10
(0.02)
1.19
(0.03)
0.15 1.07
(0.02)
1.18
(0.05)
0.23
All drug 1.52
(0.05)
1.54
(0.04)
0.83 1.47
(0.07)
1.51
(0.05)
0.75
Use at post-test
Alcohol 2.52
(0.23)
2.51
(0.10)
0.96 2.29
(0.28)
2.52
(0.16)
0.49
Marijuana 2.89
(0.28)
2.74
(0.12)
0.59 2.57
(0.37)
2.61
(0.18)
0.92
Cigarette 2.28
(0.23)
2.50
(0.13)
0.44 2.73
(0.42)
2.45
(0.20)
0.53
Hard drug 1.24
(0.08)
1.27
(0.05)
0.75 1.34
(0.19)
1.19
(0.06)
0.30
All drug 1.60
(0.10)
1.64
(0.05)
0.74 1.67
(0.20)
1.56
(0.07)
0.51

Logistic Regression
Table 4.3 shows the effect size of peer leader drug use on individual drug use.
The top portion of the table looks at both males and females together. Among this group
being of Hispanic ethnicity decreased the odds of alcohol (OR = 0.56, 95% CI 0.35 –
0.90) and all drug use at post-test (OR = 0.56, 95% CI 0.33 – 0.95). As expected baseline
drug use increased the odds of all of the drug use at post-test (p <0.001). Peer leader drug
use was not associated with drug use among males and females combined.

87

The middle and lower portions of table 4.3 depict results of logistic regression in
male and female students separately. Among male students, being of Hispanic ethnicity
lowered the odds of alcohol and all drug use (OR = 0.42, 95% CI 0.23 – 0.77 and OR =
0.36, 95% CI 0.19 – 0.69, respectively). Baseline drug use was again associated with
increasing the odds of post-test drug use (p<0.001). However, among males only, the
interaction term between network condition and peer leader use of alcohol at baseline
was positively associated with individual use at post-test (OR = 1.56, 95% CI 1.06 –
2.29). In addition, the interaction term between network condition and peer leader use of
hard drugs at post-test was positively associated with individual hard drug use at post-test
(OR = 5.13, 95% CI 1.39 – 18.94). This means that among male students in the network
condition, peer leader drug use increased the odds of a group member’s hard drug use at
post-test by 5.13 times.
Among female students, the opposite results were found. Peer leader use of
marijuana at baseline was negatively associated with marijuana use at post-test (OR =
0.78, 95% CI 0.60 – 1.00). Peer leader use of cigarettes at post-test increased the odds of
cigarette use at post-test (OR = 4.02, 95% CI 1.86 – 8.67). In addition, among females
only, the interaction term between network condition and peer leader use of marijuana
and cigarettes at post-test was negatively associated with individual use at post-test (OR
= 0.66, 95% CI 0.44 – 0.99 and OR = 0.31, 95% CI 0.14 – 0.69, respectively). What does
all this mean? Peer leader use of marijuana at baseline decreased the odds of females
using marijuana use at post-test. Females were more likely to choose peer leaders who
they truly admired. Whether it was based on popularity, achievement in class or drug use,

88

females just picked leaders who had more impact on their behavior, regardless of the peer
leader’s use. Thus after being led by these peer leaders in the classroom, females tend to
react to the program’s effects. Similarly, having peer leaders who used cigarettes at post-
test increased the odds of individual female use at post-test. Again, females are more
impacted by their peer leaders. However, among female students in the network condition,
peer leader usage at post-test decreased the odds of marijuana and cigarette use. The
network condition moderated female student drug use.
In interpreting the results in table 4.3, it is important to note a few key points.
First, the present analysis is from immediate post-test which uses measures on past 30-
day drug use. This means that interpretations cannot include observations about the
effectiveness of the interventions since the interventions happened within the 30 days and
drug use inquires cover time periods before and during the interventions. Second, the
high correlations between baseline use and follow-up use indicate that there is very little
variation to be explained by the other variables.

89

Table 4.3 Results of Logistic Regression for Peer Leader Use
Alcohol Marijuana Cigarettes Hard drugs All drugs
Males & Females (n~525)
Age at baseline 1.06 0.92 0.94 0.98 1.05
Gender (male) 0.72 0.84 0.64 0.60 0.68
SES* 0.92 0.95 0.89 1.05 0.92
Ethnicity (Hispanic) 0.56 * 0.61 0.66 0.68 0.56 *
Tnd 1.55 1.47 1.43 2.71 * 1.65
Net 1.29 1.71 1.00 4.74 2.36
Use at baseline 2.30 *** 2.76 *** 1.59 *** 10.60 *** 35.08 ***
Peerleader use at
baseline
0.90 1.00 0.93 2.89 1.15
Peerleader use at post 1.08 1.08 1.24 1.20 1.09
Net * PL use at
baseline
1.16 1.02 1.18 0.43 0.82
Net * PL use at post-
test
1.06 0.96 0.90 1.01 1.06
Males Only (n~310)
Age at baseline 1.13 0.94 1.00 1.15 1.18
SES* 0.90 0.90 0.90 0.97 0.87
Ethnicity (Hispanic) 0.42 ** 0.61 0.55 0.65 0.36 **
Tnd 1.02 1.72 2.33 1.99 1.41
Net 0,83 1.94 1.36 2.39 2.52
Use at baseline 1.93 *** 2.04 *** 1.50 *** 4.00 ** 15.70 ***
Peerleader use at
baseline
0.79 1.07 0.94 3.99 1.59
Peerleader use at post 1.09 1.01 1.06 0.59 0.72
Net * PL use at
baseline
1.56* 0.92 1.18 0.26 0.67
Net * PL use at post-
test
0.98 1.11 1.00 5.13 * 1.44
Females only (n~215)
Age at baseline 0.98 0.99 0.84 0.90 0.83
SES 0.95 1.00 0.90 1.14 0.98
Ethnicity (Hispanic) 0.92 0.68 1.04 0.66 1.74
Tnd 2.27 0.75 0.50 4.36 1.22
Net 1.27 1.12 2.58 6.78 0.90
Use at baseline 4.08 *** 9.96 *** 2.08 *** 693.34 *** 648.51***
Peerleader use at
baseline
0.92 0.78 * 1.13 1.61 0.47
Peerleader use at post 1.09 1.38 4.02 *** 4.44 2.63
Net * PL use at
baseline
0.91 1.29 0.84 0.92 0.80
Net * PL use at post-
test
1.37 0.66 * 0.31 ** 0.23 1.28
* p < 0.05, ** p < 0.01, *** p <0.001

90

DISCUSSION
Peer leaders have different impacts on male and female students. In this sample,
the odds of male students using hard drugs increased with peer leader use and this
relationship is moderated by being in the network condition. Among females, the odds of
marijuana and cigarette use decreased with peer leader use and this relationship was also
moderated by the network condition. Males and females behave in different ways in peer
network groups. Males are concerned with attributes that promote their status in the
network while females are more concerned with quality of the relationship (Benenson,
1990). It may be that with hard drugs and drug use in general, males have a tendency for
deviancy and improve their status within the network by behaving mischievously. Among
high risk males, this tendency is even more pronounced. However, among females the
need to feel central in the network by misbehaving does not exist and thus their use did
not increase with peer leader use. The act of being in a network condition, working
interactively with peers and having peer leaders assist with curriculum implementation
influenced female students to decrease use.

Limitations and Conclusion
There are several limitations to this study that must be mentioned. First, this study
looks particularly at peer leader influence by asking subjects to name people in their class
who they feel are the best leaders. It does not ask who they liked the most or who they
would most likely follow or imitate. The study did ask about best friends and people in
the class who students wanted to work with in a group. However, these questions might

91

not have been able to capture the most influential students in the class and these
influential students might be the ones leading drug use behaviors. Peer leaders might
have influence on adolescent drug use but maybe the influences would be more
prominent if subjects were asked to nominate the most influential students in the class.
Another related point to mention is that students were asked to nominate people in their
classroom and not within the school. There is a chance that students have closer friends
outside the classroom or felt that there were students in another classroom that would
make a better leader. More importantly, continuation high schools have students who got
sent to the school mid year and do not necessarily know other students well. Chances do
exist that the students’ best friends or best leaders are not even in the school.
Future studies on peer leader influence and drug use should compare different
types of peer leaders (i.e. most influential in the class vs. best leader in the class vs. most
liked) to select the group of students who would have the most impact on the behaviors of
other students.

