t. Pu. "€33

I;..oqu.:

”an {a

Iﬁnr

4;:

 

 

THESE; I

 

 

 

251(1) O 7
This is to certify that the
dissertation entitled
TOWARD AN ECO-DEVELOPMENTAL THEORY OF
ADOLESCENT SUBSTANCE USE IN VENEZUELA
.03

>_ «(P- > presented by

0: CD “5:;

< C 5

0‘.‘ C5 ’>

33 _Q ': RONALD B. COX, JR.

:1 5 i
l 2

—~———— J has been accepted towards fulﬁllment

of the requirements for the

Ph.D. degree in Familx and Child Ecology

 

 

Major Professor’s Signature
q ’ LZ‘ 200‘}
I

Date

MSU is an aﬁinnative-action, equal-opportunity employer

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATEDUE

DAIEDUE

DAIEDUE

 

-\ I'l I
'Fﬂrﬂi‘ 'r (N,

SEiPZOQt‘étmsI

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p‘lCIRC/DaleDue indd-p.1

TOWARD AN ECO-DEVELOPMENTAL THEORY OF ADOLESCENT
SUBSTANCE USE IN VENEZUELA

By

Ronald B. Cox, JR.

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Family and Child Ecology

2007

ABSTRACT

TOWARD AN ECODEVELOPMENTAL THEORY OF ADOLESCENT SUBSTANCE
USE IN VENEZUELA

By
Ronald B. Cox, JR.

This study surveyed school-attending adolescents in Caracas, Venezuela in order
to explore the prevalence of substance use and to identify risk and protective factors that
inﬂuence the age of ﬁrst drug use among this population. The theoretical premise of this
study was that the synergistic effects of the Venezuelan adolescent’s social ecology offers
a foundation from which to develop intervention strategies for the prevention and
treatment of adolescent substance use. Studies of risk and protective factors related to
adolescent substance use have identiﬁed several targets for intervention within families
and communities in the United States. These studies have led to the formation of family
therapy and prevention treatments, many of which have become standard evidence-based
interventions for youth involved with licit and illicit substances. Even though these
intervention strategies exist for US. populations, questions remain unanswered regarding
whether the same risk and protective factors related to adolescent substance use are
operative in the Venezuelan culture, and if so, how these interventions should be adapted
to be effective with this population. The purpose of this study was to assess for the extent
to which known risk and protective factors for US. populations exist among school-
attending youth in Caracas, Venezuela.

A sample of 1,831 Venezuelan youth attending 14 schools in two school districts

located on the western side of Caracas, Venezuela participated in the study. Descriptive

analyses provided prevalence rates of ﬁrst use of eight different drugs (cigarettes,
alcohol, inhalants, ecstasy, crack cocaine, heroin, non prescribed pharmaceuticals, and
marijuana) for the sample. Findings indicated that high percentages of youth (81.5%) had
consumed alcohol, and approximately one third had used cigarettes. Incidence of all other
drugs in the study had very low reported rates ranging from .3% (crack) to 3.7%
(marijuana).

Hierarchical Linear Modeling was used to test the relationships between six
variables that have been shown to covary with age of ﬁrst drug use in the US. (family
attention, externalizing behavior, peer drug inﬂuence, school climate, gender, and
socioeconomic status). Only family attention, externalizing behavior, and gender were
supported as level-1 covariates, while mean socioeconomic status was found to be a
signiﬁcant level-2 covariate of age of ﬁrst drug use. Implications for family therapy

treatment and prevention professionals are provided as well as areas for future research.

Copyright by
RONALD BLAKE COX, JR.
2007

Esta obra se dedica a:
Norelis: esposa, madre, amante, amiga, compar‘iera, e inspiracién, doy gracias a Dios
portL
Y a mis hijos, Caleb, Amber, Scarlett, y Viviana
quienes han soportado todo sin queja alguna; son mi riqueza y mi gozo.

Lo logramos!

ACKNOWLEDGEMENTS

This dissertation was the result of the synergistic interpersonal connections that
have spanned a lifetime, and thus the work of far too numerous a cadre of inﬂuential
contributors to mention within the conﬁnes of this space. Still, from a sense of deep
gratitude, I want to thank the individuals most recently and directly involved in this
effort.

To Dr. David Imig, who offered me hours of stimulating discussion that shaped
my understanding of ecological theory. Throughout my time at MSU you guided me with
your sage advice and continuously pointed out new ways of looking at the same old
problems. You’ve been a mentor and a friend. Thank you.

To Dr. Kim Maier, your class was my ﬁrst experience with statistics and you
awoke in me a desire to know more. Even more, however, I have been inspired by your
faith and touched by your kindness. You have been a great example to follow both
professionally and personally. Thank you.

To Dr. Ruben Parra, your unique ability to be incessantly afﬁrming of me as a
scholar while relentlessly challenging me to grow, combined with your keen insight have
added an important dimension to my scholarship, and more importantly, to my personal
development. Thank you.

To Dr. Adrian Blow my committee chair, who believed in my potential as a
scholar and invested enormous amounts of time and energy in my development; your

guidance and encouragement has been like a deep well, a truly invaluable resource. You

vi

went the extra mile with me and I couldn’t have asked for a more competent chair for my
dissertation. Thank you.

I am also indebted to all the support staff of F CE. Mary, Ruth, and Sarra all went
out of their way to make sure that I had access to every resource that could be mustered,
and sometimes even to the extent of personal sacriﬁce. You all provided a sense of
importance and belonging to my stay at MSU. Thank you.

To Mariela Rodriguez who helped me collect data and who went above and
beyond the call of duty; you are a friend like no other. Thank you.

To Ken Slater, my Pastor and friend from the Greater Lansing Church of Christ
who read drafts and gave excellent feedback; you have been an instrument of God in my
life. Thank you.

To my parents, Vivian and Ron, Sr., who instilled in me, through their words and
example, the work ethic and the sense of service essential to the completion of this
degree. I know that you are proud of me. But what I may not say frequently enough is
that I, too, am proud of you. Thank you.

To Al Loftis, who has been a brother and a friend; your help and generosity made
it possible for me to dedicate time to ﬁnishing this degree. Thank you.

To my children, Caleb, Amber, Scarlett, and Viviana, who gave up “Daddy time,”
and who worked tirelessly with me as a team, each doing their part toward the
accomplishment of the task, I couldn’t be more proud of each of you! Thank you.

To my wife Norelis, who saw the potential in me and believed in me long before I

even believed in myself; it was your continued support and personal sacriﬁce that has

vii

kept me going through the tough times. I admire you and love you more than you will
ever know. Thank you.

And ﬁnally, to my Lord and Savior Jesus Christ; you brought me out of the
darkness and into the light. You have given my life purpose and meaning. You have
sustained me and blessed me more than I had ever hoped. I cannot imagine where I
would be if it were not for you, for without you none of this would have been possible.
Thank you!

Each you in your own way has contributed to this work and thus it is the work of
all of us. To you and to those not mentioned in these short lines, thank you for
accompanying me on this journey. With every end, there is a new beginning. So, as I end

one phase, I begin another; one that all of you helped create. Thank you!

viii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................... xi
LIST OF FIGURES ........................................................................................................... xii
CHAPTER 1: OVERVIEW
Introduction ......................................................................................................................... 1
Adolescent Substance Use in the US. ........................................................................ 2
Treatment and Prevention Models .............................................................................. 3
The Eco-Developmental Model .................................................................................. 5
Venezuela as a Context for Adolescent Substance Use .............................................. 6
Statement of the Problem .................................................................................................... 7
Purpose of the Study ........................................................................................................... 8
Signiﬁcance of the Study .................................................................................................... 9
Theory Development .......................................................................................................... 9
Research Questions, Hypotheses, and Measures .............................................................. 12
Speciﬁc Questions and Hypotheses .......................................................................... 13
Measures ................................................................................................................... 14
Conceptual and Operational Deﬁnitions ................................................................... 15

CHAPTER II: BACKGROUND

Review of the Literature ................................................................................................... 19
Individual Inﬂuences on the Development of Substance Use in the US ................. 19
The Role of Genetics ................................................................................................ 22
Microsystem Inﬂuence on the Development of Substance Use in the US. ............. 24
Mesosystem Inﬂuence on the Development of Substance Use in the US. .............. 31
Exosystem Inﬂuence on the Development of Substance Abuse in the US .............. 33
Macrosystem Inﬂuences on the Development of Substance Use: The Case for
International Research .............................................................................................. 35

Venezuela as a Context ..................................................................................................... 39
The Qualitative Case of Venezuela ........................................................................... 40
The Quantitative Case for Venezuela ....................................................................... 48

Conclusion ........................................................................................................................ 54

CHAPTER III: METHODOLOGY

Setting ............................................................................................................................... 56
Methods ............................................................................................................................. 57
Sampling Procedures ................................................................................................ 57
Data Collection ......................................................................................................... 58
Data Processing and Quality Control ........................................................................ 63
Human Subjects Protections ..................................................................................... 63
Study Participants ..................................................................................................... 64
Measures ................................................................................................................... 66

ix

Data Analytic Plan ............................................................................................................ 92
Modeling Approaches ............................................................................................... 92
Research Questions and Hypotheses ........................................................................ 93

CHAPTER IV: RESULTS

Descriptive Statistics ......................................................................................................... 95
Preliminary Analyses ........................................................................................................ 99
Age of First Drug Use ............................................................................................... 99
Gender ..................................................................................................................... 100
Race ......................................................................................................................... 100
SES, Family Attention, Extemalizing Behavior, Peer Drug, and School Climate. 100
Missing Data ........................................................................................................... 103
Multilevel Models ........................................................................................................... 104
Age of First Drug Use: The Null Model ................................................................. 104
Age of ﬁrst drug use — Model-1 .............................................................................. 106
Age of ﬁrst drug use - Model-2 .............................................................................. 108
Age of ﬁrst drug use — Model-3 .............................................................................. 109
Age of ﬁrst drug use — Model-4 .............................................................................. 110
Age of ﬁrst drug use — Model-5 .............................................................................. 111
Age of ﬁrst drug use — Model-6 .............................................................................. 116
Age of ﬁrst drug use - Model-7 .............................................................................. 121
Age of First Drug Use — Model Diagnostics .......................................................... 135
CHAPTER V: DISCUSSION
Discussion of Results ...................................................................................................... 146
General Description of Drug Use ........................................................................... 146
Independent Variables ............................................................................................ 149
Discussion of Methodologies: Limitations ..................................................................... 162
Survey Population ................................................................................................... 162
Measurement Strategies: Use of Self-Reported Data ............................................. 162
Model Speciﬁcation and Temporal Sequencing ..................................................... 165
Discussion of Methodologies: Strengths ........................................................................ 166
Implications for Treatment and Prevention .................................................................... 168
Implications for Further Research .................................................................................. 172
Concluding Remarks ....................................................................................................... 175
APPENDICES
Appendix A: PACARDO-V 2007 (Spanish) ................................................... 177
Appendix B: PACARDO- V 2007 (English) ................................................... 186
Appendix C: MAMBI (Spanish & English). .. .195
Appendix D: IRB Approval Letter and Consent Forms ...................................... 200
Appendix E: Authorization Letters .............................................................. 217
REFERENCES ..................................................................................... 223

LIST OF TABLES

Table 3.1 Grade level in School ............................................................... 66
Table 3.2 Reliability Estimates of Psycho-Social Constructs in the PACARDO ....... 68
Table 3.3 Age of First Drug Use ............................................................... 72
Table 3.4 Descriptive Statistics of Family Attention ....................................... 78
Table 3.5 CF A Model Results .................................................................. 79
Table 3.6 Descriptive Statistics of PRDG .................................................... 81
Table 3.7 Descriptive Statistics of EXTB .................................................... 82
Table 3.8 CF A Results for EXTB ............................................................ 84
Table 3.9 Descriptive Statistics for SCLM ................................................... 85
Table 3.10 CFA Results for SCLM ........................................................... 87
Table 3.1 1 Descriptive Statistics for SES .................................................... 88
Table 3.12 Descriptive Statistics for MAMBI ............................................... 91
Table 4.1 Demographics of Sample ........................................................... 96
Table 4.2 Frequencies of Students Reporting Having Initiated Drug Use ............... 98
Table 4.3 Number of Different Drugs Consumed ........................................... 98
Table 4.4 Ages of First Initiation of Substance Use ........................................ 99
Table 4.5 Regression: Age First Use ......................................................... 101
Table 4.6 Correlations Level-1 Covariates ................................................... 102
Table 4.7 Corrleations Level-2 Covariates ................................................... 102
Table 4.8 Estimates for Models ............................................................... 113
Table 4.9 Within school sample size ......................................................... 141

xi

LIST OF FIGURES

Figure 3.1 Histogram of Age of First Use .................................................... 66
Figure 3.2 Histogram of F AM Frequency .................................................... 78
Figure 3.3 Histogram of PDRG Frequency ................................................... 81
Figure 3.4 Histogram of EXTB Frequency ................................................... 83
Figure 3.5 Histogram of SCLM Frequency .................................................. 86
Figure 3.6 Histogram of SES Frequency ..................................................... 89
Figure 3.7. Histogram ofthe MAMBI ........................................................ 91
Figure 4.1. FAM random slopes by AF U identiﬁed by MSES ........................... 119

Figure 4.2. The joint relationship of EXTB and FEM on AFU holding FAM
constant ........................................................................................... 124

Figure 4.3. The joint relationship of FAM and FEM on AF U holding EXTB

constant ........................................................................................... 126
Figure 4.4. The joint relationship of FAM and EXTB on AF U by gender .............. 128
Figure 4.5. The cross-level interaction effects of EXTB by MSES on AFU holding
FAM constant .................................................................................... 130
Figure 4.6. The cross-level interaction effects of F AM by MSES on AFU holding
EXTB constant ................................................................................... 1.32
Figure 4.7. The cross-level interaction effects of FEM by MSES on AF U holding

F AM constant .................................................................................... 134
Figure 4.8. Box plot of Level-1 residuals by each of the 14 schools ..................... 137
Figure 4.9. Scatterplot of level-1 residuals against the ﬁtted values ..................... 139
Figure 4.10. P-P plot of the level-1 residuals ................................................ 140
Figure 4.11. Normal Q-Q plot of Mahalanobis’ Distance ................................. 142

Figure 4.12 Mahalanobis’ Distance by the expected values of the order statistics. . . .. 143

xii

Figure 4.13. Empirical Bayes intercept estimate by MSES ................................ 145

xiii

CHAPTER 1: OVERVIEW
Introduction

Illicit drug use and the abuse of legal substances is a prominent concern for public
health ofﬁcials throughout the world (Corroa, Guindon, & Sharma, 2000; WHO, 2004;
WHO, 1997). It is reported that every year tobacco use alone is responsible for
approximately four million deaths worldwide, and alcohol abuse is even more costly to
human life and productivity (WHO, 2004). Research is emerging that points to the long-
term negative consequences of chronic marijuana use on selective cognitive functioning
and on negative effects for respiratory functioning similar to those caused by sustained
tobacco use (WHO, 1997). Goldman, Oroszi and Ducci (2005) in a review of the
literature on addiction research report that worldwide there are 2 billion alcohol users, 1.3
billion tobacco users, and 185 million users of illicit drugs, and that these three categories
account for 12.4% of the global deaths in 2001. In the US. alone, these authors report
that addictive drugs are the cause of approximately 590,000 deaths, and are responsible
for injury or illness to almost 40 million individuals every year. For purposes of this
study, use of any substance, licit (i.e., alcohol, tobacco) and illicit (e.g., cannabis,
cocaine, heroin, etc.), by an adolescent will be referred to as substance use or abuse
unless otherwise speciﬁed.