92

CHAPTER 5. CONCLUSION
Psychologist Herbert Kelman categorizes three different types of social influence.
These influences are compliance, identification and internalization (Kelman, 1958). In
compliance, Kelman argues that individuals adhere to influence because they want to be
liked and accepted. The individual does not follow the behavior because he/she truly
believes in the behavior but rather adheres to the behavior because he/she thinks that it is
what others want. Identification is when an individual adheres to influence because
he/she wants to be able to identify with the other person or establish a connection. The
individual is influenced to behave like someone who is well like and respected. In
internalization, Kelman argues that the individual adheres to the influence because
somehow the acceptance of such influences makes him/her happy and is in tune with
his/her values. Internalization is when the individual accepts the idea and/or behaviors
and makes it their own.
Social network influences in the form of peer approval, language and peer
modeling are in line with the three types of influences presented by Kelman. Adolescents
want to be compliant with the norms of their social networks, they want to identify
themselves with other adolescents who are well like (e.g. peer leaders) and they
eventually accept the influence and make it their own. Perception of friends’ approval on
drug use guided intentions which governed drug use. Females with bilingual best friends
were more likely to decrease drug use. Males were more likely to use hard drugs in
response to their peer leaders.

93

Levels of Social Influence
There are many levels of social influence. An ecological model developed by
McLeroy, Bibeau, Steckler and Glanz (1988) purposes five primary levels of influence on
health behavior – intrapersonal, interpersonal processes and primary group, institutional,
community and public policy (Sallis & Owen, 1997). This model proposes that multilevel
influences are better than single level influences and that selection of levels of influence
should be tailored to the health behavior that is being promoted (McLeroy et al., 1988). In
terms of understanding adolescent drug use, every level of influence should be used.
Teaching self- efficacy and self-esteem is important in building good intrapersonal
influences of behavior. Learning about positive group norms and using social networks
helps to enhance interpersonal influences. Within schools, drug prevention programs can
help teach adolescents about the consequences of drug abuse and within the community,
keeping a safe environment where kids can participate in pro-social activities will help to
detour them from experimenting with drugs. Policy makers can establish rules and
regulations about the location of liquor stores to make sure that it is as far from schools
and parks as possible.
For this paper, we see influence in the form of interpersonal and intrapersonal and
primary group processes. The aim of each set of analysis is to focus on how both levels of
influence are associated with drug use among adolescents. As a result three board
categories of interpersonal and intrapersonal influences are prominent - influences related
to language, gender and peer networks.

94

Gender
There is evidence to suggest that males and females interact differently. Child
development researchers have suggested that gender is crucial in shaping social networks
(Belle, 1989). Early adolescents showed changes in social network compositions. In early
teenage years, networks became more segregated and girls had larger networks than boys
(Feiring & Lewis, 2004). Girls also have more exclusive friendship than boys and usually
keep their friendship groups the same size while boys tend to expand their friendship
networks (Eder & Hallinan, 1978). Interviews of fourth and fifth grade boys and girls
found that position in the social network was associated with peer acceptance among the
boys and that boys were more concerned with their status in the networks while girls
cared more for attributes that were important to the relationships within the networks
(Benenson, 1990). As adults, women have smaller and stronger ties amongst their
networks than men and children tend to be more immersed in their mother’s networks
than their father’s (Fischer, 1982). Women’s social networks are more intimate and have
more disclosure than men’s networks which tend to involve shared activities and
experiences (Booth & Hess, 1974). When advice is needed, women seek advice from
multiple networks while men tend to seek the advice of a few key individuals such as
their wives (Antonucci & Akiyama, 1987; Veroff et al., 1981). However, what’s similar
for both men and women is that the quality of support is more important than the quantity
(Antonucci & Akiyama, 1987).
Why males and females interact differently is an important question to answer
which won’t be attempted here. However, what’s interesting is that both men and women

95

have gender roles to follow and this role might somehow influence the way they interact.
Gender roles and identities seem to be highly affected by normative cultural influences
(Schmitz & Diefenthaler, 1998). In 1974, psychologist Sandra Bem developed Bem’s
Sex Role Inventory to measure how well individuals fit into their tradition gender
specific roles such as masculine, feminine, androgynous, or undifferentiated (Bem, 1974;
Holt & Ellis, 1998). She argues that individuals attribute certain characteristics and
personality traits as either masculine or feminine. They then, intentionally and
unintentionally, behave in ways that they perceived are acceptable for their gender. For
example, a young woman might want to get involved with a conversation about football
but feels that it is a conversation for the “boys” so she ends up keeping quiet although she
has a lot to say. An adolescent boy might not want to drink beer at a friend’s party but
believes it is normal for boys to drink alcohol at parties so he follows suit. Both
individuals are following gender roles.
It is possible that, in the present study population, male and female adolescents
react differently to bilingualism and peer leaders because they are innately different.
However, it is also possible that they react to these influences differently because they are
trying to follow their gender roles. Males in peer groups increased their hard drug use
because they perceived it as the thing to do. Bilingual females tend to value family and
cultural norms more than men so they are less likely to use drugs. Adolescent girls also
have stronger networks than boys so they are highly influenced by their bilingual peers.
Although there is no data from this sample regarding delinquency, it is possible that
males in continuation high schools are more delinquent than females. Males may have

96

been sent to these schools for violent and aggressive behavior including use of illicit
drugs. Females may have been sent to the schools are problems such as teenage
pregnancy which prevented them from attending more traditional high schools.

Language
Chapter 3 of this paper focuses on language use and how it influences behavior.
Based on the results, we can see that the use of two languages serves to provide multiple
lines of influence on individual behavior. Not only does language provide us with a form
of communication, using multiple languages allows us to be immersed in multiple
cultures with core values and stigmas. Among Hispanics, drug use and drug addiction has
a very negative connotation. It is so negative that some Hispanics refuse to take
antidepressants in fear that others in their networks would perceived it in the wrong way
(Interian et al., 2007).
Acting concurrently with language are cultural norms. In Latino cultures the sense
of “familismo” and “simpatia” are very strong influences of how an individual behaves.
“Familismo” refers to a sense of obligation that an individual has to support the
immediate family and extended family which can include grandparents, aunts, uncles and
cousins (Marin & Marin, 1991). “Simpatia” is a term referring to the need to maintain
peaceful social interactions and avoid conflict (Marin & Marin, 1991; Zea et al., 1994).
An individual who is bilingual will be in social networks where these traits are highly
valued and strictly enforced. Delinquent behaviors that may disrupt peaceful interactions
are discouraged. Obligations are to the family so behaviors like drug use which are

97

deemed as possibly hurting the family are also discouraged. Bilingual adolescent females
who are immersed in these values and norms are thus less likely to participate in anti-
social behaviors.

Network Level
Lastly, network levels of influence govern how we interact with the people
around us. A study looking at 549 youths in fourth grade, seventh grade, tenth grade and
college showed that influence is a function of age. Parents provided the most support and
influence in fourth grade while same sex friends and parents are listed as providing equal
support in seventh. Among tenth grade students same sex friends are listed as providing
the most support and influence. In college, romantic partners, mothers and friends were
listed as providing the most support (Furman & Buhrmester, 1992). Like other
adolescents, the adolescents in this study population are influenced by their friends,
particularly same sex friends. Bilingual girls are highly influenced by their girl friends.
Boys in peer groups are highly influenced by their leaders. These adolescents seek
approval from their peers and as shown, peer approval has a strong impact on intentions
and behavior.
When speaking of network influences, it is a good idea to mention Latane’s Social
Impact Theory which states that the more important the group is to the individual, the
more likely the individual will conform to the group’s norms (Latane, 1981). It has
already been established that peer groups are important to adolescents, maybe more so
than parents. Both male and female adolescents want to conform to the norms of their

98

groups. Conformity is desirable especially in adolescent social networks since many do
not yet know what they truly want and are yet to establish their own sets of core values
and beliefs.