Adolescent substance use is of particular concern because early initiation predicts
later misuse (Spoth, Guyll, & Day, 2002). For example, if the current trend among
adolescent tobacco use were to continue, it is predicted that 250 million children living

today will die of tobacco-related causes (Warren, Riley, Asma, Eriksen, Green, et al.,

2000). Alcohol use among youth is associated with signiﬁcant increases in
suicides, motor vehicles accidents, and drownings (WHO, 2004). Cannabis use among

youth is also linked to increases in motor vehicles accidents (WHO, 1997).

Adolescent Substance Use in the US.

In the US, Johnston, O’Malley and Bachman (2001) report that 7.4% of 8th
grade students and 20.6% of 12th grade students smoked cigarettes daily, 14.1% of 8th
grade students and 30.0% of 12th grade students engaged in binge drinking (deﬁned by
having 5 or more drinks on a single occasion in the past two weeks), and 19.5% of 8th
grade students and 40.9% of 12th grade students used illegal drugs in the past year.

Problem behaviors including substance abuse among youth in middle school
frustrate learning and increase susceptibility to antisocial inﬂuences. This in turn places
them at risk for school failure, involvement in the criminal justice system, and health
problems (Dryfoos, 1990; Jessor, & Jessor, 1977; Simons-Morton, Crump, Saylor, & Yu,
1999). Evidence suggests that the transition into middle school is a particularly critical
time for youth. Prior to middle school (i.e., preadolescents in elementary school) behavior
problems are uncommon, but show considerable increase during adolescence (Johnston,
O’Malley, & Bachman, 1994). For example, these authors report that less than 10% of
sixth graders (approximately 11 to 12 years of age) have used tobacco compared to about
30% of eighth and 60% of 11 graders. Similarly, only 5% of sixth graders have used
alcohol, compared to almost 70% of eighth graders. Unfortunately, for some children
early adolescence is the beginning of a downward spiral from which they never recover

(Eccles, Lord, & Midgley, 1991).

Treatment and Prevention Models

Treatment and prevention models have been developed in order to interrupt the
downward spiral in which many youth ﬁnd themselves. Both treatment and prevention
sciences are built on the idea that there are empirically identiﬁable patterns of behavior or
contexts that serve as risk or protective factors (Hawkins, VanHorn, & Arthur, 2004) in
the development of adolescent substance abuse. A risk factor is something that increases
the chance that substance-abusing behavior will occur. A protective factor acts as
moderator, or mediator buffering or reducing the effect of risk exposure and is, therefore,
more than simply the opposite of a risk factor. (Hawkins, Catalano, & Miller, 1992).
Studies have identiﬁed mental health (Clark & Winters, 2002; Colby, Lee, Lewis-
Esquerre, Esposito-Smythers, & Monti, 2004; Swadi, 1999), parental and family
relationships (Chassin, Ritter, Trim, & King, 2003; Stanton & Todd, 1982), peer
relationships (Bauman & Ennett, 1994; Hussong 2002), school bonding (Hill & Werner,
2006; Murguia, Zeng-yin, & Kaplan, 1998), religion (Chen, Dormitzer, Bejarano, &
Anthony, 2004), and neighborhood environment (Duncan, Duncan, & Strycker, 2002) to
be important factors in the development of substance use among adolescents. Treatment
and prevention interventions attempt to reduce speciﬁc risk factors and increase
protective factors in an effort to sway the developmental trajectory of the adolescent
toward health. However, in order for these strategies to be efﬁcient they must be based on
a foundation of empirical research that is conducted within a cultural context (Castro,
Barrera, Martinez, 2004; Hecht, Marsiglia, Elek, Wagstaff, Kulis, et al., 2003). This study

tested how four of the most consistently identiﬁed risk and protective factors in the

literature inﬂuenced the onset of adolescent substance abuse in Venezuela in a sample of
school-attending youth.

The bulk of the literature emphasizes two aspects of the parental relationship as
predictive of substance abuse onset among adolescents: Parental warmth or
supportiveness and parental monitoring (Barnes, Reifman, Farrell, & Dintcheff, 2000).
Following the example of Anthony and colleagues (Dormitzer, Gonzalez, Penna,
Bejarano, Obando, et al. 2004) in an international study of risk factors for school-
attending adolescents in Central America, this study will combine these two dimensions
(i.e., parental supportiveness and parental monitoring) into a composite variable called
family attention.

Externalizing behavior is a mental health construct that refers to a grouping of
behavior problems manifested in children’s outward behavior and that depict the child
negatively acting on her or his external environment (Eisenberg, Cumberland, Spinrad,

F abes, Shepard, et al. 2001), externalizing behavior has been consistently linked to
substance use among teens (e.g., Kaplow, Curran, & Dodge, 2002; Schuckit, et al., 2003).

Peer relationships are a robust predictor of adolescent substance abuse in the
literature (Bauman & Ennett, 1994) with youth who are more embedded in peer contexts
with delinquent youth being more likely to use substances themselves (Hussong, 2002).

Additionally, research has found that school climate, comprised of a combination
of a positive afﬁliation toward school and characteristics of the environment in which the
school is nested, may serve a protective function against many antisocial behaviors

(Ennet, Flewelling, Lindrooth, Norton, 1997; Hill & Werner, 2006).

The Eco-Developmental Model

Ecological theory as set forth in Bronfenbrenner’s Eco-developmental model
(Bronfenbrenner, 1979; Bronfenbrenner & Ceci, 1994) offers a useful framework to
examine how risk and protective factors interact to inﬂuence adolescent development and
drug use in differing cultures. Bronfenbrenner posits that an individual interacts with
different contexts to form and guide development, and that these contexts are nested
within four layers or systems of inﬂuence in which the individual lives. These four layers
are the microsystem, mesosystem, exosystem, and macrosystem. The layers evolve in
increasing levels of abstraction from direct inﬂuences to more indirect inﬂuences on the
developing individual. The microsystem is comprised of elements in the individual’s
immediate environment such as family, peers, and school. The mesosystem refers to how
these microsystems interact to inﬂuence the individual’s development. The exosystem
refers to systems that exert their inﬂuence on the individual indirectly through the
microsystem (e. g. aparent’s work inﬂuences the parent who inﬂuence the child; a
teacher’s relationship with school administrators affects the teacher’s interaction with
child). The macrosystem is the most abstracted of the systems and refers to inﬂuences
such as cultural values, national economics, and policies.

Environments are meaningful not only for what they actually contain, but for the
meaning that is created within them. For human ecologists, environments are
“subjectively experienced. . .. [People] perceive, interpret, and create their meaning”

(Bubolz & Sontag, 1993, p. 23). Social contexts, therefore, have a wide-ranging inﬂuence

on an individual’s decision to engage in substance use. From this perspective,
environments are not determinants of human behavior but create constraints as well as
opportunities (Bubolz & Sontag, 1993). Development is not something that just happens
to children. Rather, they are active participants in the contexts in which development
occurs. People can respond, change, act on, and modify their environment, and thus, in
this sense, contribute to their own development (Bronfenbrenner, 1995).

From an ecological perspective, risk and protective factors for adolescent
substance abuse are the result of the interaction of an individual with his or her context.
Therefore, substance abuse can be deﬁned as the “phenotypic expression of the
interaction of a genetic predisposition(s) (genotype) to substance abuse, certain personal
or environmental risk factors, and the psychopharmacological effects of the drugs
themselves” (Brook, Brook, & Pahl, 2006, p. 39). While psychiatric treatment that
includes a pharmacologic regimen might be used to treat the underlying
pathophysiological predispositions and comorbid psychiatric disorders present,
prevention and treatment models intervene to change environmental risk and protective
factors, as well as behavioral effects of the drugs themselves (e.g., craving, relapse
prevention, etc.). In order to maximize the effectiveness of these models, the
interventions should be adapted to the individual and his/her speciﬁc context, which
includes a careful consideration of the cultural variations that exist within psychosocial

domains.

Venezuela as a Context for Adolescent Substance Use

This study took place in the country of Venezuela. Subjective reports of substance
abuse among youth in Venezuela are alarming. Some studies have begun to shed light on
the prevalence of adolescent substance abuse in the Spanish Speaking countries of the
Americas (e.g., Dormitzer, et al., 2004). However, between country variability in
prevalence rates and the variance in adolescent substance use explained by risk and
protective factors precludes assumptions of homogeneity based on a common language
and cultural heritage. Different national histories, governmental policies, economics,
geographic locations, and the like are macrosystemic effects that inﬂuence the
exosystems, mesosystems, and microsystems that comprise the proximal processes that in
turn interact with a genotype to determine the developmental trajectory of the individual.
Venezuela constitutes a distinct context that warrants careful consideration in order to

culturally adapt or develop effective prevention and treatment models.

Statement of the Problem

A review of the literature reveals that adolescent substance abuse is not caused by
any single agent, but is the result of the interplay of several factors that interact with the
characteristics of the developing adolescent (Hawkins, Catalano, & Miller, 1992; Swadi,
1999). Although drug use is a global problem, few studies exist that detail either the state
of drug use among Venezuelan youth, or what risk and protective factors might operate to
inﬂuence adolescents to use substances. For example, a recent study sponsored by the
Inter-American Drug Abuse Control Commission of the Organization of American States
looked into the prevalence of substance use among school-attending youth in the

countries of South America. Unfortunately, Venezuela was not included in this study.

Of those studies that do exist on drug use among Venezuelan youth, the
information they provide is very limited in depth and in scope. For example, few
demographic variables are provided, and only scant information is given on which drugs
are most frequently consumed. Additionally, methodological errors render some of the
results dubious. These and other gaps in the extant literature merit further research into
adolescent drug use in Venezuela.

Empirically driven prevention and treatment models have been shown to be
efﬁcacious in reducing problem behaviors among youth (Ozechowski & Liddle, 2000) in
the US. However, the basic research necessary to develop a culturally appropriate
version of these models is still lacking for Venezuela. Given the global prevalence of
substance misuse and its trail of human suffering and misery, it is important to extend this

knowledge into other countries.

Purpose of the Study

The purpose of this study was to begin to lay the empirical foundations necessary
for the development of treatment and prevention models of adolescent substance use in
Venezuela. Since a sample that would be representative of the nation of Venezuela was
out of the scope of the present study, this research may be viewed as a pilot study in one
section of the capital city of Caracas. In order to establish causal paths in the onset of
adolescent substance use longitudinal data are necessary (Heise, 1970). The present study
used a cross-sectional design, and as a result, is viewed as exploratory.

The study identiﬁed the age of ﬁrst use of eight different drugs for school-

attending youths ages 11 to 19 in fourteen Venezuelan high schools from the capital city

of Caracas. The study also explored whether the relationship between four risk and
protective factors known to be associated with adolescent drug use in the US. is
operative as well for Venezuelan school-attending youth in Caracas. Ecological theory is
used to conceptualize the manner in which these factors inﬂuence the development of

adolescent substance use in Caracas, Venezuela.

Signiﬁcance of the Study

There are several beneﬁts to conducting a study that identiﬁes risk and protective
factors linked to substance use in Venezuela. First, in order to inform developers of
effective and cost efﬁcient prevention strategies or treatment interventions, research that
describes prevalence of substance use among youth and the mechanisms that operate to
inﬂuence its onset and maintenance is needed. Second, research that clearly deﬁnes the
problem of adolescent substance use in Venezuela will aid policy makers and educators
in their attempt to guide youth into responsible citizem'y. Third, given the scarcity of
research into this topic in Venezuela, this study will serve as a starting point for future
work by identifying pitfalls and promises in conducting research in Venezuela. Fourth,
studying adolescent substance use in other cultures (e. g., Venezuela), may produce
information that increases our understanding of the mechanism at work in our own U.S.

culture.

Theory Development
Bronfenbrenner (1979) originally conceptualized human development as “a set of

nested structures, each inside the next, like a set of Russian dolls” (p. 87). The “nested

structures” or environments that he identiﬁed to explicate contextual inﬂuences on child
development were: the microsystem, the mesosystem, the exosystem, and the
macrosystem. The microsystem consists of persons who consistently interact directly
with the developing child. The mesosystem involves linkages between the child’s
microsystems or reciprocal inﬂuences between contexts (e.g., family, school,
neighborhood). The third level of inﬂuence, the exosystem, involves settings in which the
child’s development is indirectly inﬂuenced through interaction between a microsystem
and external system (e.g., mother’s workplace). This context is the point at which society
has inﬂuence upon what goes on within the family. The ﬁnal level of inﬂuence, the
macrosystem, involves the general culture in which the individual lives including values
and belief systems that inﬂuence the child’s development.

Later, Bronfenbrenner recognized that a person—context model was insufﬁcient to
address the challenges of delineating and understanding process. He espoused the
process-person-context model of human development, which “permits analysis of
variations in developmental processes and outcomes as a joint function of the
characteristics of the environment and of the person” (Bronfenbrenner, 1989, p. 197). He
also proposed the conceptualization of the chronosystem to encompass the evolving
interconnected nature of the person, environment process over time. These additions led
to the identiﬁcation of the process-person-context-time model (PPCT) of human
development.