Other Contextual Factors
Gender, language and network level influences via peer modeling are some of the
social network factors that influence drug use behaviors in adolescents. However, are
there any other contextual factors which were not directly touched upon in this paper?
Yes, family structure and environmental influences such as school and neighborhoods are
all social network factors which influence drug use.
Chapter 1 has mentioned that among Hispanic adolescents family influence and
support are negatively associated to substance use (Coombs et al., 1991; Pabon, 1998;
Rodriguez & Weisburd, 1991). A study comparing the effects of a family-based drug
abuse prevention program to an adolescent group therapy program showed that the
family-based intervention was more effective and significantly reduced adolescent drug
use (Liddle et al., 2001). A study in the United Kingdom showed that adolescents who
live with both parents, hence a strong family structure, are less likely to use alcohol,
tobacco and illicit drugs (Miller, 1997). Adolescents with poor child-parent relationships
commit more delinquent acts than those with good child-parent relationships (Leung &
Lau, 1988).
Living in areas where drugs are prevalent puts adolescents at risk for becoming
drug users. Studying self-report data of 1, 416 middle school students, researchers found

99

that adolescents living in low SES neighborhoods were 5.6 times more likely to have
been offered cocaine as compared to adolescents living in high SES areas (Crum et al.,
1996). Another study looking at inner city adolescents showed that involvement in drug
or street culture activities directly effects drug involvement for African American and
Puerto Rican (Dembo et al., 1979). These examples suggest that the environment can be
an important source of influence on adolescent drug use. This is probably the reason why
nuisance businesses like liquor stores and pubs/bars are located far from schools and
playgrounds. It is also the reason why there are bans on tobacco advertisements on
television and sports venues. Unfortunately, there are a large number of adolescents who
are not shield from these bans and become influenced by what they see and experience.

Implications for Prevention Programs
Results of this paper identify social network influences as an important predictor
of substance use among high risk adolescents. In particular, social influences of peer
networks affect the intention and behaviors of adolescents in regards to drug use. Future
prevention programs among adolescents should include some form of social influence
into their curriculum. As discussed in Chapter 1, effective prevention programs provide
training for adolescents to resist social-influence to engage in substance abuse (Botvin,
1990). Effective programs are also aimed at changing normative beliefs (Gottfredson,
1997; Hansen, 1992) and utilize peer-leaders or peer group (Black et al., 1998; Rooney &
Murray, 1996). All of these characteristics are related social influence and thus, it is a
very important concept to address in future prevention programs.

100

Future programs should also pay attention to gender differences. Results of this
analysis showed that male and female adolescents behave differently from one another.
Males and females interact differently in social networks and value different
characteristics of their social networks. Males and females also communicate differently
and choice of language also varies between the two genders. Program developers should
pay attention to these differences and try to tailor their programs to meet the needs of
both males and females.
In addition, there has not been a lot of documented research in the field of
bilingualism and drug use. Acculturation and drug use have been extensively studied due
to acculturations’ linked to other negative behaviors like dietary intakes (Ayala et al.,
2008). However, the research on bilingualism and its protective role is sparse. We see in
this paper that bilingualism is protective and that females with bilingual friends are less
likely to use drugs. Although a causal model was not developed, this paper proves that
language use is linked to drug use. A longitudinal study looking at the effects of
bilingualism on adolescent behaviors such as drug use, aggression, depression, dietary
intake and social competency would be interesting to see.
In conclusion, I feel that this paper accomplished its goal of showing that social
network influences in the forms of peer approval, language and peer leaders have an
affect on adolescent drug use. In fact, it is funny that we had to conduct these analyses to
prove what we should have known all along. Don’t we already know that what we do and
how we think are the products of social influence placed upon us by friends, family and
society? It is true that we were born with genetic materials that pre-determined who we’ll

101

be but that is only the start. As we grow from infancy to adulthood, the people and things
around us shape how we behave and perceived as acceptable behaviors. Our parents,
schools, friends, work place and spouses and children all influence who we are and how
we live.

102

REFERENCES

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human
Decision Processes, 50, 179-211.

Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior.
Englewood Cliffs, NJ: Prentice Hall.

Akers, R., Krohn, M., Lanza, K., & Radosevich, M. (1979). Social learning and deviant
behavior: A specific test of a general theory. American Sociological Review, 44,
636-655.

Almodovar, A., Tomaka, J., Thompson, S., Mckinnon, S., & O'Rourke, K. (2006). Risk
and protective factors among high school students on the US/Mexico border.
American Journal of Health Behavior, 30(6), 745-752.

Amirkhanian, Y., Kelly, J., Kabakchieva, E., Kirsanova, A., Vassileva, S., Takacs, J., et
al. (2005). A randomized social network HIV prevention trial with young men
who have sex with men in Russia and Bulgaria. AIDS 19(16), 1897-1905.

Anderssen, N., & Wold, B. (1992). Parental and peer influences on leisure-time physical
activity in young adolescents. Research Quarterly for Exercise and Sport 63(4),
341-348.

Andrews, J. A., Tildesley, E., Hops, H., & Li, F. (2002). The influence of peers on young
adult substance use. Health Psychology, 21(4), 349-357.

Antonucci, T., & Akiyama, H. (1987). An examination of sex differences in social
support among older men and women. Sex Roles, 17(11-12), 737-749.

Ardila, A. (1998). Bilingualism: A neglected and chaotic area. Aphasiology, 12(2), 131-
134.

Ary, D., Tildesley, E., Hops, H., & Andrews, J. (1993). The influence of parent, sibling,
and peer modeling and attitudes on adolescent use of alcohol. Substance Use and
Misuse, 28(9), 853-880.

Aseltine, R. H. (1995). A reconsideration of parental and peer influences on adolescent
deviance. Journal of Health and Social Behavior, 36(2), 103-121.

Ayala, G., Baquero, B., & Klinger, S. (2008). A systematic review of the relationship
between accultration and diet among Lations in the United States: Implications
for future research. Journal of the American Dietetic Association, 108(8), 1330-
1344.

103

Bachman, J., Wallace, J., O'Malley, P., Johnston, L., Kurth, C., & Neighbors, H. (1991).
Racial/ethnic differences in smoking, drinking, and illicit drug use among
American high school seniors, 1976-1989. American Journal of Public Health,
81(3), 372-377.

Bagozzi, R. (1981). Attitudes, intentions, and behavior: A test of some key hypotheses.
Journal of Personality and Social Psychology, 41(4), 607-627.

Baker, J., Parasuraman, A., Grewal, D., & Voss, G. (2002). The influence of multiple
store environment cues on perceived merchandise value and patronage intentions.
The Journal of Marketing, 66(2), 120-141.

Barik, H. C., & Swain, M. (1976). A longitudinal study on bilingual and cognitive
development. International Journal of Psychology, 11(4), 251-263.

Barnes, J. (1954). Class and committees in a Norwegian Island Parish. Human Relations,
7, 39-58.

Baron, R., & Kenny, D. (1986). The moderator - mediator variable distinction in social
psychological research: conceptual, strategie and statistical considerations.
Journal of Personality and Social Psychology, 51(6), 1173-1182.

Beck, A. J., Kline, S. A., & Greenfeld, L. A. (1988). Survey of youth in custody, 1987.
Rockville, MD: National Institute of Justics.

Belle, D. (1989). Gender differences in children's social networks and supports. In D.
Belle (Ed.), Children's social networks and social supports (pp. 173-190): John
Wiley & Sons, Inc.

Bem, S. (1974). The measurement of psychological androgyny. Journal of Consulting
and Clinical Psychology, 42, 155-162.

Beman, D. (1995). Risk factors leading to adolescent substance abuse. Adolescence,
30(117), 201-208.

Ben-Zeev, S. (1977). The influence of bilingualism on cognitive strategy and cognitive
development. Child Development, 48(3), 1009-1018.

Benenson, J. (1990). Gender differences in social networks. The Journal of Early
Adolescence, 10(4), 472-495.

Berndt, T. J. (1979). Developmental changes in conformity to peers and parents. .
Development Psychology, 15, 608-616.

104

Berndt, T. J., & Keefe, K. (1995). Friends' influence on adolescents' adjustment to school.
Child Development, 66, 1312-1329.

Bialystok, E. (2001). Bilingualism in development: Language, literacy, and cognition.
Cambridge, England: Cambridge University Press.

Bialystok, E., Craik, F. I. M., Klein, R., & Viswanathan, M. (2004). Bilingualism, aging
and cognitive control: Evidence from the Simon task. Psychology and Aging, 19,
290-303.

Biddle, B. J., Bank, B. J., & Marlin, M. M. (1980). Social determinants of adolescent
drinking: what they think, what they do and what I think and do. Journal of
Studies on Alcohol, 41, 215-241.

Biglan, A., Metzler, C., Wirt, R., Ary, D., Noell, J., Ochs, L., et al. (1990). Social and
behavioral factors associated with high-risk secxual behavior among adolescents.
Journal of Behavioral Medicine, 13(3), 245-261.

Billy, J., & Udry, J. (1985). The influence of male and females best friends on adolescent
sexual behavior. Adolescence, 20(77), 21-32.

Black, D. R., Tobler, N., & Sciacca, J. P. (1998). Peer helping/involvement: an
efficacious way to meet the challenge of reducing alcohol, tobacco, and other
drug use among youth? Journal of School Health, 68(3), 87-93.

Bloomfield, L. (1933). Language. New York: Henry Holt.

Booth, A., & Hess, E. (1974). Cross-sex friendship. Journal of Marriage and Family,
36(1), 38-47.

Booth, M., Castro, F., & Anglin, D. (1990). What do we know about Hispanic substance
abuse? A review of the literature. In R. Glick & J. Moore (Eds.), Drugs in
Hispanic communities. New Brunswick, NJ: Rutgers University Press.

Botvin, E. (1990). Substance abuse prevention: Theory, practice, and effectiveness. In M.
Tonry & J. Wilson (Eds.), Drugs and Crime (pp. 461-519). Chicago: University
of Chicago Press.

Botvin, G., & Baker, E. (1993). Factors promoting cigarette smoking among black youth:
A causal modeling approach. Addictive Behaviors, 18, 397-405.

Botvin, G., Dusenberry, L., Baker, E., James-Ortiz, S., Botvin, E., & Kernet, J. (1992).
Smoking prevention among urban minority youth: assessing effects on outcome
and mediating variables. Health Psychology, 11(290-299).

105

Broidy, L., Nagin, D., Tremblay, R., Bates, J., Brame, B., Dodge, K., et al. (2003).
Developmental trajectories of childhood disruptive behaviors and adolescent
delinquency: a six-site, cross-national study. Developmental Psychology, 39(2),
372-378.

Brooks, J., Whiteman, M., Finch, S., & Cohen, P. (1996). Young adult drug use and
delinquency: childhood antecedents and adolescent mediators. Journal of
American Academy of Child & Adolescent Psychiatry, 35(12).

Buckner, J., Mallott, M., Schmidt, N., & Taylor, J. (2006). Peer influence and gender
differences in problematic cannabis use among individuals with social anxiety.
Anxiety Disorders, 20, 1087-1102.

Cairns, R., Cairns, B., Neckerman, H., Gest, S., & Gariepy, J. (1988). Social networks
and agressive behavior: peer support or peer rejection? Developmental
Psychology, 24(6), 815-823.

Carlson, J. J., Eisenmann, J. C., Pfeiffer, K. A., Jager, K. B., Sehnert, S. T., Yee, K. E., et
al. (2008). (S)Partners for Heart Health: a school-based program for enhancing
physical activity and nutrition to promote cardiovascular health in 5th grade
students. . BMC Public Health, 8(420).

Census. (2008). Current Population Survey, Annual Social and Economic Supplement
2008. internet released data sept. 2009.

Chassin, L., Presson, C. C., Bensenberg, M., Corty, E., Olshavsky, R., & Sherman, S. J.
(1981). Predicting adolescents' intentions to smoke cigarettes. Journal of Health
and Social Behavior, 22(445-455).

Chassin, L., Presson, C. C., Sherman, S. J., Montello, D., & McGrew, j. (1986). Changes
in peer and parental influence during adolescence: Longitudinal versus cross-
sectional perspectives on smoking initiation. Development Psychology, 22(3),
327-334.

Cheshire, J., & Gardner-Chloros, P. (1998). Code-switching and the sociolinguistic
gender pattern. International Journal of the Sociology of Language, 129(1), 5-34.

Childers, T. L., & Rao, A. R. (1992). The influence of familial and peer-based reference
groups on consumer decision. The Journal of Consumer Research, 19(2), 198-211.

Christakis, N., & Fowler, J. (2007). The spread of obesity in a large social network over
32 years. The New England Journal of Medicine, 357(4), 370-379.

106

Cohen, S., Gottlieb, B., & Underwood, L. (2004). Social relationships and health.
American Psychologist, 59(8), 670-853.

Coie, J., Terry, R., Zakriski, A., & Lochman, J. (1995). Early adolescent social influences
on delinquent behavior. . In J. McCord (Ed.), Coercion and punishment in
longterm perspectives. New York: Cambridge University Press.

Colzato, L. S., Bajo, M. T., Van den Wildenberg, W., Paolieri, D., Nieuwenhuis, S., La
Heij, W., et al. (2008). How does bilingualism improve executive control? A
comparion of active and reactive inhibition measures. Learning, Memory, and
Cognition, 34(2), 302-312.

Coombs, R. H., Paulson, M. J., & Richardson, M. A. (1991). Peer vs. parental influence
in substance use among Hispanic and Anglo children and adolescents. Journal of
Youth and Adolescents, 20(1), 73-88.

Costenbader, E., Astone, N., & Latkin, C. A. (2006). The dynamics of injection drug
users' personal networks and HIV risk behaviors. Addiction, 101(7), 1003-1013.

Crowe, A. H., & Bilchik, S. (1998). Drug identification and testing summary. Retrieved
February 21, 2009, from http://ojjdp.ncjrs.org/pubs/drugid/contents.html

Crum, R. M., Lillie-Blanton, M., & Anthony, J. C. (1996). Neighborhood enviroment and
opportunity to use cocaine and other drugs in late childhood and early
adolescence. Drug & Alcohol Dependence, 43(3), 155-161.

Cuijpers, P. (2002). Effective ingredients of school-based drug prevention programs. A
systematic review. Addictive Behaviors, 27(6), 1009-1023.

De La Rosa, M. (1988). Natural support systems of Hispanics: a key dimension for their
well-being. Health Social Work, 15, 181-190.

Dembo, R., Farrow, D., Schmeidler, J., & Burgos, W. (1979). Testing a causal model of
environmental influences on the early drug involvement of inner city junior high
school youths. The American Journal of Drug and Alcohol Abuse, 6(3), 313-336.

Deutsch, M., & Gerard, H. (1995). A study of normative and informational social
influences upon individual judgement. J Abnorm Soc Psychol, 51, 629-636.

Diebold, A. (1961). Incipent Bilingualism. Language, 37(1), 97-112.

Dishion, T., McCord, J., & Poulin, F. (1999). When intervention harm: peer groups and
problem behavior. American Psychologist, 54(9), 755-764.

107

Dishion, T., Spracklen, K., Andrews, D., & Patterson, G. (1996). Deviancy training in
male adolescent friendships. Behavior Therapy, 27, 373-390.

Donaldson, S., Graham, J., & Hansen, W. B. (1994). Testing the genearlizability of
intervening mechanism theories: understanding the effects of adolescent drug use
prevention interventions. Journal of Behavioral Medicine, 17(195-216).

Dryfoos, J. G. (1990). Adolescents at risk: Prevalence and prevention. New York:
Oxford University Press.

Duncan, O. D., Haller, A. O., & Portes, A. (1968). Peer influences on aspirations: a
reinterpretation. The American Journal of Sociology, 74(2), 119-137.

Duncan, T. E., Tildesley, E., Duncan, S. C., & Hops, H. (1995). The consistency of
family and peer influences on the development of substance use in adolescence.
Addiction, 90(12), 1647-1660.

Durlak, J. (1995). School-based prevention programs for children and adolescents.
Thousand Oaks: Sage.

Eckert, P. (1988). Adolescent social structure and the spread of linguistic change.
Language in Society, 17(2), 183-207.

Eder, D., & Hallinan, M. (1978). Sex differences in children's friendships. American
Sociological Review, 43, 237-250.

Elkind, D. (1978). Understanding the young adolescent. Adolescence, 13(49), 127-134.

Ellickson, P., Collins, R., & Bell, R. (1999). Adolescent use of illicit drugs other than
marijuana: How important is social bonding and for which ethnic groups?
Substance Use and Misuse, 34(3), 317-346.