As Bronfenbrenner further developed the theory, process came to occupy an
increasingly important role. He emphasized that discernible differences in individual

development, not only across but also within societies, result from the interplay between

10

individual and environment effects. In his bioecological theory of human development
(Bronfenbrenner & Ceci, 1994), he embraced both sides of the nature vs. nurture
argument and posited that individuals possess heritable genetic qualities whose potential
is actualized through progressively more complex reciprocal interaction with persons,
objects, and symbols in the immediate environment through mechanisms known as
proximal processes. According to Bronfenbrenner, the magnitude and the developmental
effectiveness of proximal process are seen to vary as a joint function of the characteristics
of the setting in which they take place, the persons living in that environment and the
nature of the developmental outcomes under investigation. In other words, the focus of
this model is on the “how” certain kinds of genetic potentials (genotype) are actualized to
determine distinct developmental outcomes (phenotype) of effective psychological
functioning. Not all of the genotypic possibilities that the child inherits will necessarily
progress into a phenotypic form. Which phenotypes ultimately emerge will depend on the
interaction between the principal proximal settings of the developing child (mesosystem).
Contexts inﬂuence the proximal processes through resources that are made available and
in terms of the degree of stability and consistency provided over time for their effective
functioning.

From this perspective, the developing child begins with an inherited genetic
potential that follows a path. However, from the very outset the path through which
genotypes are transformed (their potential actualized) into phenotypes (developmental
outcomes) is the mechanism of proximal processes. These processes are driven by a
genetic pattern that selectively attends, acts, and responds, while simultaneously being

shaped by ongoing reciprocal interaction with persons, objects, and symbols in the

11

immediate environment over time. Even developmental changes like puberty that would
seem to be biologically based and thus acontextual, have been shown to be mediated by
family, peer, and school inﬂuences (Simmons & Blyth, 1987). Therefore, social contexts
are always causally involved to some extent in every aspect of human development
(Bateson, Jackson, Haley, & Weakland, 1968).

In the following pages, it is illustrated how various aspects of Bronfenbrenner’s
theory on the ecology of human development was used to guide this research.
Speciﬁcally examined was the role context plays in the onset of adolescent drug use in
Caracas, Venezuela. Accordingly, the study explored adolescent development in the
microsystems of the family, the peer group, the interaction between these systems (what
Bronfenbrenner calls the mesosystem), and the differential effects of this occurring
within a school (a level of analysis Bronfenbrenner calls the exosystem) and within the
culture of a Spanish-speaking, urban, South American city (part of what Bronfenbrenner
calls the macrosystem). This study was limited to a cross sectional design. As a result, the
ecological development over time (what Bronfenbrenner referred to as the chronosystem)

was not considered.

Research Questions, Hypotheses, and Measures
The research questions posed in this study and speciﬁc hypotheses related to them
are presented below. Hypotheses were advanced in areas in which previous research in
the US. and other countries has indicated relationships. Other questions were considered
exploratory in nature; therefore, no hypotheses were formulated for them. The contextual

variables in this study are: family attentiveness, externalizing behavior, peer

12

relationships, and school climate, and are deﬁned below. Individual/demographic
variables in this study are: gender, SES, and race. Legal substances are tobacco and
alcohol, , and illegal substances are marijuana, prescription drugs, cocaine, crack, heroin,
inhalants, or ecstasy. The dependent variable is age of ﬁrst use of a substance, and is a
continuous outcome.
Speciﬁc Questions and Hypotheses
1. What percentage of Venezuelan youth use each of the following drugs: tobacco,
alcohol, marijuana, cocaine, crack, heroin, amphetamines, inhalants, ecstasy, or
prescription?

Data analysis for question 1: Descriptive.
2. Does age of ﬁrst drug use vary by individual variables?
2.1. Does age of ﬁrst actual drug use vary by gender?
Hypothesis: The age of ﬁrst drug use will vary by gender.
2.2. Does age of ﬁrst drug use vary by SES?
Hypothesis: The age of ﬁrst drug use will vary by SES.
2.3. Does age of ﬁrst drug use vary by race?
Hypothesis: The age of ﬁrst drug use will vary by race.
2.4. Does age of ﬁrst drug use vary by family attention?
Hypothesis: The age of ﬁrst drug use will vary by family attention.
2.5. Does age of ﬁrst drug use vary by externalizing behavior.

Hypothesis: The age of ﬁrst drug use will vary by externalizing behavior.

13

U)

2.6. Does age of ﬁrst drug use vary by peer drug use?

Hypothesis: The age of ﬁrst drug use will vary by peer drug use.
2.7. Does age of ﬁrst drug use vary by school climate?

Hypothesis: The age of ﬁrst drug use will vary by school climate.

Data analysis for questions 2.1- 2.7: Hierarchical Linear Modeling (HLM)

. Are school characteristic related to the onset of drug use?

3.1. Does School Condition help to explain the variance in age of ﬁrst drug use?
Hypothesis: School Condition will be related to age of ﬁrst drug use.

3.2. Does Mean SES help to explain the variance in age of ﬁrst drug use?
Hypothesis: Mean SES will be related to age of ﬁrst drug use.

3.3. Does Mean School Climate help to explain the variance in age of ﬁrst drug use?
Hypothesis: Mean School Climate is related to age of ﬁrst drug use.

Data analysis for question 3.1-3.3: HLM.

Measures

The PA CARDO- V

The PACARDO-V is an adapted version of the PACARDO questionnaire for use

in Venezuela. The PACARDO (which stands for P_Anama, Central America, and

Bepublica mmincana) questionnaire was developed for use in a NIDA-funded grant

“Cross-National Research in Clusters of Drug Use” (Dormitzer, et al., 2004). It is a

standardized self-administered questionnaire for adolescents ages 12-17 and was used in

14

studies that included nationally representative samples of students in Central America,
Panama, and the Dominican Republic (N = 12,797). The Spanish version of the
PACARDO-V has been provided in Appendix A, and the English version of the
PACARDO-V has been provided in Appendix B.
The MAMBI

The MAMBI (Which stands for Guia de Observacion Medio AMBIente del
Salon, Colegio y Vecindario, or Observational Guide for the Classroom, School, and
Neighborhood Environment) was developed for use in a NIDA-funded grant “Cross-
National Research in Clusters of Drug Use” (Dormitzer, et al., 2004) and is an
observational guide to be ﬁlled out by the teacher, and/or school administrators. The
purpose of the MAMBI is to assess for the environmental conditions in which the
children are studying (e.g., Are there enough desks and chairs for each student to have
one? Is there barbed wire or broken glass on the top of the walls that surround the
school?) A copy of the MAMBI in both Spanish and English has been provided in

Appendix C.

Conceptual and Operational Deﬁnitions
Individual Level Variables
Family attention.
Conceptual — this variable taps two dimensions that have been shown to be
important in the onset of adolescent substance using behavior: (a) the extent
to which the youth’s relationship with parents or caretakers reﬂects positive

communication, warmth, and cohesion, and (b) the extent to which the

15

youth’s relationship with parents or caretakers reﬂects positive boundary
setting, monitoring, and involvement.

Operational — the average score on items 14, 15, 17, 25 and 16, 18, & 20,
respectively on the PACARDO-V questionnaire. The scores were
standardized to have a mean of 0 and a standard deviation of 1 for ease of
interpretability. A positive score indicates above average family attention.

Externalizing behavior.

Conceptual — extent of youth participation in delinquent acts and risky behavior.

Operational — the average core on items 40, 41, 42, & 43, 48 on the PACARDO-
V questionnaire. The scores were standardized to have a mean of 0 and a
standard deviation of 1 for ease of interpretability. A positive score indicates
above average externalizing behavior.

Peer drug use.

Conceptual — extent of drug use among the youth’s peer group.

Operational — the average score on items 30, 31, & 33-36. on the PACARDO-V
questionnaire. The scores were standardized to have a mean of 0 and a
standard deviation of 1 for ease of interpretability. A high score indicates
above average afﬁliation with a peer group that would expose the youth to
drugs.

School climate.

Conceptual — student perception of their sense of acceptance and belonging to

their school.

16

Operational — the average score on items 44, 50, 51, & 52 on the PACARDO-V
questionnaire. The scores were standardized to have a mean of 0 and a
standard deviation of 1 for ease of interpretability. A high score indicates
above average school climate.

Socioeconomic status (SES).

Conceptual - An individual’s or group’s position within a hierarchical social
structure. Socioeconomic status depends on a combination of variables, and
will be deﬁned in this study through student response to caretaker’s
education level, type of neighborhood of residence (housing
project/barrio/casa, urbanization/apartment, quinta), number of vehicles
owned by immediate family, and the number bedrooms in their place of
residence.

Operational — the average score on items 6, 9, 7, 12, & 13 on the PACARDO-V
questionnaire. The scores were standardized to have a mean of 0 and a
standard deviation of 1 for ease of interpretability. A positive score indicates
above average SES.

Race.

Conceptual — In social science and popular understanding, race is thought to refer
to phenotypical differences between groups of people, while ethnicity
denotes cultural differences. In a review of international census forms,
Morning (in press) found that only the United States uses separate questions
to measure its citizens’ race versus their ethnicity. In Venezuela, as in most

South American countries, ethnicity is used to refer to indigenous

17

populations while race is secondary category coming after the word color
and referring to skin tone (Hooker, 2005; Morning, in press). Therefore, for
the present study race will be conceptualized as skin tone and will use 4
popular designations from Venezuelan culture (i.e., Negra, Morena, Blanca,
and Indigina or Black, Brown, White, and Indigenous respectively).

Operational — item ﬁve on the PACARDO-V questionnaire.

 

Second Level Variables
Mean school climate.
Conceptual — The extent to which the school maintains an environment that
fosters a sense of belonging and acceptance among students.

Operational — the mean student scores within school j of student climate measure.

 

Mean peer drug use.

Conceptual — Average aggregate drug use by peers in a given school.
Operational — the mean score of students in school j of peer drug use measure.
Mean SES.

Conceptual — Average SES of students in school j of SES measure.

Operational - the mean score of students in school j of SES measure.

 

School condition .

Conceptual — an index of the general environment of the school building,
resources for the students, and area adjacent to school property.

Operational — the composite score grouped by school of items 2-40 from the

MAMBI questionnaire.

18

CHAPTER II: BACKGROUND
Review of the Literature

The premise of this study is that the synergistic effects of the adolescent’s social
ecology will offer a more ﬁrm foundation from which to develop intervention strategies
for the prevention and treatment of adolescent substance abuse in the Venezuelan culture.
It is common knowledge that parents exert a powerful inﬂuence, albeit positive or
negative, over the development of their offspring (e.g., Hirschi & Gottfredson, 1993).
However, parental inﬂuence does not occur in a social vacuum, and the effect of the
parent-child relationship cannot be fully understood except within the context of social
factors (Kraemer, Stice, Kazdin, Offord, & Kupfer, 2001; Von Bertalanffy, 1968). The
effect that parents leverage over their offspring is likely to be moderated by the
intersection of peer groups, which in turn are nested within schools and neighborhoods;
all of which interact within an overarching culture.

Much research has been done to identify risk factors in adolescents related to the
development of drug and alcohol problems in order to promote an understanding of the
complex causal chains involved. This chapter will review the relevant literature regarding
how individual characteristics, and inﬂuences from families, peer groups, schools,
neighborhoods, and the Venezuelan culture effect the onset of substance use in

adolescents.

Individual Inﬂuences on the Development of Substance Use in the US

Some family systems theorists have argued that all behavior must be understood

within its systemic context (e.g., Bateson, 1972; Watzlawick, Weakland, & Fisch, 1974).

19

From this epistemology, it is illogical to view a behavior problem as an attribute of an
individual. Following then, in order for a behavior to be characterized as aggressive, it
must occur within a sequence of interactions and be assessed in a given social context
that would enable such a conclusion. For example, in American Football, one player
bumping helmets with another could be viewed as normal and within the rules of the
game, aggressive and punishable by a penalty, or celebratory depending on the context in
which it occurs.

In the context of human development, systems theory asserts that what is
considered developmentally appropriate behavior is relative to a socially constructed
standard within a determined context. For example, even with problems that are certainly
the result of some biological abnormality or genetic marker such as childhood autism,
there is evidence that behavioral problems have some association with family
relationships and school environment (Morgan, 1988). This view does not hypothesize
causality such that the parents cause the child’s autism in a linear fashion. Linear
thinking assigns a direct cause to problems and, consequently, assigns blame. Systems
thinking, on the other hand, emphasizes the bidirectional nature of behavior in that the
child’s behavior, whatever its assumed cause, will impact that of the parents, which will
in turn affect the child and so on (Watzlawick, Beavin-Bavelas, & Jackson, 1967).
Whereas individual variables should not be viewed in isolation of their context, the
systems-based ecodevelopmental model acknowledges that individual members do
contribute uniquely to behavior problems, and should be considered for their implications

in both the conceptualization and treatment of problem behavior (Bronfenbrenner, 1989).

20

Several variables have been identiﬁed in the literature as mechanisms through which
individual characteristics may operate to form risk factors for the onset of alcohol and
other drug use (AOD). Swadi, (1999), in a review of the literature points to numerous
personality attributes that have been associated with AOD such as poor self control, high
levels of novelty seeking, risk taking, ambition, negative affect, impulsiveness, hard
working, self reliant, feeling capable and accepted, unsociable, untrustworthy, rebellious,
and impulsive.

In another review of the literature, Colby, Lee, Lewis—Esquerre, Esposito-Smythers,
and Monti (2004) suggest two important cognitive factors: alcohol outcome expectancies
(i.e., greater endorsement of positive alcohol expectancies is associated with higher
quantity and frequency of drinking) and craving (i.e., low urge-speciﬁc coping skills that
are related to increased drinking). However, Martino and colleagues’ (2006) ﬁndings
suggest that attempts to alter adolescent alcohol expectancies are likely to fail unless they
address the inﬂuence of immediate social factors on these beliefs.

Other researchers have stressed the role of gender and ethnicity in the development
of adolescent alcohol use (Grifﬁn, Botvin, Scheier, Diaz, & Miller, 2000). These authors
found that Black youth reported the fewest risk factors and the lowest alcohol use, White
youth reported the most risk factors and the highest alcohol use, and Hispanic youth
reported the fewest protective factors and intermediate levels of alcohol use. Females
were found to have a reduced tendency to develop drinking problems across all
ethnicities in comparison to males.