Ennett, S. T., & Bauman, K. E. (1993). Peer group structure and adolescent cigarette
smoking: a social network analysis. Journal of Health and Social Behavior, 34(3),
226-236.

Ennett, S. T., Bauman, K. E., Hussong, A., Faris, R., Foshee, V., Cai, L., et al. (2006).
The peer context of adolescent subtance use: findings from social network
analysis. Journal of Research on Adolescence, 16(2), 159-186.

Epstein, J. N., Saunders, B. E., Kilpatrick, D. G., & Resnick, H. S. (1998). PTSD as a
mediator between childhood rape and alcohol use in adult women. Child Abuse
and Neglect, 22, 223-234.

108

FDA. (2004). Federal Food, Drug, and Cosmetic Act. Retrieved February 21, 2009,
from http://www.fda.gov/opacom/laws/fdcact/fdctoc.htm

Feiring, C., & Lewis, M. (2004). The transition fro middle childhood to early
adolescence: Sex differences in the social network and perceived self-competence.
Sex Roles, 24(7-8), 489-509.

Feliciano, C. (2001). The benefits of biculturalism: Exposure to immigrant culture and
dropping out of school among Asian and Latino Youths. Social Science Quarterly,
82(4), 865-879.

Feunekes, G., de Graaf, C., Meyboom, S., & van Staveren, W. (1998). Food choice and
fat intake of adolescents and adults: associations of intakes within social networks.
Prevention Medicine, 27(5), 645-656.

Fillmore, M. T., & Rush, C. R. (2002). Impaired inhibitory control of behavior in chronic
cocaine users. Drug and Alcohol Dependence, 66(3), 265-273.

Fischer, C. (1982). To dwell among friends: personal networks in town and city. Chicago:
University of Chicago Press.

Fisher, W. (2006). Predicting contraceptive behavior among university men: The role of
emotions and behavior intentions. Journal of Applied Social Psychology, 14(2),
104-123.

Fok, M. S., & Wong, D. Y. (2005). A pilot study on enhancing positive coping behavior
in early adolescents using a school-based project. Journal of Child Health Care,
9(4), 301-313.

Fratiglioni, L., Wang, H., Ericsson, K., Maytan, M., & Winblad, B. (2000). Influence of
social network on occurrence of dementia: a community-based study. The Lancet,
355(9212), 1315-1319.

Frauenglass, S., Routh, D., Pantin, H., & Mason, C. (1997). Family support decreases
influence of deviant peers on Hispanic adolescents' substance use. Journal Of
Clinical Child Psychology, 26.

Freeberg, A., & Stein, C. (1996). Felt obligation towards parents in Mexican-American
and Anglo-American young adults. Journal of Personal and Social Relationships,
13(3), 457-471.

Fritz, D. J. (2000). Adolescent smoking cessation: how effective have we been? J Pediatr
Nurs, 15(5), 299-306.

109

Fry, R., & Lowell, B. (2003). The value of bilingualism in the U.S. labor market.
Industrial and Labor Relations Review, 57(1), 128-140.

Furman, W., & Buhrmester, D. (1992). Age and sex differences in perceptions of
networks of personal relationship. Child Development, 63(1), 103-115.

Gamboa, L., Kahramanian, M., Morales, L., & Bautista, D. (2005). Acculturation and
Latino health in the United States: a review of the literature and sociopolitical
context. Annual Review of Public Health, 26, 367-397.

Gardner, M., & Steinberg, L. (2005). Peer influence on risk taking, risk preference, and
risky decision making in adolescence and adulthood: an experimental study.
Developmental Psychology, 41(4), 625-635.

Gil, A., Vega, W., & Dimas, J. (2006). Acculturative stress and personal adjustment
among Hispanic adolescent boys. Journal of Community Psychology, 22(1), 43-54.

Goodman, E., & Huang, B. (2002). Socioecononic status, depressive symptoms and
adolescent substance use. Archives of Pediatrics and Adolescent Medicine, 156,
448-453.

Gottfredson, D. (1997). School-based crime prevention. In D. Sherman, D. Gottfredson,
D. MacKenzie, J. Eck, P. Reuter & S. Bushway (Eds.), Preventing crime: What
works, what doesn't, what's promising: A report of the United States Congress.
Washington, D.C.: US Department of Justice. Office of Justice Programs.

Gottfredson, D., & Wilson, D. (2003). Characteristics of Effective School-Based
Substance Abuse Prevention. Prevention Science, 4(1), 27-38.

Gottfredson, G., & Gottfredson, D. C. (2001). What schools do to prevent delinquency
and promote safe environments. Journal of Educational and Psychological
Consultation, 12(3), 313-344.

Graham, J. W., Marks, G., & Hansen, W. B. (1991). Social influence processes affecting
adolescent substances use. Journal of Applied Psychology, 76, 291-298.

Grube, J., & Wallack, L. (1994). Television beer advertising and drinking knowledge,
beliefs, and intentions among schoolchildren. American Journal of Public Health,
84(2), 254-259.

Hansen, W. (1992). School-based substance abuse prevention: A review of the state of
the art in curriculum: 1980-1990. Health Education Research, 7, 403-430.

110

Hansen, W. B., & McNeal, R. (1997). How DARE works: an examination of program
effects on mediating variables. Health Education and Behavior, 24, 165-176.

Haugen, E. (1953). The Norwegian Language in America. Philadelphia.

Hawkins, J., Lishner, D., & Catalana, R. (1985). Childhood predictors and the prevention
of adolescent substance abuse. In C. Jonse & R. Battjes (Eds.), Etiology of drug
abuse: Implications for prevention (DHHS Publication No. ADM 85-1385).
Washington, D.C.: U.S. Government Printing Office.

Hawkins, J. D., Catalana, R., & Miller, J. Y. (1992). Risk and protective factors for
alcohol and other drug problems in adolescence and early adulthood: Implications
for substance abuse prevention. . Psychological Bulletin, 112(1), 64-105.

Hawkins, J. D., Lishner, D., Catalana, R., & Howard, M. O. (1985b). Childhood
predictors of adolescent substance abuse: Toward an empirically grounded theory.
Journal of Children in a Contemporary Society 18, 11-48.

Headen, S. W., Bauman, K. E., Deane, G. D., & Koch, G. G. (1991). Are the correlates of
cigarette smoking initiation different for black and white adolescents? Am J
Public Health, 81(7), 854-858.

Heaney, C., & Israel, B. (1997). Social networks and social support. In K. Glanz, F.
Lewis & B. Rimer (Eds.), Health Behavior and Health Education (2nd ed.). San
Francisco: Jossey-Bass Publishers.

Holt, C., & Ellis, J. (1998). Assessing the current validity of the Bem Sex-Role Inventory.
Sex Roles, 39(11-12), 929-941.

House, J. (1981). Work stress and social support. Reading, MA: Addison-Wesley.

Hyde, J., & Linn, M. (1988). Gender differences in verbal ability: A meta-analysis.
Psychological Bulletin, 104(1), 53-69.

Interian, A., Martinez, I., Guarnaccia, P., VEga, W., & Escobar, J. (2007). A qualitative
analysis of the perception of stigma among Latinos receiving antidepressants.
Psychiatric Services, 58(12), 1591-1594.

Jenkins, J., & Zunguze, S. (1998). The relationship of family structure to adolescent drug
use, peer affiliation, and perception of peer acceptance of drug use. Adolescence,
33(132), 811-8222.

Jessor, R., & Jessor, S. (1977). Problem behavior and psychosocial development: a
longitudinal study of youth. New York: Academic Press.

111

Johnson, C. A., Pentz, M. A., Weber, M. D., Dwyer, J. H., Baer, N., MacKinnon, D. P., et
al. (1990). Relative effectiveness of comprehensive community programming for
drug abuse prevention with high-risk and low-risk adolescents. Journal of
Consulting and Clinical Psychology, 58, 447-456.

Johnson, K., Whitbeck, L., & Hoyt, D. (2005). Predictors of social network composition
among homeless and runaway adolescents. Journal of Adolescence, 28(2), 231-
248.

Jorgensen, S., King, S., & Torrey, B. (1980). Dyadic and social network influences on
adolescent exposure to pregnancy risk. Journal of Marriage and Family, 42(1),
141-155.

Juvonen, J. (1996). Self-presentation tactis: promoting teach and peer approval: the
function of excuses and other clever explanations. In J. Juvonen & K. Wentzel
(Eds.), Social motivation: understanding children's school adjustment (Vol. 43-
65). New York: Cambridge University Press.

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

Kaplan, H. B., Martin, S. S., & Robbins, C. (1984). Pathways to adolescent drug use:
self-derogation, peer influence, weakening of social controls, and early substance
use. Journal of Health and Social Behavior, 25, 270-289.

Kelly, J., St. Lawrence, J., Stevenson, Y., Hauth, A., Kalichman, S., Diaz, Y., et al.
(1992). Community AIDS/HIV risk reduction: the effects of endorsements by
popular people in three cities. American Journal of Public Health, 82(11), 1483-
1489.

Kelman, H. (1958). Compliance, identification and internalization: three processes of
attitude change. Journal of Conflict Resolution, 1, 51-60.

Kilpatrick, D. G., Acierno, R., Saunders, B., Resnick, H., Best, C. L., & Schnurr, P. P.
(2000). Risk factors for adolescent substance abuse and dependence: Data from a
national survey. Journal of Consulting and Clinical Psychology, 68(1), 19-30.

Kokkevi, A., Richardson, C., Florescu, S., Kuzman, M., & Stergar, E. (2007).
Psychosocial correlates of substance use in adolescence: a cross-national study in
six European countries. Drug and Alcohol Dependence, 86, 67-74.

Komro, K., Perry, C., Williams, C., Stigler, M., Farbakhsh, K., & Veblen-Mortenson, S.
(2001). How did Project Northland reduce alcohol use among young adolescents?
Analysis of mediating variables. Health Education Research, 16, 59-70.

112

Latane, B. (1981). The psychology of social impact. American Psychologist, 36, 343-365.

Latkin, C. A. (1998). HIV prevention with drug-using populations. Public Health Reports,
113, 151-160.

Lenneberg, E. (1967). Biological Foundations of Language. New York: John Wiley &
Sons, Inc. .

Leopold, W. F. (1961). Patterning in children's language learning. In S. Saporta (Ed.),
Psycholinguistics. New York: Holt, Rinehart & Winston.

Leung, K., & Lau, S. (1988). Effects of self-concept and perceived disapproval of
delinquent behavior in school children. Journal of Youth and Adolescence, 18(4),
345-359.

Levitt, E. E. (1971). Reasons for smoking and not smoking given by school children.
Journal of School Health, 41(101-105).

Liddle, H. A., Dakof, G. A., Parker, K., Diamond, G., Barrett, K., & Tejeda, M. (2001).
Multidimensional family therapy for adolescent drug abuse: results of a
randomized clinical trial. American Journal of Drug & Alcohol Abuse, 27(4), 651-
688.

Lieberson, S. (1970). Language and ethnic relations in Canada. New York: Wiley.

Lierman, L., Young, H., Kasprzyk, D., & Benoliel, J. (1990). Predicting breast self-
examination using the theory of reasoned action. Nursing Research, 39(2), 97-101.

Linsey, B. (1997). Peer education: a view point and critique. Journal of the College of
Health, 45, 187-189.

Liu, N., Zhou, D., Li, B., Ma, Y., & Hu, X. (2006). Gender related effects of heroin abuse
on the simple reaction time task. Addictive Behaviors, 31(1), 187-190.

Lopez, M. (2002). Does bilingual education affect educational attainment and labor
market outcomes? : School of Public Affairs University of Maryland.

Luster, T., & Small, S. A. (1994). Factors associated with sexual risk-taking behaviors
among adolescents. Journal of Marriage and the Family, 56(3), 622-632.

Mackey, W. (2000). The description of bilingualism. In L. Wei (Ed.), The Bilingualism
Reader (pp. 26-54). New York: Routledge.

113

MacKinnon, D. P., Johnson, C. A., Pentz, M., Dwyer, J. H., Hansen, W. B., Flay, B., et al.
(1993). Mediating mechanisms in a school-based drug prevention program: first
year effects of the Midwestern Prevention Project. Health Psychology, 10, 164-
172.

Mager, W., Milich, R., Harris, M. J., & Howard, A. (2005). Intervention groups for
adolescents with conduct problems: Is aggregation harmful or helpful? Journal of
Abnormal Child Psychology, 33(3), 349-362.

Marcoux, B., & Shope, J. (1997). Application of the theory of planned behavior to
adolescent use and misuse of alcohol. Health Education Research, 12(3), 323-331.

Marin, G., & Marin, B. (1991). Research with Hispanic population. Newbury Park, CA:
Sage.

Marin, G., Perez - Stable, E., & Marin, B. (1989). Cigarette smoking among San
Francisco Hispanics: the role of acculturation and gender. American Journal of
Public Health, 79, 196-199.

Marin, G., Sabogal, F., Marin, B., Otero-Sabogal, R., & Perez - Stable, E. (1987).
Development of a short acculturation scale for Hispanics. Hispanic Journal of
Behavioral Sciences, 9(2), 183-205.

Markides, K. (1986). Sources of helping and inter-generational solidarity: A three
generation study of Mexican Americans. Journal of Gerontology, 41, 506-511.

Marouzeau, J. (1951). Lexique de la terminologie linguistique (3rd ed.). Paris: Geuthner.

Marshall, B. D. L., Fairbairn, N., Li, K., Wood, E., & Kerr, T. (2008). Physical violence
among a prospective cohort of injection drug users: A gender-focused approach.
Drug & Alcohol Dependence, 97, 237-246.

Mathieson, K. (1991). Predicting user intentions: comparing the technology acceptance
model with the theory of planne behavior. Information Systems Research, 2(3),
173-191.

McLeroy, K., Bibeau, D., Steckler, A., & Glanz, K. (1988). An ecological perspective on
health promotion programs. Health Education Quarterly, 15, 351-377.

Mellanby, A., Rees, J., & Tripp, J. (2000). Peer-led and adult-led school health education:
a critical review of of available comparative research. Health Education Research,
15, 533-545.

114

Merriam-Webster. (2009). "bilingualism". Retrieved December 22, 2009, from
http://www.merriam-webster.com/dictionary/bilingualism

Mertens, J. R., Lu, Y. W., Parthasarathy, S., Moore, C., & Weisner, C. M. (2003).
Medical and psychiatric conditions prevalent among alcohol and drug treatment
patients in an HMO. Archives of Internal Medicine, 163(20), 2511-2517.

Miller, P. (1997). Family structure, personality, drinking, smoking and illicit drug use: a
study of UK teenagers. Drug & Alcohol Dependence, 45(1-2), 121-129.

Min, M., Farkas, K., Minnes, S., & Singer, L. T. (2007). Impact of childhood abuse and
neglect on substance abuse and psychological distress in adulthood. Journal of
Traumatic Stress, 20(5), 833-844.

Mindel, C. (1980). Extended familism among urban Mexican Americans, Anglos, and
Blacks. Hispanic Journal of Behavioral Sciences, 2, 21-34.

Montano, D. E., Kasprzyk, D., & Taplin, S. H. (1997). The theory of reasoned action and
the theory of planned behavior. In K. Glanz, F. Marcus Lewis & B. Rimer (Eds.),
Health Behavior and Health Education (Vol. 2nd Edition). San Francisco: Jossey-
Bass Publishers.

Mooney, D. K., Fromme, K., Kivlahan, D. R., & Marlatt, G. A. (1987). Correlates of
alcohol consumption: Sex, age, and expectanies relate differentially to quantity
and frequency. Addictive Behaviors, 12, 235-240.

Moore, M., & Chase-Lansdale, P. (2001). Sexual intercourse and pregnancy among
African American girls in high-poverty neighborhoods: The role of family and
perceived community enviroment. Journal of Marriage and Family, 63(4), 1146-
1157.

Mounts, N., & Steinberg, L. (1995). An ecological analysis of peer influence on
adolescent grade point average and drug use. Developmental Psychology, 31(5),
915-922.

Mouw, T., & Xie, Y. (1999). Bilingualism and the academic achievement of first- and
second-generation Asian Americans: Accomodation with our without
assimilation? American Sociological Review, 64(2), 232-252.