Researchers have found that among childhood characteristics predicting adolescent

alcohol use disorders, childhood mental health issues, (including conduct, attention

21

deﬁcit hyperactivity, major depressive, and anxiety disorders) were all prominent factors
in the onset of alcohol difﬁculties (e. g., Clark & Winters, 2002). However, some have
suggested that a common genetic and environmental inﬂuence is responsible for the
association between childhood mental disorders and alcohol and other drug use disorders
(Waldman & Slutske, 2000). Tarter, Kirisci, Mezzich, Cornelius, Pajer, et al. (2003)
reported on a unidimensional trait they called neurobehavioral disinhibition, an index
formed from measures of affect, behavior, and cognition. They found that
neurobehavioral disinhibition was successful at discriminating between boys at high
average risk from those at low average risk of substance use at ages 10—12, and predicted
substance use disorder at age 19 with 85% accuracy. Clark (2004) also suggests that a
more parsimonious approach to the association between childhood mental health
disorders and substance use is to consider an underlying common liability trait termed

“psychological dysregulation.”

The Role of Genetics

The role of genetics in the development of alcohol abuse and alcoholism has gained
much momentum since Jellinek’s early work (J ellinek, 1946). Research seems to indicate
that problem drinking is a heritable family disorder with a genetic origin (e.g., Cloninger,
1987). In his widely cited study, Cloninger (1987) posited two types of alcoholism, Type
I and Type II. The Type II alcoholic was set forth to distinguish those individuals who
have a stronger genetic predisposition to, (a) initiate alcohol-seeking behavior, (b) have
earlier onset of alcoholism, and (c) progress at a different rate from susceptibility to loss

of control after drinking begins. National twin studies suggest that genetic factors explain

22

much of the variance in the development of alcoholism and other drugs (e.g., Hettema,
Corey, & Kendler, 1999; Prescott & Kendler, 1999). Advances in genetic studies have
led to the identiﬁcation of some underlying genes that are substance-speciﬁc, such as the
alcohol metabolic genes, and it is hoped that such advances will eventually lead to more

successful treatment approaches (Goldman, Oroszi, & Ducci, 2005).

The implication of these studies is that individual characteristics are an important
consideration in the understanding of the etiology of adolescent substance abuse.
However, as Bronfenbrenner (1995) points out, characteristics of the individual are often
studied as developmental outcomes, but seldom conceptualized as sources of variation in
the person’s susceptibility or risk to the developmental effects of proximal processes (i.e.,
enduring patterns of interaction between the person and his/her environment). In regards

to his model Bronfenbrenner (1995) states,

What is most revealing about proximal processes, however, is not the gain
in predictive power that they provide, but their substantive and theoretical
signiﬁcance as the mechanisms of organism-environment behavioral
interaction that drive development, and the profound ways in which these
mechanisms are affected by characteristics of the developing person and
of the environmental context in which the interaction takes place (pp.

626).

As such, no single factor or event can be said to "cause" addiction, genetically or otherwise.
From this perspective then, causality ceases to be linear and becomes reciprocal in nature.
The parents bring to the family of procreation certain values, traditions, rules, and

boundaries from their families of origin (Bowen, 1974; Sullivan, 1953). While developing

23

both emotionally and physically within their context of peers, school, and community
activities, the child brings inﬂuences from other systems into the family. The child reacts to
his or her parent's behaviors, provoking the parents to react in turn, and so on, in a multiple
reciprocal fashion (Cox & Ray, 1994). It is to these interactions that the focus of this section

will now turn.

Microsystem Influence on the Development of Substance Use in the US.

Family as a Context

Family relationships have been found to play a major role in the development of
adolescent substance abuse (Chassin, Ritter, Trim, & King, 2003; Hawkins, et al., 1992;
Stanton & Todd, 1982). Poor parenting practices have been consistently associated with
increased substance use and delinquency in adolescents (e.g., Belcher & Shinitzky, 1998;
Calvert, 1997). Inconsistent discipline is positively related to development of drug use
(Gonnan-Smith, Tolan, Zelli, & Huesmann, 1996; King & Chassin, 2004). Reduced
parental monitoring is associated with higher rates of adolescent misbehavior including
the transition into substance use (Chilcoat, Breslau, & Anthony, 1996; Steinberg, Fletcher
& Darling 1994) and increases in delinquency and aggression (Patterson, & Stouthamer-
Loeber, 1984). Parental support was found to promote a protective function against
adolescent substance use (Wills, Resko, Ainette, & Mendoza, 2004), as was positive
parent-child communication (Anderson & Henry, 1994). In other studies increases in
family cohesion (Hussong & Chassin, 1997), the parent-adolescent attachment
relationship (Brook, Brook, Gordon, Whitemnam, & Chohen, 1990), authoritative

parenting style (Baumrind, 1991; Fletcher, Darling, Steinberg, & Dombusch, 1995), and

24

parent-adolescent autonomous-relatedness (Samuolis, Hogue, Dauber, & Liddle, 2005)
were related to decreases in adolescent substance use.

In addition to the quality of family relationships, research has also examined the
link between family structure and behavior problems in adolescents. For example, Blum
and colleagues (2000) using data from the National Longitudinal Study of Adolescent
Health found that youth from single-parent families were at greater risk than youth from
two-parent families on every health risk behavior studied. However, the explanatory
power of these analyses was so small that it only “marginally advances our understanding
of the factors that contribute to the behaviors under study” (p. 2).

In the one study that was found that looked at the interaction between family
functioning and family structure and its effect on adolescent substance abuse, Grifﬁn et
al. (2000) found that family structure was moderated by gender. Results indicated that
boys from single-parent families engaged in more problem behaviors compared to girls
and to youth of either gender from two-parent families. However, increased parental
monitoring buffered these effects for boys in the single parent families.

Zhou, King, and Chassing (2006) looked at how family history density of
alcoholism (F HD) interacted with a measure of family functioning (family harmony) over
time to impact the development of adolescent substance use disorders (SUD). They found
that family harmony had a protective effect for the development of SUD for low to
moderate levels of FHD. However, this effect was limited to the development of
substance use disorders apart from alcohol dependence and lost its potency for higher

levels of FHD.

25

Barnes, Reifman, Farrell, and Dintcheff (2000) summarize the literature on
parental socialization and child outcomes into two key constructs: parental support
(communication that would indicate to the child that they were loved and accepted) and
parental control (behaviors intended to promote child behavioral compliance to parental
expectations). In a six-wave longitudinal study, the authors found a signiﬁcant link
between parental support and adolescent outcomes. Surprisingly, they found that neither
coercive control nor parental inductive control (telling and explaining to adolescents why
they should not do something) to be signiﬁcant predictors of positive child outcomes.
Only parental monitoring (e. g., did parents know the whereabouts of their adolescent
children) emerged to be a signiﬁcant predictor of desired child adolescent outcomes.
These ﬁndings are consistent with Baumrind’s (1991) typology, which conceptualizes
authoritative parenting as those parents who combine boundary setting (monitoring) with
responsiveness (support).

The implication of these studies is that family attentiveness does serve as a
context for gaining insight into the onset of adolescent substance abuse. Following
Bronfenbrenner’s bio-ecological model (1994), parental style and practices interact with
individual characteristics to create a willingness on the part of the child to be socialized,
and this willingness is predictive of behavioral outcomes (Darling & Steinberg, 1993).
Peers as a Context

Developing a network of friends is an important part of early adolescence (Ianotti,
Bush, & Weinfurt, 1996). Adolescent prosocial development, (Simmons & Blyth, 1987),
and moral development (Schonert-Reichl, 1999) are both inﬂuenced by peer

reinforcement. Carlo, Fabes, Laible, and Kupanoff (1999) suggest a unique inﬂuence

26

from peer interaction that does not exist in adult-adolescent interaction due to the more
equal status between peers. They observed that peers reciprocate peer prosocial
behaviors, such that cycles of prosocial behavior are formed.

Peer inﬂuences have traditionally been a robust predictor of adolescent substance
abuse (Bauman & Ennett, 1994; Hawkins, et al., 1992). Research has provided support
for at least two theories to explain the relationship between peer inﬂuences and substance
use: Individual Characteristics Model, and the Peer Inﬂuence Model (Curran, Stice, &
Chassin, 1997; Vitaro, Tremblay, Kerr, Pagani, & Bukowski, 1997; Wills & Cleary,
1999). In the former, adolescents involved in delinquent behavior select friends who are
also involved in deviant behavior. This conforms to the adage “birds of a feather ﬂock
together” and precludes the idea of an individual being corrupted by hanging around the
“wrong” crowd. In the second view, deviant friends inﬂuence a new group member to
adopt delinquent behavior through peer pressure. That is a young person who is not
involved in delinquent behavior is inﬂuenced to adopt this behavior due to his association
with the group.

Hussong (2002) examined adolescent peer interaction along three dimensions
predicting adolescent substance use: best friendships, peer cliques, and social crowds.
She found that the strongest of the three dimensions was the extent of substance use by
the adolescent’s best friend. However, the dimensions had an additive effect such that
youth who were more embedded in peer contexts were more likely to use substances
themselves.

Steinberg, Darling, and Fletcher (1995) found that peer groups exert an inﬂuence

on school achievement above and beyond that of the family. Parents were found to have

27

the most important inﬂuence on a youth’s long-term educational plans, but peers
inﬂuenced more powerfully their day-to-day activities in school (e. g., how much time
spent on homework, level of enjoyment of school, etc.). They found that an important
predictor of academic success is the level of agreement in values between an adolescent’s
family and peer group. Other researchers have also found a relationship between school
performance and risk of substance-using behaviors (Resnick, Bearman, Blum, Bauman,
Harris, et al., 1997). Thus, it would appear that peer relationships are an important
variable in understanding the etiology of adolescent substance use. Later in this section,
the intersection between the microsystems and how they affect adolescent substance
abuse is discussed.
Schools as a Context

Contrary to popular belief, research seems to support a general trend toward
increased prosocial behavior among children as they get older (Fabes, Carlo, Kupanoff,
& Laible, 1999). However, early adolescence is a time of rapid, and sometimes difﬁcult
physical, cognitive, and psychosocial maturation, which many individuals have
difﬁculties navigating (Carlo et al., 1999). For example, middle school was conceived as
a means of making the transition into secondary education less turbulent (Simons-Morton
et al., 1999); but, early adolescents can have a particularly difﬁcult experience moving
into a new academic environment, and their prosocial development may be hindered in
the face of multiple changes (Eccles, Lord, & Midgley, 1991).

Johnston et a1. (1994) describe a sizeable increase in problem behaviors during
the transition into adolescence. For example, they report that prior to middle school less

than 10% of sixth graders have used tobacco and 5% have used alcohol. However, by the

28

eighth grade the ﬁgures jump to 30% and 70% respectively. Simons and Blyth (1987)
found that the number of life transitions is negatively correlated with grades and
participation in extracurricular activities for both boys and girls, and with self-esteem for
girls. Negative motivational and behavioral characteristics are also associated with early
adolescent transitions (Eccles, Lord, & Midgley, 1991).

Hirschi’s (1969) seminal work in social bonding theory posits that attachment to
socially conforming institutions such as the school provides a protective function against
deviant peer groups. He identiﬁed four elements of social bonding that if present would
deter deviant or delinquent behavior: attachment, commitment, involvement, and belief.
When youth have an attachment to a prosocial institution they are more able to refuse to
engage in deviant and delinquent behavior. Commitment refers to the personal time and
energy invested in the institution, and more investment leads the youth to uphold the
institution’s norms and ideals. Involvement deals with literal hours in the day; the more
time spent in institutional activities the greater the attachment. Finally, belief refers to the
extent that the youth agrees with the legitimacy of the institution’s values and norms. The
more agreement, the more likely the youth will be to internalize these beliefs and engage
in them as a personal choice.

Subsequent research has found that a positive afﬁliation toward school may serve a
protective function against many antisocial behaviors (Hill & Werner, 2006). School
attachment has been deﬁned in the literature as a sense of affection toward and enjoyment
of school (Hill & Werner, 2006) or as a basic expression of the human need to ﬁt in
(Anderman, 2002), and is associated with positive outcomes such as school completion

and success (Marcus & Sanders-Reio, 2001 ). Low school attachment, on the other hand,

29

is associated with negative outcomes such as aggressive behavior, delinquency (Grifﬁn,
Botvin, Scheier, Doyle, & Williams, 2003), and substance abuse (Murguia, Zeng-yin, &
Kaplan, 1998; Najaka, 2001). The effects of school attachment on adolescent behavior
have been found to be similarly correlated in other cultures (J unger & Marshall, 1997).

Schools do not only effect the developing youth directly through the creation of a
personal bond. Rather, Petronis and Anthony (2003) argue that there is a “contagion
effect” related to how contextual inﬂuences can explain geographic concentrations of
drug use in a certain school when compared to another. According to the contagion
model, students within schools develop similar substance use habits through social
interactions with other peers (Murray & Harman, 1990). A school climate of norms and
attitudes toward drug use may be transmitted from peer to peer so to encourage or
dissuade substance use making varying substance abuse rates noticeable across schools
(Kumar, O'Malley, Johnston, Schulenberg, & Bachman, 2002). For example, Henry and
Slater (2007) found that regardless of a student’s personal level of school attachment,
students who attend schools where the pupils overall tend to be well attached are less
likely to use alcohol. Other researchers have found that a sense of community in the
classroom and school enhances prosocial development (Solomon, Battistich, Watson,
Schaps, & Lewis, 2000).

However, the contagion model is insufﬁcient in and of itself to explain all of the
variance between schools since it must assume some initial across school variability in
attitudes toward drugs to which other students are exposed. Therefore, a second
explanation is that signiﬁcant sociodemographic characteristics of the school or in the

neighborhood in which the school is nested are operating to inﬂuence the onset of

30

youthﬁil substance use (Ennet, et al., 1997). Studies show that adolescent problem
behaviors such as rates of delinquency, teenage pregnancy, substance abuse, and low
education are higher in disorganized and impoverished schools and neighborhoods (e. g.,
Furstenberg, 1994; Hawkins, Catalano, & Miller, 1992).

Together these studies emphasize the importance of looking at individual
characteristics as well as the the effect of family, peers, and schools at the individual
level as a microsystem inﬂuence. However, these studies also present a second level
contextual effect, more akin to an exosystem inﬂuence. This section now turns to the

interaction of microsystem inﬂuences on adolescent substance use.

Mesosystem Inﬂuence on the Development of Substance Use in the US.

F amily-Peer-School Interactions

Grifﬁn, et al. (2003) suggest that parents inﬂuence their child’s peer network
through the formation of conventional values in their adolescent. These youth then go on
to seek out friends who hold similar values. Patterson and colleagues similarly suggest
that the patterns formed in parent-child interactions are replicated in other settings such
as school and the peer group, which in turn become self-reinforcing (Dishion, Patterson,
Stoolmiller, & Skinner, 1991). This coincides with Steinberg and associates’ (1995)
ﬁndings that parents are the most salient inﬂuence on their children’s long-term
educational goals, but that peers are a more persuasive inﬂuence on the day-to-day
activities that directly affect adolescent school performance. Interestingly, they also
found this relationship was moderated by ethnicity with minority students relatively more

inﬂuenced by their peers than European-American youth. Due to the segregated nature of

31

schools, minority youth ﬁnd their choices of peer groups to be restricted. For example,
Asian-American youngsters reported the highest level of peer support for academic
achievement but the lowest levels of parental involvement in school related activities. In
contrast, African-American parents score among the highest in regards to parental
involvement in their child’s school, but African-American youth ﬁnd it difﬁcult to gain
membership into the “brains” peer group. Therefore, the negative effects of a lack of
parental involvement for Asian-American students was offset by the homogenizing
inﬂuence of their peer group, and for African-American students, the positive beneﬁts of
supportive parents was offset by a lack of support from their peer network. Thus at the
mesosystem level (intersection of the family and peer microsystems) the macrosystemic
inﬂuence of a culture that promotes segregation was a moderating factor.

In similar vein, Eccles and colleagues (1993) suggest that parenting styles that
tend to be more authoritarian and that are unresponsive to adolescents’ developmental
needs for increases in autonomy may amplify the risk for adolescent substance use in part
due to a decrease in school attachment. Lagerway and Phillips (2003), in a study on
Latinos, comment that student success was related to parent’s encouragement to do well
in an effort to combat racial stereotyping. Other researchers found that school attachment
differed among racial groups (Johnson, Crosnoe, & Elder, 2001) and that student’s
perception of discrimination on the part of teachers and administrators created
institutional barriers that affected levels of school attachment (Conchas, 2001; Martinez,
DeGarmo, & Eddy, 2004). Peer selection, likewise, is associated with school attachment
with members of deviant peer groups being more likely to have lower levels of school

performance and to drop out (Carlo et al., 1999).

32

In a study of contextual interactions between neighborhoods, families, peers, and
schools, Cook, Herman, Phillips, and Settersten (2002) found that there was no one
context that stood out, but that each context had an independent and modest effect on
adolescent outcomes. They concluded that contexts do matter, but that their effect is
additive instead of nonlinear. In the aggregate, the quality of a school reﬂected the quality
of the neighborhood, homes, and peer groups with the reverse being equally true.
Therefore, when a context was positive it tended to protect the youth, and when it was
negative it placed the youth at greater risk. Cumulatively, more negative contexts spelled
greater risk for poor outcomes.

Taken together these studies underline how the interaction between different
microsystems inﬂuences the development of substance use in adolescence. There is also a
need for more research into how individual characteristics like race, or a speciﬁc genetic
marker may be conceptualized as sources of variation that inﬂuence the person’s
susceptibility to the developmental effects of proximal processes that operate within the

clustering effects of a context.

Exosystem Inﬂuence on the Development of Substance Abuse in the US.
Neighborhood as a Context for Family and Peer Interaction
Few studies have systematically examined neighborhood inﬂuences on substance
abuse problems (Duncan, Duncan, & Strycker, 2002). Systemic theory suggests that
when members of a community form local social ties, their ability for community social
control is augmented (Sampson, 1988). Sampson, Raudenbush, and Earls (1997) testing

this theory found that a latent construct of collective efﬁcacy deﬁned as a willingness to

33

intervene on behalf of the common good and a neighborhood sense of cohesion and trust,
exerted a substantial effect on multiple measures of violence in disadvantaged
neighborhoods. Duncan et al., (2002) found a negative relationship between levels of
social cohesion and perceived problem with youth alcohol and drug use in the
neighborhood. Moreover, they report that the interclass correlations for perceived social
cohesion lend credibility for looking at drug and alcohol problems among youth at the
neighborhood and individual levels of analysis. Nash and Bowen (1999) did not measure
actual neighborhood crime rates or social control but only the adolescent’s perception of
each. They found that adolescent’s perception of neighborhood crime served as a risk
factor and their perception of neighborhood informal social control acted as a protective
function for their own prosocial behavior.

In a different approach, Fletcher and colleagues (1995) examined the extent that
adolescents are inﬂuenced by the parenting style of their peer’s parents living in the same
community independent of the adolescent’s own parents parenting style. They found that
a preponderance of parental authoritativeness (Baumrind, 1991) in the adolescent’s peers
is related to a variety of healthy adjustment indicators beyond the contribution made by
the adolescent’s own parents. The link, however, was not direct, but was indirectly
transferred through the peer’s choice for non-deviant peers. Surprisingly though, if the
friend is already engaged in delinquent activities, the prevalence of authoritativeness
among the friends’ parents directly reduces delinquent behavior. Thus, it seems that the
inﬂuence of multiple authoritative parents in a neighborhood creates a community-wide

impact against delinquent behavior (Sampson & Groves, 1989).

34

As pointed out by Steinberg et a1. (1995), the norms and parental monitoring that
inﬂuence adolescent development are most effective when there is an overlapping of
adult and youth social networks. That is, parents not only know their children’s friends,
but also the parents of their children’s friends. Furstenberg (2005) echoed these ﬁndings
and asserted that parental efforts to rear children are more successful when they reside in
communities that have a high consensus and intergenerational closure (overlapping social
networks between parents and children) in regards to child rearing. However, while
socially integrated parents are generally more successful in their child rearing attempts
Steinberg et a1. (1995) point out an important caveat:

Although we tend to think of social integration as a desirable endpoint,

its desirability depends on the nature of the people that integration brings

one into contact with. There are many communities in contemporary

America in which it may be more adaptive for parents to be socially

isolated than socially integrated. Indeed, some of F urstenberg’s (1990)

recent work on family life in the inner city of Philadelphia suggests that

social isolation is often deliberately practiced as an adaptive strategy by

many parents living in dangerous neighborhoods (p. 459).

Macrosystem Inﬂuences on the Development of Substance Use: The Case for
International Research
Research that remains culture-bound is at odds with the goal of scientiﬁc
investigation—generalizability. Much has been written regarding the need for cultural

adaptations of extant prevention and treatment models for the diverse ethnic groups

35

residing within the United States (e.g., Castro, Barrera, & Martinez, 2004; Resnicow,
Soler, Ahluwalia, Braithwaite, & Butler,2000; Turner, Wieling, & Allen, 2004). While
the need to tailor substance use prevention and treatment programs to the social
characteristics of the target population is clear, how to ascertain the pertinent
characteristics of a given culture is considerably murkier. Resnicow, et a1. (2000)
describe two levels of cultural sensitivity that should be considered: surface structure,
which involves matching intervention materials and messages to observable, “superﬁcial”
characteristics of a target population, and “deep structure,” which involves incorporating
the cultural, social, historical, environmental, and psychological forces that inﬂuence the
target health behavior in the proposed target population. It is this “deep structure” that
will be considered in what follows.

Cultural values and beliefs refer to the implicitly or explicitly expressed ideas
regarding what is good, right, and desirable in a society and on which the speciﬁc norms
for appropriate behavior are founded in a given group of individuals (Schwartz, 1999).
Bronfenbrenner cited by Luscher (1995) lays out a proposition detailing the effects of the
interaction between culture and proximal processes in human development:

Maj or determinants of the contents and effects of proximal processes are

systems of belief and knowledge about human development and how it

takes place. These systems exist on three levels. From a developmental

perspective, they originate in the broader sociocultural and institutional

structures of the larger society, both formal and informal. These systems

of belief and knowledge are then transmitted, through a variety of

pathways, into the more immediate settings of family, school, peer group

36

and workplace, where they exert their direct effects on proximal

processes. Finally, through the operation of these processes over an

extended period of time, systems of belief are internalized and become

characteristics of the developing person, and, as such, inﬂuence the

course of that person’s subsequent development (p. 573).

Accordingly, knowledge and beliefs are seen as cultural phenomena (or as Luscher
suggests, the culture itself) that are transferred from generation to generation through
reciprocal interaction in the immediate environment. Once ingrained in the society these
beliefs form the macrosystem inﬂuence on the developing person (Bronfenbrenner,
1979). Inherent in Bronfenbrenner’s macrosystemic inﬂuence is the recognition of the
heterogeneity that resides within any culture as well as the inﬂuence of culture being
anchored in the context of a historic period in time. That is, the traits of the individual
(genotype) interact with the microsystem, mesosystem, exosystem, macrosystem, and
chronosystem to produce a developmental outcome (phenotype).

Schwartz (1999) in a theory of cultural values proposed seven types of values on
which cultures can be compared by considering three dimensions that he proposes
confront all societies. The ﬁrst cultural dimension is Conservatism vs. Autonomy.
Conservatism describes cultures in which the person is viewed as entity embedded in the
collectivity and ﬁnds meaning in life through social relationships and participating in a
shared way of life. Autonomy is subdivided into Intellectual Autonomy and A ﬁective
Autonomy, and describes cultural emphasis on the desirability of individuals
independently pursuing their own ideas (curiosity, broadmindedness, creativity) and

affectively positive experience (pleasure, exciting life, varied life) respectively. A second

37

dimension is Hierarchy vs. Egalitarianism. Hierarchy describes a cultural emphasis on
the legitimacy of an unequal distribution of power, roles and resources (social power,
authority, humility, wealth). Egalitarianism, on the other hand, describes a cultural
emphasis on transcendence of selﬁsh interests in favor of voluntary commitment to
promoting the welfare of others (equality, social justice, freedom, responsibility,
honesty). The third dimension speaks to the relation of humankind to the natural and
social world expressed by Master vs. Harmony. Mastery describes a cultural emphasis on
getting ahead through assertive behavior (ambition, success, daring, competence).
Harmony describes a cultural emphasis on ﬁtting harmoniously into the environment
(unity with nature, protecting the environment, world of beauty).

Schwartz (1999) in a test of his theory surveyed 49 nations of the world. Included in
his survey were three Latin American countries: Mexico, Brazil, and Venezuela. The
results place all three Latin American countries very close to the intersection of the three
dimensions (i.e., the world average on all values). In his report, Schwartz did not offer an
interpretation of the ﬁndings for the Latin American countries; however, two equally
plausible interpretations would seem to ﬁt the data. First would be to infer values that
could be considered the mid-point of the dimensions such as an avoidance of extreme
positions, valuing relative ﬂexibility, being more present oriented, and spontaneous. A
second and perhaps more conservative position might be to assume considerable within
group cultural heterogeneity among the three countries such that they effectively

canceled each other out and regressed toward the mean.

38

Venezuela as a Context

Venezuela is a very diverse society. Historically, around the time of the Second
World War Venezuela became an extremely attractive destination for immigrants from
around the world. The mild climate, beautiful scenery, and petroleum rich land created an
appealing environment for many. Among others, large immigrations from Italy,
Germany, Spain, Portugal, China, Israel, Argentina, Chile, and Columbia came looking to
make their fortunes, and many did. From the 19605 until the beginning of the 19803 it
was not unusual for middle-class Venezuelans to take weekend shopping trips to Houston
or Miami and to send their children to study in Europe or the United States. They were
given the nickname “Dame Dos” among store patrons, which means “give me two,” from
their practice of saying, “Oh that’s cheap, give me two.” Politically, Venezuela has
traditionally been a social democracy run by two primary parties, but with smaller more
radical groups having a noticeable inﬂuence. Economically, the country has operated
under a quasi free enterprise system. In the early 19705, the government nationalized oil
production and set price controls on most products and created a national health care
system for the poor. Property rights were ﬁercely protected, but the poor were allowed,
“squatters rights” on government land. In the 19903 after more than a decade of low oil
prices, the International Monetary Fund intervened to encourage a more free-market
economy and divestment of government owned enterprises in the face of an escalating
national debt. The government, at that time, owned the largest bank in the country, the
only international airline, the telephone, water, and energy companies amid others. The
economic turmoil was not without its social unrest. Numerous strikes, protests, and riots

plagued the country throughout the 19905. Populist movements began to gain adherents,

39

and the country suffered an attempted coup d’Etat that ultimately ended with the
president being impeached.

The military leader of the failed coup, after receiving a presidential pardon that
released him and his compatriots from prison, formed a political party that won the 1998
elections. The Hugo Chavez administration began to undo much of what the International
Monetary Fund had imposed. The country began to move once again toward what some
have described as a more collectivistic and hierarchical orientation (Hofstede, 2001;

Triandis, 1995).

The Qualitative Case of Venezuela

Culture joins with social structure, history, demography, and ecology in complex
reciprocal relations that inﬂuence every aspect of how we live (Schwartz, 2006). Still,
measuring culture can be difﬁcult. Schwartz and other cultural theorists (Schwartz, 2006)
look at children’s stories, at the systems of law, at the ways economic exchange is
organized, or at socialization practices to reveal the cultural orientations in a society.
When researchers try to identify culture by studying the literature of a society or its legal,
economic, family, or governing systems, what they seek, are underlying values
(Schwartz, 2006).

A pre-dissertation fellowship provided by the MSU International Studies
Program, allowed the author to travel to Caracas, Venezuela in the summer of 2006 to
collect pre-research data in the form of interviews and focus groups. In an attempt to gain
a better understanding of the characteristics of low-income families in Venezuela and

their effect on the development of adolescent substance abuse, ﬁve ethnographic records

40

were developed from focus groups with high school teachers (2 groups), low-income
parents of adolescents (2 groups), and psychotherapists (1 group) who work with youth
and families. The preliminary ﬁndings suggest several areas of fruitful investigation.
First, there seems to have occurred an important shift in cultural values regarding
parenting in Venezuela. All ﬁve focus groups concurred that a determining factor in the
onset of adolescent substance abuse in Venezuela has been an increase in parental
perrnissiveness expressed through a lack of parental monitoring and consistent discipline.
In one group of parents comprised exclusively of mothers of adolescents, the participants
conveyed that they had been reared in strict homes with rigid boundaries, but that they
were much more permissive with their own children. The observations of these mothers
are particularly important since Venezuela has been described as a “matrifocal” society
(Recagno-Puente, 1998). Particularly in lower income families, the conjugal system is
unstable, fathers are peripheral, and mothers are venerated which leaves the mother, aunt,
or grandmother as the affective and organizational center of the family (Lodo-Platone,
2004). In part, this shift may be attributed to recent swells of women in the workplace. As
a result of the declining economy, increasingly more women have left their traditional
roles as homemakers to aid in the family’s ﬁnances. Older siblings, grandmothers, aunts,
and neighbors often form networks of extended kin in order to attend to younger children.
One Venezuelan researcher (Lodo-Platone, 2004) in a qualitative study of familial
organization in low-income communities looked at ﬁve aspects of family organization:
(a) daily problem solving, (b) family communication, (0) behavioral patterns in the
designation of responsibilities with the household, (d) authority, supervision and control

standards, and (e) affective relationships and the reciprocal expressions of feelings. Her

41

ﬁndings regarding the prevalent patterns of family functioning under these ﬁve categories
are described below along with the corresponding ﬁndings from focus groups conducted
by the author.