Myers, R., Hyslop, T., Jennings-Dozier, K., Wolf, T., Burgh, D., Diehl, J., et al. (2000).
Intention to be tested for prostate cancer risk among African American men.
Cancer Epidemiology, Biomarkers & Prevention, 9, 1323-1328.

115

Nash, S., McQueen, A., & Bray, J. (2005). Pathways to adolescent alcohol use: family
environment, peer influence, and parental expectations. Journal of Adolescent
Health, 37, 19-28.

Neaigus, A. (1998). The network approach and interventions to prevent HIV among
injection drug users. Public Health Reports, 113(S1), 140-150.

Needle, R., Hamilton, M., Wilson, M., Reineck, R., Lazar, A., & Mederer, H. (1986).
Interpersonal influences in adolesent drug use: the role of older siblings, parents
and peers. International Journal of the Addictions, 21(7), 739-766.

Neumark-Sztainer, D., Story, M., Hannan, P. J., & Rex, J. (2003). New Moves: a school-
based obesity prevention program for adolescent girls. . Prevention Medicine,
37(1), 41-51.

Newcomb, M. D., & Bentler, P. M. (1988). Consequences of adolescent drug use: impact
on the lives of young adults. Thousand Oaks, CA: Sage Publications, Inc.

Newcomb, M. D., Chou, C., Bentler, P. M., & Huba, G. J. (1988). Cognitive motivations
for drug use among adolescents: Longitudinal tests of gender differences and
predictors of change in drug use. Journal of Counseling Psychology, 35, 426-438.

Newcomb, M. D., Maddahian, E., Skager, R., & Bentler, P. M. (1987). Substance abuse
and psychosocial risk factors among teenagers: associations with sex, age,
ethnicity, and type of school. American Journal of Drug & Alcohol Abuse, 13(4),
413-433.

Newsom. (2008). Testing Mediation with Regression Analysis: University of North
Carolina.

Novacek, J., Raskin, R., & Hogan, R. (1991). Why do adolescents use drugs? Age, sex
and user differences. Journal of Youth and Adolescents, 20(5), 475-492.

O'Callaghan, F., Callan, V., & Baglioni, A. (1999). Cigarette use by adolescents: attitude-
behavior relationships. Substance Use and Misuse, 34(3), 455-468.

Olsson, G., Nordstrom, M., Arinell, H., & von Knorring, A. (1999). Adolescent
depression: social network and family climate - a case-control study. Journal of
Child Psychology and Psychiatry, 40, 227-237.

OMB. (1997). Revisions to the standards for the classification of federal data on race and
ethnicity. Retrieved February 24, 2010, from
http://www.whitehouse.gov/omb/fedreg_1997standards/

116

Orpinas, P., Basen-Enqquist, K., Grunbaum, J., & Parcel, G. (1995). The co-morbidity of
violence-related behaviors with health-risk behaviors in a population of high
school students. Journal of Adolescent Health, 16(3), 216-225.

Ortega, A., Rosenheck, R., Alegria, M., & Desai, R. (2000). Acculturation and the
lifetime risk of psychiatric and substance abuse disorders among Hispanics.
Journal of Nervous & Mental Disease, 188(11), 728-735.

Pabon, E. (1998). Hispanic adolescents delinquency and the family: a discussion of
sociocultural influences. Adolescence, 33.

Patterson, P., & Spreng, R. (1997). Modeling the relationship between perceived value,
satisfaction and repurchase intentions in a business-to-business, services context:
an empirical examination. International Journal of Service Industry Management,
8(5), 414-434.

Pavlenko, A., & Piller, I. (2001). New directions in the study of multilingualism, second
langauge learning and gender. In A. Pavlenko, A. Blackledge, I. Piller & M.
Teutsch-Dwyer (Eds.), Multilingualism, Second Language Learning and Gender.
New York: Mouton de Gruyter.

Peal, E., & Lambert, M. (1962). The relation of bilingualism to intelligence.
Psychological Monographs, 76(546), 1-23.

Pentz, M. A. (Ed.). (1994). Target populations and interventions in prevention research:
What is high risk? Rockville, MD: U.S. Department of Health and Human
Services.

Perry, C., Williams, C., Komro, K., Veblen-Mortenson, S., Stigler, M., Munson, K., et al.
(2002). Project Northland: long-term outcomes of community action to reduce
adolescent alcohol use. Health Education Research, 17(1), 117-132.

Piaget, J., & Inhelder, B. (1969). The psychology of the child. New York: Basic Books.

Portes, A., & Schauffler, R. (1994). Language and the second generation: bilingualism
yesterday and today. International Migration Review, 28(4), 640-661.

Possel, P., Horn, A. B., Groen, G., & Hautzinger, M. (2004). School-based prevention of
depressive symptoms in adolescents: a 6-month follow-up. Journal of Amer
Academy of Child & Adolescent Psychiatry, 43(8), 1003-1010.

Rodriguez, O., & Weisburd, D. (1991). The integrated social control model and ethnicity.
Criminal Justice and Behavior, 18(4), 464-479.

117

Rogers, R. (1983). Cognitive and physiological processes in fear appeals and attitude
change: A revised theory of protection motivation. . In B. Cacioppo & L. Petty
(Eds.), Social Psychophysiology: A Sourcebook. London: Guilford.

Rogler, L., Cortes, D., & Malgady, R. (1991). Acculturation and mental health status
among Hispanics: convergence and new directions for research. American
Psychologist, 46(6), 585-597.

Rooney, B., & Murray, D. (1996). A meta-analysis of smoking prevention programs after
adjustment for errors in the unit of analysis. Health Education Quarterly, 23, 48-
64.

Saint-Jean, G., Martinez, C., & Candrall, L. (2008). Psychosocial mediators of the impact
of acculturaion on adolescent substance abuse. Journal of Immigrant and
Minority Health, 10(2), 187-195.

Sallis, J. F., & Owen, N. (1997). Ecological Models. In K. Glanz, F. Marcus Lewis & B.
Rimer (Eds.), Health Behavior and Health Education. San Francisco: Jossey-Bass
Publishers.

SAMHSA. (2009). Results from the 2008 National Survey on Drug Use and Health:
National Findings.Unpublished manuscript, Rockville, MD.

Schifter, D., & Ajzen, I. (1985). Intention, perceived control, and weight loss: An
application of the theory of planned behavior. Journal of Personality and Social
Psychology, 49(3), 843-851.

Schmitz, K., & Diefenthaler, S. (1998). An examination of traditional gender roles among
men and women in Mexico and the United States (pp. 139-143): University of
Wisconsin La Crosse.

Shedler, J., & Block, J. (1990). Adolescent drug use and psychological health. A
longitudinal inquiry. American Psychologist, 45(5), 612-630.

Sheeran, P. (2002). Intention-Behavior Relations: A conceptual and empirical review.
European Review of Social Psychology, 12, 1-36.

Sheeran, P., & Taylor, S. (1999). Predicting intentions to use condoms: A meta-analysis
and comparison of the theories of reasoned action and planned behavior. Journal
of Applied Social Psychology, 29(8), 1624-1675.

118

Smith, M. U., Dane, F. C., Archer, M. E., Devereaux, R. S., & Katner, H. P. (2000).
Students together against negative decisions (stand): Evaluation of a school-based
sexual risk reduction intervention in the rural south. AIDS Education and
Prevention, 12(1), 49-70.

Stanton-Salazar, R., & Dornbusch, S. M. (1995). Social capital and the reproduction of
inequality: Information networks among Mexican-origin high school students.
Sociology of Education, 68(2), 116-135.

Stark, R., & Bainbridge, W. (1980). Networks of faith: interpersonal bonds and
recruitment to cults and sects. The American Journal of Sociology, 85(6), 1376-
1395.

Story, M., Lytle, L., Birnbaum, A., & Perry, C. (2002). Peer-led, school-based nutrition
education for young adolescents: feasibility and process evaluation of the TEENS
study. Journal of School Health, 72(3), 121-127.

Strait, S. (1999). Drug use among Hispanic youth: Examining common and unique
contributing factors. Hispanic Journal of Behavioral Sciences, 21(1), 89-103.

Sussman, S. (1996). Development of a school-based drug abuse prevention curriculum
for high-risk youths. Journal of Psychoactive Drugs, 28, 169-182.