First, daily problem solving is characterized by very little planning to avoid
possible problems. The lack of resources often makes it difﬁcult for families to plan
effectively for the future. Problems tend to be dealt with as they occur and solutions are
improvised on the spur of the moment. In the author’s own research a fatalistic mentality
was found among the poor in Venezuela that often leads them to an “eat drink and be
merry, for tomorrow we may die” view of the world. This perception of the efﬁcacy of
human deliberation and actions has important implications for the implementation of
parenting and other programs that rely upon coaching individuals to be more purposeful
in their interactions.

Second, communication patterns tend to center around the events of the day.
Lodo-Platone proposes that discussion about shared problems is avoided due to
inadequate communication skills, which lead to poor results. The families that were
interviewed by the author spoke of having to leave their homes often between 5:30 and
6:00 in the morning in order to avoid the long lines that form for transportation out of the
neighborhood. Then after a full day at work, they battled in long lines and overcrowded
buses to return home again. Time at home in the evening was occupied in the daily
chores and in preparation for the next day of work. Therefore, it is not surprising that
members would avoid the more intense forms of communication and prefer to watch
television and relax in any free time that they may have. Additionally, housing layouts

often do not provide for private areas that would lend themselves to deeper forms of

42

communication. Participants in my focus groups conveyed that poorer families are
frequently limited to renting a single room in which all home interaction takes place.
Implementation of programs that help families learn communication skills must be
sensitive to the physical and time constraints on these families.

Third, there are few routine behavioral patterns related to the designation of
responsibilities within the household. Members show great ﬂexibility in performing
different roles and functions in order to maintain system stability. This is especially true
in regards to childcare and protection. However, Lodo-Platone was not speciﬁc if this
referred to males and females, or only females.

Fourth, the mother tends to be the authority ﬁgure in the household. However, due
to her time spent at work, any adult or older sibling may exercise control or supervisory
functions over the younger children. This often results in little consistency in discipline or
the expectation of appropriate behavior. The participants that I interviewed echoed this
tendency. One mother spoke of her 13 year-old daughter having male friends in her house
before she arrived from work. Another spoke of young children playing in the
neighborhood streets with no parental supervision. Again, there was an expression of “but
what can I do, I have to work, and there is no one to watch them.” In Venezuela, one
often hears the expression that a child is “hijo de papa y mama.” (son or daughter of
father and mother). The expression refers to a “good” child who was properly reared
under the supervision and love of a functioning parental system. Poor families have a
vision of what a positive environment for children entails, but they frequently struggle to

provide it given their ﬁnancial constraints and the context in which they live.

43

Fifth, affective relationships and reciprocal expression of feelings tend to be
implicit rather than explicit among poor families. Feelings of loyalty to each other were
found to be very important. I also found that expressions of love, concern, and respect are
paramount to these families. Children are taught from the earliest ages to “pedir la
bendicion” (ask for the blessing) from any adult relative whenever they enter or leave
their presence as a sign of respect. Not doing so in most families would be a message that
you are not family to me. The message is also exclusive to family members and would
not be extended to a neighbor even if she or he provided a care-giving function.
Participants related that the term “respect” is multifaceted and very important in
Venezuelan culture and is a strength that future programs that are designed to aide
families should incorporate if they hope to be successful.

A second prominent point that emerged from the focus groups was how the
inclination toward extended kin networks coupled with the topography of Caracas lead to
characteristics of neighborhood development. Participants spoke of neighborhoods that
started as “squatters” were often built from extended family networks. As one member of
the family managed to get his shanty built, he let his relatives know so that they would
come and put up their own shanties beside the original one, or enlarge it. Soon, several
small contiguous units would be built in which families would be in close proximity in
order to facilitate childcare and share resources. Numeric growth of the family would
trigger projects to expand the shanty to accommodate more family members. Men from
the family or emerging neighborhood would often band together to help each other build
their own ceramic-block houses. With the growth of the neighborhood, expansion of the

existing structures was limited to building upward. Amazingly, developed poor

44

neighborhoods in Caracas are full of three, four, and ﬁve storied houses built on 650 (or
greater) inclines by men with less than a high school education. Interestingly, Hernandez-
Ponce and Reimel, (2004) found that quality of life measures among Venezuelan poor
were positively related to home ownership, and the adequacy of the home to
accommodate family size. It seems that these values held by the Venezuelan poor (i.e.,
extended kin networks, and home ownership) have interacted to produce densely
populated sectors of low-income families that are mainly led by single mothers. As
neighborhoods grow, they also form what is called “Associacion de Vecinos” or
Neighborhood associations. These associations are recognized by Venezuelan law and
often determine the disposition of government provided resources (Hemandez-Ponce &
Reimel, 2004). There is usually no formal means to announce community meetings or
decisions besides word-of—mouth. Therefore, as the neighborhood grows, residents living
on the outskirts become increasingly disenfranchised and uninvolved in the decision-
making process at the community level.

With the growth of the neighborhood also comes delinquency. Pockets form in
the neighborhoods where drugs and prostitution are unhindered, and the community
cohesions begins to deteriorate. Participants spoke of how drugs are sold in some areas as
openly as one would purchase bread and milk. All ﬁve groups estimated that 65%-85%
of homes in poorer neighborhoods have a member that either consumes or sells illegal
drugs. This intersection between family dynamics and neighborhood qualities gives rise
to characteristics in the peer structures of the youth that inhabit them.

A third issue from the focus groups was the inﬂuence of peer groups.

Schoolteachers, parents, and psychotherapists underscored the important inﬂuence of

45

peers on Venezuelan youth. Schools play an important role in the development of peer
groups in Venezuela. Public schools are nested in neighborhoods and therefore often
reﬂect the neighborhood in which they exist. Teachers spoke of how dangerous schools
are becoming and how they often are afraid to evaluate students negatively for fear of
repercussions from a drug-dealing parent or perhaps even the youth themselves. Other
focus group members reported that youth in poorer neighborhoods are often recruited by
older drug-trafﬁckers because of their relative immunity before the law and their need for
income. Apparently these youth are promoted within the drug organization for merit
much like a military organization. After a certain “rank” youth are given “command”
over a certain number of other youth, and armaments in order to guard drug trafﬁcker’s
territory. Many families often turn a blind eye to this behavior because the income the
youth brings home is needed for their survival. Those youth who become involved in the
drug trade generally respect youth and adults who do not. However, this relationship can
be tenuous given the immaturity of gun-carrying youth. Still, the majority of violence
occurs between the different groups who cross territorial lines and those unfortunate ones
that get caught in the crossﬁre.

Venezuelan youth, despite their precarious situation, do not report neighborhood
violence as their primary stressor. A study of 2,121 youth from all social strata of Caracas
found that self-reported causes of stress for youth are: 1) Bad grades, 2) My mother
becomes sick, 3) Fights within my family, 4) Lies that people tell me, and 5) My brother
or sister or another member of my family are harmed by someone. It is interesting to note
that the youth, on average, did not list themselves being harmed in the ﬁrst ﬁve positions.

It was not until the 6th position that they expressed concern over their own personal well-

46

being (Davila & Guarino, 2001). Unfortunately, the study did not provide a demographic
breakdown for the sample to know what percentage of the participants came from poorer
neighborhoods, which may have elucidated an interesting interaction effect.

Finally, schools were also seen to play an important role in the development of
substance-abusing behavior. Both teachers and parents spoke of the lack of parental
involvement in children’s schooling. Teachers stated that attendance to parent-teacher
conferences was usually limited to about 2-5% of the parents. Parents expressed teachers’
unwillingness to schedule conferences at times that wouldn’t conﬂict with their work,
and that they already knew what they were going to say so “gpa’ que ir?” (why go?).
Parents are coerced to attend two annual meetings under threat of non-admittance into the
school at the beginning of the school year, or not releasing grades at the end of the school
year. Both teachers and parents described the relationship as being adversarial. Lodo-
Platone (2004) found similar attitudes on the part of parents she interviewed. She
described parental feelings of PTA meetings as “teachers’ scoldings that produce mutual
distrust.” (p. 81).

Lodo-Platone (2004) suggested a dropout rate of approximately 42% from the 1St
to the 6th grade. Unfortunately, she did not provide a citation to substantiate her claim.
The teachers that I interviewed also expressed concerns over high desertion rates, but
placed the highest rates of desertion occurring after 7’h grade. In Venezuela students
progress from elementary school into high school without the transition of middle school
as is common in the US. school system. According to teacher reports, there exists a
strong school attachment among elementary students, but this attachment wanes after the

transition into high school. They attribute this decline to the structure of high school

47

whose changing class schedule does not permit a bond to develop between the teacher
and the student. Also, there is a general lack of extracurricular activities that might allow
the adolescent to develop a bond or sense of belonging. Additionally, a majority of
schools in Venezuela hire teachers on an hourly basis to teach a given subject, which
creates a transitory impermanent culture within a school. Teachers do not form any bond
or loyalty to a school and are therefore less inclined to promote a sense of belonging and
stability among students. Likewise, students’ class schedules are often spotted with
inactive hours where they have no scheduled activities or classes to attend. During
“down” times students congregate in areas of the building and socialize, or leave the
premises to engage in other activities.

The teachers interviewed also reported a sharp increase in student drug use after
entrance into high school. If this results to be a general trend, it would seem that in the
absence of a secure attachment at home and the loss of attachment to the school, coupled
with low parental involvement with school, and low parental monitoring, and low school
structure, teens may be left to meet their emotional needs among peers. This scenario, has
repeatedly been shown to increase risk of substance-abusing behavior among teens in the
US. (Bauman & Ennett, 1994; Hawkins, Catalano, & Miller, 1992), and suggests a basis

for initial theory development for the onset of substance use among teens in Venezuela.

The Quantitative Case for Venezuela
Limited quantitative research has been conducted in Venezuela regarding
substance abuse in general and less yet in regards to adolescent use. One report by the

Comision Nacional Contra El Uso Ilicito de las Drogas [National Commission Against

48

the Illicit Use of Drugs] (CONACUID, 2006) examined several drug related behaviors of
patients in residential treatment programs across the nation of Venezuela. They found
that of the 6,374 patients 19.19% are below 20 years of age, 89.93% are male, 75.4% do
not have a high school education, 68.95% are single (never married), 57.17% are
unemployed, 12.49% are students, and 75.28% entered treatment voluntarily. These
patients initiated their substance abuse with Marijuana (36.57%), and Alcohol (30.67%),
and then Cocaine (15.05%), tobacco (9.54%), and crack and other drugs (8.06%).
Approximately eighty-six percent initiated drug use before their twentieth birthday with a
mean age of initiation for Marijuana 15.6, Alcohol 15.1, Cocaine 18.6, Crack 18, and
tobacco 13.9. The drug use for which they most frequently sought out treatment was
Crack 50.44%, Cocaine 20.03%, Marijuana 13.65%, and Alcohol 6.92%. While the
generalizability of this study is limited due to the clinical nature of the sample, the
indications are that initiation into substances start in early to mid adolescence, and that
Venezuelans begin their use somewhat differently than in the US. Studies of substance
initiation in the US. tend to support a gateway theory with 84.7% of the sample initiating
use with tobacco and alcohol and then progressing into marijuana and harder drugs
(Golub & Johnson, 2001). Also interesting is the high percentage of individuals seeking
treatment for crack and the apparent low occurrence of treatment for alcohol dependence
reported in the CONACUID study. Results from the 2004 National Survey on Drug Use
& Health (SAMHSA, 2005) state that of the 3.8 million persons who received treatment
in the US. for alcohol or drugs in the past year, more than half (2.4 million) received

treatment for alcohol use during their most recent treatment, 1.0 million persons (26%)

49

received treatment for marijuana, 884,000 persons (23%) for cocaine, 424,000 persons
(11%) for pain relievers, and 283,000 persons (7%) for heroin.

Another study attempted to identify risk factors for licit and illicit drug use in a
population of Venezuelan youth between the ages of 12 and 17 in Naguanagua, a small
urban population in the north central region of the country (Osorio, Ever, Ortega de
Medina, & Pillon, 2004). This study reported that family and mental health factors were
high-risk for drug use with severity scores of 80.41% and 63.67% respectively.
Recreation, behavior problems, and school adjustment were only moderate-risk with
severity scores of 48.98%, 46.73%, 39.39% respectively. Peers and social competencies
were found to be low-risk with severity scores of 31.63%, and 31.02% respectively. The
overall problem density for substance use was 3.67%.

The Osorio et a1, study used 8 of the 10 domains from the Drug Use Screening
Inventory (Kirisci, Mezzich, & Tarter, 1995) to assess for problem areas. The problem
density (severity) score for each domain on the Drug Use Screening Inventory (DUSI) is
obtained by dividing the number of yes endorsements by the number of items. The
resulting value, multiplied by 100, yields the problem density score that has a range from
0 to 100% in each of 10 domains. It can be seen, then, that problem density scores are
solely descriptive of responses to a given domain. Therefore, without further statistical
analysis it is impossible to infer risk for substance abuse from these ﬁndings.

This study had several other shortcomings that created difﬁculties. First, the
demographic information reported was limited to number, age, and gender of subjects.
Second, there was not a detailed explanation of the instrument used, and thus, how the

factors were operationalized for Venezuela. The authors report that the DUSI was

50

validated for Venezuela, but the reference they provide is not a published work. Third,
there was not a description of domain by drug, which might allow a comparison of how
the different factors were associated with drug abusing behavior. Fourth, the study gave
an indication of the overall severity of drug use, but not which drugs were being abused
by adolescents in that area of Venezuela. Fifth, this study was conducted in an area
outside the target population for the current study, which limits the generalizability of the
ﬁndings. Still the study provides an initial look into how these students perceive the
situations assessed by the domains of the DUSI.

A third study conducted in Venezuela looked at risk factors for the abuse of
alcohol among youth (Navarro & Pontillo, 2002). The study was conducted in the same
general location as the previously cited study, north central Venezuela. These authors
also used the DUSI to assess for alcohol abuse, however, in addition to the factors cited
in the previous study these authors also looked at self-esteem as a correlate to adolescent
drug abuse. The ﬁndings from this study varied signiﬁcantly from the previous one.
Problem density scores were social competency 72.2%, school adjustment 63.4%, peer
relationships 49.9 %, psychiatric/emotional 44.6%, and family 44.3%. Self-esteem was
found to be high, with 77% of the sample scoring at this level.

This study shares the same difﬁculties as those mentioned for the previous study,
with one important addition. The original sample was 500 adolescents. However, 199
(39.8%) were excluded from the study. One hundred and seventeen (117) were excluded
for their score on the DU SI lying sub-scale, and 82 were excluded for incomplete

demographic information or for missing data greater than 20%. The authors make no

51

attempt to explain how the exclusion of almost 40% of their sample might affect the
interpretation of their ﬁndings.

A fourth study conducted in Venezuela used an adapted version of the DUSI to
measure risk factors for substance abuse among students in the Department of
Architecture in the Universidad de Zulia (University of Zulia) in Maracaibo, Venezuela
(Gonzalez, 2005). The author reports that 55% of the students report using legal drugs
such as alcohol and tobacco and 2.5% use illicit drugs such as cannabis, cocaine, or
heroin. They state that 21.5% of the students use stimulants in order to stay awake to
study. However, which stimulants were used and their legal status was not stated. The
study indicated that 40.5% of the students stated that their best friend used drugs, but
again did not differentiate between legal and illegal.

Besides many of the previously mentioned difﬁculties, the sample used in this
population varies considerably from the population of interest for the proposed research.
In Venezuela, public education is theoretically free. That is, there is no matriculation fee,
but there is great demand and little supply. For this reason, typically only those people
willing or able to pay for their inside connection are admitted. Additionally, supplies that
one needs for school and normal living expenses are not provided such that for the poor
the struggle is uphill. Therefore, the sample used in this study represents an elite group of
young peOple with characteristics that vary widely from a typical adolescent in Caracas.

A study conducted in neighboring Colombia used an ecological approach to look
at frequency of marijuana among adolescents (Brook, Brook, De La Rosa, Duque,
Rodriguez, et al., 1998). These authors found support for the domains family, personality,

and peers having direct effect on adolescent marijuana use. Interestingly, they found that

52

the developmental path leading to drug use among Colombian youth is largely similar to
that found among White, African-American, and Puerto Rican adolescents living in the
United States. Still several cultural differences affecting adolescent substance between
the two countries were noted: exposure to violence showed a stronger association with
subsequent drug use among Colombian youth, there is greater drug availability in
Colombia, the impact of the peer group on the youngster's behavior in Colombia is more
pronounced than in the United States, and both familialismo and religion, had a stronger
protective function against drug use in Colombia than in the United States.

Estimates of adolescent tobacco use from the Global Youth Tobacco Survey
(GYTS) ranged a great deal among Latin American countries (Martin, & Peruga, 2002).
The highest estimates of the cumulative incidence of youth who had ever smoked tobacco
in Latin America were in Chile, and ranged from 68% in the city of Valparaiso to 72% in
the city of Santiago. In Uruguay, estimates ranged from 39% in the city of Colonia to as
high as 57% in the city of Montevideo. In the city of Buenos Aires, Argentina, 55% of
youth had sampled tobacco. In Peru, estimates ranged from 46% in the city of Trujillo to
5 5% in the city of Lima. In Bolivia, estimates ranged from 54% in the city of Santa Cruz
to 50% in the city of Cocha‘bamba. Lower estimates were found in Venezuela (22%) and
in Cuba (34%).

A cross-country comparison study of adolescent substance abuse in seven Central
American countries and the Dominican Republic, known as the PACARDO project, also
found considerable between country variability in patterns of drug use (Dormitzer, et al.,
2004). For example, the odds ratio estimate for alcohol use in the Dominican Republic

using Guatemala as a reference was 15.9. Additionally, these researchers found that

53

estimates of school-level clustering indicated that alcohol use clusters non-randomly

within schools in all of the PACARDO countries.

Conclusion

Ecological theory has been presented as a frame through which the interaction
between differing contexts can be tested to show how risk and protective factors might be
used as the basis of a program to deter adolescent substance abuse. The primary contexts
that have emerged from the literature are the family, peers, school, and neighborhood.
The current study explored risk and protective factors related to adolescent substance use
in Caracas, Venezuela. International studies such as the current one, stand poised to shed
light on important areas of interest in the struggle to create more healthy environments
for the development of future generations. Questions such as how poverty and other
variables affect families across cultures in relation to drug abuse, or what differences
government policy makes in families’ ability to protect their children from the onset of
drug abuse are important in an coo-developmental family therapy approach. Additionally,
discoveries regarding the risk and protective factors of adolescent substance abuse in
Venezuela constitute important advances for the citizens of Venezuela.

Little research has been conducted on drug abuse in Venezuela and less yet on
adolescent populations (personal communication with Elvia Rincon director of research
for the Oﬁcinal Nacional Antidroga — National Ofﬁce Against Drugs — July, 2006).
Currently, family therapists are being trained and are practicing their profession in

Venezuela. However, without serious research on the characteristics of the Venezuelan

54

population, therapists are left to adapt empirically the theories and techniques of their
trade (Feldman, 1989).

Research done in the US. and other countries suggest that family attentiveness,
externalizing behavior, peer relationships, and school environment are four primary
variables that might begin to explain considerable variability in age of onset of substance
use among adolescents in Caracas. However, the between-country variability among
Latin American countries found in the GTYS and PACARDO studies underscores the
importance of a solid research base that serves to identify prevalence of substance use,
clustering patterns by contexts or demographic variables, and risk and protective factors
of each country. Assuming a homogeneous population due to a common language or any
other single characteristic runs the risk of missing the mark and thus wasting precious
resources and time. More research is needed to determine the patterns of substance abuse
among Venezuelan youth. What drugs are used, the progression of use, contexts
surrounding use, demographic characteristics that vary with use, and interpersonal factors
associated with drug use are largely unknown, but essential to the identiﬁcation of risk
and protective factors that will inform the development of prevention and treatment

interventions.

55

CHAPTER III: METHODOLOGY

This chapter provides an in-depth description of the procedures employed in the
study. First, the setting of the study, data collection procedures including comments on
human subjects protocols, and a description of the participants are presented. Second,
conceptual and operational deﬁnitions of the variables studied are deﬁned. Third, an

overview of the data analytic plan and speciﬁc hypotheses to be tested are stated.

Setting

Caracas Venezuela is an urban metropolis on the northern coast of the South
American continent. The greater metropolitan area is densely populated with
approximately ﬁve million inhabitants. There are eight school districts in the
metropolitan area and approximately 3,000 schools. Approximately one-third (1,000) of
those are secondary schools. Caracas has developed such that pockets of poor and
afﬂuent neighborhoods are present within each school district. However, it is assumed
that some districts (i.e., those on the east side of the city) will have a higher concentration
of afﬂuence than others. Unlike US. public schools, public schools in Caracas are
populated almost exclusively with children from low to lower middle-class families and
some private schools are considered missions that reach out to the poorer segments of
society. For example, C olegios Fe y Alegria, (Faith and Happiness Schools) are Catholic
missions that are subsidized by the government and serve primarily the lower classes.
Caraquefios (people from the city of Caracas) are ethnically very heterogeneous with

little acknowledged racial discrimination. In this study, race will be deﬁned by skin tone

56

categorized roughly by dark, medium, light, and indigenous. Discrimination by social
class is much more prevalent and is acknowledged publicly. Socioeconomic status will be
determined by the caretaker’s education level, type of neighborhood of residence, number

of vehicles owned by immediate family, and number of bedrooms in the home.

Methods
Sampling Procedures

Given the exploratory nature of this study, the availability of funding, and the
purpose of the study being to collect pilot data, two school districts geographically
proximal from the western portion of the city of Caracas were selected from which to
draw a sample of 15 schools. First, schools were stratiﬁed in each district by grades
taught (i.e. 7th through 11’“), with only those containing the target population of children
ages 11-18 being selected. Next, schools were stratiﬁed by funding type (i.e., private or
public). Approximately 40% of schools in these districts are private. However, in order to
maintain comparable group sizes, private schools with average class sizes of less than 20
students per classroom were eliminated from the sampling frame. All public high schools
had classroom sizes in excess of 20 students per classroom and were, therefore, retained
in the sampling frame. In order to ensure that the sample reﬂects the population with
respect to the stratiﬁcation variable, a proportional allocation procedure of private to
public schools was performed. The procedure resulted in six private schools and nine
public schools to be randomly sampled from the pool of schools. During the data
collection phase of the study one public school was excluded due to logistical concerns.

However, given the late stage in the study that the school was dropped it was impossible

57

to replace the school leaving a total of six private schools and eight public schools. Each
school had multiple sections for each of the ﬁve grades. Therefore, the section to be
sampled was selected randomly from the pool of sections at each grade level. The total
population of students present the day of study within each classroom was sampled.
Since districts included in the sample were not randomly selected, a selection bias
may have been introduced into the sample. Caracas is a very diverse city, and Venezuela
is even more so with geographical and cultural variants that hold the potential to
inﬂuence participant responses. Moreover, the sample was also limited to school-
attending youth and cannot be generalized to those who have dropped out, or never
attended. Inferences about adolescent drug use are, therefore, considered generalizable

only to the school districts sampled.

Data Collection

Data collection at each school followed a protocol developed by the researcher in
collaboration with sources in Venezuela familiar with the education system in that
country. Administrative authorizations were sought and received the previous year (2006)
during a plausibility study. The ﬁrst step in protocol was a visit with principals of schools
to explain the study, present them with the appropriate authorizations from the Regional
Director and the District Superintendent (authorizations from the regional director and
district superintendents were obtained during the summer of 2006 during a feasibility
study funded by a pre-dissertation fellowship), ask for their participation, and select the
classrooms to be sampled. A second visit to the school was made to meet with each

teacher to explain to them the study and to leave with them the Parental Informed

58

Consent forms to be sent to the parents of each child in the classroom to be sampled.
Additionally, during the second visit the school principal was given ﬁve packets, each of
which contained a UCRIHS approved informed consent form and a copy of the MAMBI
to be ﬁlled out by administrators or teachers in the school and collected at the time of the
youth assessment.

A research team of ﬁve high school teachers (lead assessors) and ﬁve university
students (assistants) was recruited to assist in data collection. The members of the
research team participated in a 6-hour training session directed by the primary
investigator two days prior to the commencement of data collection.

Within each designated classroom, the lead assessor and assistant followed the
three-part assessment protocol designed to improve quality and accuracy of the study,
and to decrease missing values. For the ﬁrst part, the school principal accompanied the
assessors and their assistants to the classroom and introduced them to the teacher and
youths seated in the classroom. The school principal then left the classroom in charge of
the team’s lead assessor.

Within the classroom, the lead assessor’s ﬁrst tasks were to describe the survey
and to establish trust and rapport, prior to distribution questionnaires. This ﬁrst part of the
assessment protocol was structured in a manner that encouraged youth to voice concerns
about the anonymity of the study data; the idea was that these concerns should be made
public and discussed with resolution in the form of increased trust and rapport. The
accuracy and completeness of self-report youth survey data depend upon youth being
conﬁdent that their answers are anonymous. The lead assessor and the assistant

distributed a formal youth assent form, pre-scripted with IRB-approved sentences to elicit

59

assent. The youth assent form was read to youth pausing for questions or comments after
each section of the form. Next, youth were asked to return the IRB-approved Informed
Parental Consent forms that were designed both to inform the legal guardian of the child
regarding the study and to elicit the guardian’s consent for their child’s participation in
the study. Youth whose parents objected to their participation were identiﬁed.
Additionally, youth were asked if their parents had expressed a desire for them not to
participate, but had not returned the form. At this moment youth who were not
participating in the study because of parental objection, or simply because they chose not
to participate, left the classroom along with the teacher and assistant assessor. The
assistant assessor left with the teacher and children to help ensure the children made it to
the pre-designated area during the assessment and to provide the teacher with instructions
about completion of standardized ratings (described below); once the teacher started to
make these ratings, the assistant returned to aide the lead assessor in the classroom.

After working through issues of trust and rapport the second part of the protocol
involved the actual assessment. First, the lead assessor walked through the classroom and
distributed a stack of anonymous pre-printed questionnaire forms, a blank Scantron
answer sheet, and survey pencils. Youth were instructed not to put their names on any
part of the forms. Since many youth were not familiar with Scantron answer sheets, the
lead assessor instructed the youth to ﬁll out the name section of the form with the school
name. This served as a practice so that youth became familiar with the data collection
technique before actually responding to the survey. Next, youth were invited to pick
randomly from a small container a pre-printed 5—digit number. The ﬁrst two digits of the

number identiﬁed school. The next digit identiﬁed the classroom (i.e., grades 7-11). The

60

last two digits (01-60) distinguished between cases. Allowing the students to select
randomly a number that would identify their responses ensured that their anonymity was
safeguarded. The youth were then instructed to record their number on the Scantron
answer sheet in the area marked “PID” and to record it on the section ﬁve answer sheet.

The PACARDO-V was subdivided into ﬁve sections in order to create natural
breaks for students to rest, and to catch any mistakes and correct them before they
became too egregious. The lead assessor secured youth attention and began reading the
questions out loud. By reading the questions, the lead assessor was able to ﬁnd a pace
that was comfortable for the students. As the assessor read each question, the students
followed along and marked their own answers on the Scantron sheet. The assistant, after
giving instructions to the teacher returned to the classroom to help ensure that privacy
was respected and order was maintained (e. g., by providing quiet answers to idiosyncratic
questions from individual students). This approach was designed to overcome inter-
individual variations in literacy, and was intended to reduce what sometimes occurs as
‘racing’ to the end of a self-administered questionnaire and resultant marking errors.
Reading the questions at a rapid but comfortable pace also helped to increase privacy and
reduce disruptions by helping the students stay on task. The survey lasted approximately
55 minutes from start to ﬁnish.