Sussman, S., Dent, C. W., Galaif, E. R., Stacy, A. W., Newman, T., Moss, M. A., et al.
(1997). Implementation and process evaluation of a student "school-as-
community" group. Evaluation Review, 21(1), 94-123.

Sussman, S., Dent, C. W., & Stacy, A. W. (1999). The association of current stimulant
use with demographic substance use, violence-related, social and intrapersonal
variables among high risk youth. Addictive Behaviors, 24(6), 741-748.

Sussman, S., Dent, C. W., & Stacy, A. W. (2002). Project Towards No Drug Abuse: a
review of the findings and future directions. American Journal of Health Behavior,
26(5), 354-365.

Sussman, S., Earleywine, M., Wills, T., Cody, C., Biglan, T., Dent, C., et al. (2004). The
motivation, skills, and decision-making model of "drug abuse" prevention.
Substance Use & Misuse, 39(10-12), 1972-2016.

Sussman, S., Skara, S., & Ames, S. L. (2008). Substance abuse among adolescents.
Substance Use & Misuse, 43(12-13), 1802-1828.

119

Sussman, S., Sun, P., McCuller, W. J., & 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(2), 155-162.

Taylor, S., Buunk, B., & Aspinwall, L. (1990). Social comparison, stress, and coping".
Personality and Social Psychology Bulletin, 16, 74-89.

Thompson, J., Shroff, H., Herbozo, S., Cafri, G., Rodriguez, J., & Rodriguez, M. (2007).
Relations among multiple peer influences, body dissatisfaction, eating
disturbances and self-esteem: a comparison of average weight, at risk of
overweight and overweight adolescent girls. Journal of Pediatric Psychology,
32(1), 24-29.

Tobler, N., & Stratton, H. (1997). Effectiveness of school-based drug prevention
programs: A meta-analysis of the research. Journal of Primary Prevention, 18(1),
71-128.

Triandis, H. (1977). Interpersonal Behavior. Monterey, CA: Brooks/Cole Publishing
Company.

Tudge, J., & Rogoff, B. (1999). Peer influences on cognitive development: Piagetian and
Vygotskian perspectives. In P. Lloyd & C. Fernyhough (Eds.), Lev Vygotsky:
Critical assessments. London: Taylor & Francis.

Turner, C., Ku, L., Rogers, S., Lindberg, L., Pleck, J., & Sonenstein, F. (1998).
Adolescent sexual behavior, drug use and violence: Increased reporting with
computer survey technology. Science, 280(867-873).

Urberg, K., Degirmencioglu, S., & Pilgrim, C. (1997). Close friend and group influence
on adolescent cigarette smoking and alcohol use. Developmental Psychology,
33(5), 834-844.

Valente, T. (1995). Network models of the diffusion of innovations. Cresskill, New
Jersey: Hampton Press, Inc.

Valente, T., Hoffman, B., Ritt-Olson, A., Lichtman, K., & Johnson, C. A. (2003). The
effects of a social network method for group assignment strategies on peer led
tobacco prevention programs in schools. American Journal of Public Health,
93(11), 1837-1843.

Valente, T., & Pumpuang, P. (2007). Identifing opinion leaders to promote behavior
change. Health Education and Behavior, 34(6), 881-896.

120

Valente, T., Ritt-Olson, A., Stacy, A. W., Unger, J. B., Okamoto, J., & Sussman, S.
(2007). Peer acceleration: effects of a social network tailored substance abuse
prevention program among high-risk adolescents. Addiction 102, 1804-1815.

Varlinskaya, E. I., Spear, L. P., & Spear, N. E. (2001). Acute effects of ethanol on
behavior of adolescent rats: role of social context. Alcoholism: Clinical and
Experimental Research, 25(3), 377-385.

Vega, W., Khoury, E., Gil, A. G., & Warheit, G. (1995). Cultural conflicts and problem
behaviors of Latino adolescents in home and school environments. Journal of
Community Psychology, 23, 167-179.

Veroff, J., Douvan, E., & Kulka, R. (1981). The inner American: a self-portrait from
1957-1976. New York: Basic Books.

Vincent, M. A., & McCAbe, M. P. (2000). Gnder differences among adolescents in
family and peer influences on body dissatisfaction, weight loss, and binge eating
behaviors. Journal of Youth and Adolescents, 29(2), 205-221.

Voorhees, C., Murray, D., Welk, G., Birnbaum, A., Ribisi, K., Johnson, C., et al. (2005).
The role of peer social network factors and physical activity in adolescent firls.
American Journal of Health Behavior, 29(2), 183-190.

Watson, D., & Tharp, R. (2002). Self-directed behavior: self-modification for personal
adjustment. Belmont, CA: Wadsworth.

Webb, J., Baer, P., & McKelvey, R. (1995). Development of a risk profile for intentions
to use alcohol among fifth and sixth graders. Journal of the American Academy of
Child & Adolescent Psychiatry, 34(6), 772-778.

Weiss, B., Caron, A., Ball, S., Tapp, J., Johnson, M., & Weisz, J. R. (2005). Iatrogenic
effects of group treatment for antisocial youth. Journal of Consulting and Clinical
Psychology, 73(6), 1036-1044.

Whitaker, D. J., & Miller, K. S. (2000). Parent-adolescent discussions about sex and
condoms. Journal of Adolescent Research, 15(2), 251-273.

White, D., & Pitts, M. (1998). Educating young people about drugs: a systematic review.
Addiction, 93, 1475-1487.

Wilson, D., Gottfredson, D., & Najaka, S. (2001). School-based prevention of problem
behaviors: A meta-analysis. Journal of Quantative Criminology, 17(3), 247-272.

121

Woolard, K. (1997). Between friends: Gender, peer group structure, and bilingualism in
urban Catalonia. Language in Society, 26(4), 533-560.

Zea, M., Quezada, T., & Belgrave, F. (1994). Latino cultural values: their role in
adjustment to disability. Journal of Social Behavior and Personality, 9.

Zogg, J., Ma, H., Dent, C. W., & Stacy, A. W. (2004). Self-generated alcohol outcomes
in 8th and 10th graders: Exposures to vicarious sources of alcohol information.
Addictive Behaviors, 29(1), 3-16.

Abstract (if available)

Abstract Social network influences predict how we act and behave. Most people tend to comply with social norms and beliefs so that they are accepted by society. Characterized by physical, emotional and mental changes, adolescents are no different when it comes to social conformity. Peer influences become a strong predictor of how adolescents act and behave. Many of these individuals behave in ways that are encouraged and acceptable by peers despite the delinquent nature of the behavior. The purpose of this dissertation is to understand adolescent social influence and contextualized social networks as it relates to drug use.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

Social network influences on depressive symptoms among Chinese adolescents

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

Investigating factors that influence peer relationships and obesity during middle childhood

PDF

Adolescent social networks, smoking, and loneliness

PDF

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

PDF

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

PDF

Subclinical atherosclerosis in overweight Latino youth: influence of cardiometabolic risk factors

PDF

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

PDF

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

PDF

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

PDF

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

PDF

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

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

Post-traumatic growth among high-risk youth: predictors, impact of stressful life events, and relationship with changes in substance use behaviors

PDF

The egocentric social networks of Black adolescents and sources of racial socialization messages

PDF

Association of traffic-related pollution and stress on childhood lung function

PDF

The vicious cycle of inactivity, obesity, and metabolic health consequences in at-risk pediatric populations

PDF

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

Asset Metadata

Creator Kwan, Patchareeya Pumpuang (author)

Core Title Contextualizing social network influences on substance use among high risk adolescents

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Preventive Medicine (Health Behavior)

Publication Date 04/28/2010

Defense Date 03/08/2010

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

Tag Adolescents,bilingualism,continuation high schools,drug use,high risk adolescents,OAI-PMH Harvest,peer influences,peer leaders,social networks,substance use

Place Name California (states)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Valente, Thomas W. (committee chair), Rice, Eric (committee member), Riggs, Nathaniel (committee member), Sun, Ping (committee member), Sussman, Steven (committee member)

Creator Email patchareeya@gmail.com,pumpuang@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2972

Unique identifier UC1227121

Identifier etd-Pumpuang-3431 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-311765 (legacy record id),usctheses-m2972 (legacy record id)

Legacy Identifier etd-Pumpuang-3431.pdf

Dmrecord 311765

Document Type Dissertation

Rights Kwan, Patchareeya Pumpuang

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