Section ﬁve of the PACARDO-V assessed for age of ﬁrst opportunity for
substance use and for actual age of ﬁrst substance use. Given the nature of the responses
(i.e., numeric ages) and difﬁculty involved in recording these ages on the Scantron, youth
were instructed to record their responses to section ﬁve on the survey form. Youth

responses were later transferred from the section ﬁve form to the Scantron answer sheet

61

by the research team using a 100% veriﬁcation procedure (i.e. after being transferred
from one format to the next, each section ﬁve form and its respective Scantron was
checked again by a member of the research team for accuracy).

The construct domains relevant to this dissertation covered in the ll2-item
questionnaire are described in detail below along with their corresponding items. Section
1 of the questionnaire assessed for demographic data. Section 2 assessed for questions
regarding relationships with parents, friends, school, neighborhood, and general social
adaptation. Section 3 assessed for the youths perception of the relative risk (physical or
otherwise) of consuming drugs, and the degree perceived accessibility of different drugs.
Section 4 assessed the frequency in which youth are involved in different activities. It is
in section 4, item 81 that youth are ﬁrst asked directly about their own consumption of a
particular drug. This strategy was implemented with the expectation that youth who are
now familiar with the format of the questionnaire would be more apt to respond
truthfully. Section 5 assessed age of ﬁrst opportunity for substance use and age when a
substance was ﬁrst used.

The third part of the protocol was the closing session, during which the assessors
collected the completed questionnaires in a manner that reduced data collection errors
and helped promote a sense of anonymity. In speciﬁc, youth were asked to place their
Scantron answer sheet on top of the Section 5 sheet and hand them directly to the lead
assessor or the assistant as they passed through the room. While collecting the answer
sheets the assessors veriﬁed that the ID numbers were correctly ﬁlled out and were
present on both sheets before placing them in a large envelope. The assessors sealed and

packed away the envelope before engaging in closing exercises that included collecting

62

the PACARDO-V, expressing gratitude and hope that the youths would agree to
participate in future assessments of this type. Youths were given a ballpoint pen with the
logo of Michigan State University as a token of appreciation for their participation.
Additionally, a new laser printer was donated to the school as a sign of appreciation for

the participation of students, teachers, and administrators.

Data Processing and Quality Control
After data collection, all survey data from section 5 that reported on the youth age
at ﬁrst use of a substance were transferred to Scantron answer sheets with 100%
veriﬁcation. Each Scantron answer sheet was checked for accuracy, and cleaned of any
stray marks that might have inﬂuenced the precision of the scanning machines. The

answer sheets were then scanned into a database by the MSU scoring ofﬁce.

Human Subjects Protections
The researcher obtained permission from the Regional Director of the Federal

District of Caracas and the corresponding superintendents of the two school districts that
were sampled as well as a letter of collaboration from the principal of each of the schools
sampled. Additionally a letter of support was received by the Universidad Simon
Rodriguez (the only Venezuelan university that has a post graduate degree in substance
abuse). Permission to conduct the study was received from the Michigan State University
Committee on Research Involving Human Subjects (UCRIHS, IRB# 07-320). A copy of
the IRB approval letter along with approved consent forms has been provided in

Appendix D. A Copy of the letters from Venezuelan authorities may be found in

Appendix E with the exception of the letters from the principals of the 14 schools, in
which case only one example is provided.

The MSU IRB approved a waiver of parental consent protocol, which involved
contact with parents via a letter from the primary researcher sent home in the days prior
to the assessment session. This letter explained purposes and contents of the survey, and
requested the parent/ guardian to return the form expressing their desire in regards to their
child’s participation in the study. The parent/ guardian was also informed that in the event
that they did not return the form, their child would be allowed to participate if she/he
chose to do so, and that the school principal would act as an advocate for their child. A
total of 24 parents/ guardians requested that their child not participate in the study
representing approximately 1% of the parents contacted. A total 189 of the parents
returned the forms at all representing approximately 8.5% of the total population
contacted.

During the assessment session, even if a child’s guardian had consented to allow
them to participate, an active assent process was in place. That is, youths could decline to
answer any and all questions if they did not wish to participate, or they could mark a “no
response” option on the survey form. In actual practice, non-participation in this form
was atypical: no students left all items blank and only three students marked more than

50% of the survey item responses as “no response.”
Study Participants

A total of 1,831 students ages 11-19 were surveyed from 14 schools in two

districts from the western part of Caracas, Venezuela. Questions on the ﬁrst use of a fake

64

drug (Cadrina) were included in the PACARDO-V questionnaire. Among the 1,831
respondents, only 8 (0.4%) reported use of Cadrina. Under the assumption that mis-
statements about a fake drug may signal presence of falsely positive reports about other
drug experiences or general response errors in the questionnaires completed by these
participants, they were excluded from the study. Additionally, the three students that had
more than 50% missing data were excluded, leaving a total of 1,820 respondents. A total
of 960 respondents (52.5%) were female with 18 (1%) subjects not reporting gender.
Regarding race, the majority (58.8%) of the participants identiﬁed themselves as
“Morena” or brown (n=1074), 34.8% identiﬁed with “Blanca ” or white (n=636), 3.3%
identiﬁed with “Negra " or Black (n=60), and only 1.5% (n=27) identiﬁed with
“Indigena ” or Indigenous. Thirty individuals (1.6%) failed to respond to the item on
race.

Due to inadequate space on the response form, age was subdivided into ﬁve levels
and measured as a categorical variable. The ﬁrst age level was from ages 11 to 12 (5.9%,
n=107). The second age level was 13 to 14 (32.9%, n=601). The third age level was from
15 to 16 (39.8%, n=727). The fourth age level was 17 to 18 (19.9%, n=364). The ﬁfth
level was age 19 or above (1.1%, n=20). Only eight people (4%) failed to respond to the
item regarding age.

The sample was drawn from the ﬁve grades (7th — 11m) that make up high school
in Venezuela and is described in Table 3.1. Number of students per grade level seems to

be equally distributed across the sample.

65

Table 3.1 Grade Level in School

 

 

Valid Cumulative
Grade Frequency Percent Percent Percent
7th 340 18.7 18.7 18.7
8th 345 19.0 19.0 37.7
9th 379 20.8 20.8 58.5
10th 375 20.6 20.6 79.2
11th 379 20.8 20.8 100.0
Total 1818 99.9 100.0
Missing 2 .1
Total 1820 100.0

 

 

 

Teachers and school administrators were also asked to respond to an instrument
that surveyed impressions regarding the environment of the school. A total of 57 school
administrators or teachers responded to the 50-item questionnaire, or approximately four
instruments per school. No demographic information was collected for teachers or

administrators.

Measures
The PA CA RDO
The PACARDO (which stands for P_Anama, Qentral America, and Bepublica
mmincana) questionnaire was developed for use in a NIDA-funded grant “Cross-
National Research in Clusters of Drug Use” (Dormitzer, et al., 2004). In its original form
it is a standardized self-administered questionnaire and was administered to nationally
representative samples of students in Central America, Panama, and the Dominican

Republic (N = 12,797).

66

The original instrument has 224 items placed in 19 zones or modules. Initial
modules assessed general health constructs and social adaptation, such as are tapped by
questions about headaches, positive moods, and getting along with other youths. The ﬁrst
questions about afﬁliation with drug-involved peers appear in the eighth module, after 88
questions on other aspects of youth health and well-being. Questions about the youth’s
own drug involvement begin at PACARDO question 162 in the 15th module, which starts
out asking about legal consumption of alcoholic beverages, and tobacco. Subsequent
modules address illegal drug activities (e. g., marijuana, coca paste), prior to a concluding
20-item module modeled after Johanson’s Behavioral Repertoire Rating Scale (J ohanson,
Duffy, & Anthony, 1996) that assesses for frequencies of differing activities (e.g., going
to religious activities, doing housework). In order to protect against false positive reports
about drug experiences or general response errors in the questionnaires completed by the
participants, questions on the ﬁrst chance to try and ﬁrst use of a fake drug (Cadrina) are
included in the PACARDO questionnaire. The PACARDO instrument was pre-tested
prior to its use in all seven countries. Prior to item-metric and psychometric optimization,
the psychometric scales were ﬁrst analyzed at the aggregate level. Exploratory factor
analysis revealed that the scales generally were consistent across the seven countries.
Table 3.2 represents examples of the internal consistency and reliability coefﬁcients for

constructs of the PACARDO as recorded by Donnitzer (2004).

67

Table 3.2 Reliability Estimates of Psycho-Social Constructs in the PACARDO
(Dormitzer, 2004).

 

 

 

Estimated
Reliability No. of
Construct Name KR-20 Items Example Items (True-False response format)
, ~ oHas estado de ma] humor?
Irrrtable/ Crabby 0'72 8 ~ “Have you been in a bad mood?”
~ oDurante los ultimos 6 meses, ohas sentido
Positive Mental 0 71 9 bien?
Health ' ~ “During the last 6 months, have you felt very
happy?”
, , ~ oTe has sentido nervioso?
Mixed DISUCSS 0'71 9 ~ Have you felt nervous a lot?
~ oDurante el ultimo aﬁo, has herido o hecho
Extemalizing ~. dafio a los animales?
Behavior 0'83 19 ~ During the past year, have you harmed
animals?
~ oSiempre pides permiso a tus padres cuando
_ , sales de la casa a divertirte?
Famrly Attention 0'70 8 ~ Do you always ask your parents for
permission when you go out and have fun?
~ oTus amigos han robado, o han causado daﬁo
Deviant Peer a proposito a las cosas de otras personas?
Afﬁliation 0'80 8 ~ Have your friends stolen things or damaged

others’ property on purpose?

~ Algunos de mis amigos han fumado

Peers who use 0.77 6 marihuana.

drugs ~ Some of my friends have smoked marijuana.
~ oDurante los ultimo seis meses, tus notas
escolares han sido mejores que las de la
POOT SCROOI 0.78 20 mayoria de las de tus compaﬁeros de clase?
Adaptatron ~ During the past 6 months, Do you cut school
more than two days a month?
~ Hay suﬁcientes lugares seguros para carninar
Neighborhood a o jugar en mi barrio o vecindario.
Disadvantage 0'73 8 ~ There are plenty of safe places to walk or

spend time outdoors in my neighborhood.

 

68

 

The primary instrument employed in this study, the PACARDO-V (with the
addition of the V for Venezuela) was developed from the original PACARDO. Authors
of the PACARDO and researchers who used it in the ﬁeld reported that the 224 items
were excessive and that students were potentially answering the last sections of the
instrument without much thought due to fatigue (J. Anthony, C. Dormitzer, & P. Obando,
personal communication March, 2007). In order to avoid this problem in the present
study, items regarding general health issues (e. g., during the past 6 months, I have not felt
nauseated) and general mental health condition (e. g., Have you felt nervous a lot?) were
deleted. Other subscales were reduced using a conﬁrmatory factor analysis with
categorical dependent variables procedure on MPlus 4.1 software (Muthen & Muthen,
2006), the results of which are reported below in the description of each variable. Full
Information Maximum Likelihood (FIML) was used to estimate the parameters from data
with missing values. The ﬁnal version, PACARDO-V contains 112 items.

The original items of the PACARDO, and MAMBI, measures were subject to a
translation, back-translation, and harmonization process and were pilot tested within the
seven PACARDO countries (Dormitzer, et al., 2004). Items from the PACARDO were
modiﬁed in the PACARDO-V to reﬂect idiosyncrasies of the Venezuelan culture and
language use. For example, the term “pasta base” referring to coca base was change to
“bazuco” for Venezuelan participants. Likewise, in the MAMBI, items were modiﬁed for
increased accuracy and comprehension. The PACARDO-V and the MAMBI were then
pilot tested on Venezuelan adolescents and teachers in order to assess the face validity of
the instrument and to ensure cultural ﬁt and accuracy before their actual implementation

in the study. An iterative process was used to reﬁne the instruments. First, after the initial

69

changes were implemented, a small group of acquaintances of the author were asked to
read through the survey instruments and make comments or suggestions regarding the
readability of each item. Next, after those changes had been incorporated, each school
principal was asked to read through the PACARDO and make comments regarding the
readability of each item. Two of the private schools formed a committee comprised of
school psychologists, administrators, and teachers to assess the accuracy of the items.
Third, the author and a Venezuelan research assistant who is a school teacher and
psychologist evaluated each of the suggestions and made the appropriate changes to the
instruments.
Dependent Variables

Occurrence of ﬁrst drug use is the main response variable for this dissertation
research. Occurrence of ﬁrst drug use is measured in response to the standardized item,
“How old were you the ﬁrst time you tried (name of drug)?” for each drug in the study
(alcohol, tobacco, inhalants, prescription medication (not prescribed to the youth),
cocaine and any of its derivatives (i.e., crack, coca-base), ecstacy heroin, cadrina, and
marijuana. In the case of multiple drug use, the youngest age of ﬁrst use was recorded.
Age of ﬁrst drug use (AFU) is a continuous variable that ranged from 0 to 18 (0 = never
used). Focus groups conducted in Venezuela, and personal interviews with
administrators, teachers, psychotherapists, parents, and others revealed numerous
anecdotal pieces of evidence for early onset of substance use. For instance, one teacher
spoke of a 4-year-old who was given drugs to sell every day in order to have lunch
money. Other individuals spoke of a practice in eastern Venezuela of putting small

amounts of alcohol into male children’s bottles, or fathers allowing their sons to drink

70

from their beverages in order to initiate them into the ostensibly masculine trait of
drinking. Additionally, youth who reported extremely early ages of drug initiation
reported use of only one drug at this age, and did not show any other patterns of
falsifying or exaggerating their responses. Nevertheless, since youth reporting ﬁrst drug
use from ages 1-3 would most likely need to rely on a third person report due to memory
limitations of very young children (i.e., 1-3), the six cases that reported these ages were
considered outliers and were coded as missing values. Cases that reported ﬁrst use at age
4 and up were retained in the study. Table 3.3 records frequencies of the reported ages
and Figure 3.1 represents these graphically. Due to the high frequency of 0 responses
(i.e., never used), the distribution of AFU is bimodal. A bimodal distribution creates
difﬁculties for regression analysis violating the assumption of homogeneity of variance
and normality. Therefore, AFU was adjusted so that the 0 responses were excluded from
the analysis. This adjustment allowed for the assumptions of homogeneity and normality

to be reasonably met.

71

Table 3.3 Age of First Drug Use

 

 

 

 

Age of First Drug Use Adjusted Age of First Drug Use
Agea Frequency Percent Age Frequency Percent

0 293 16.10 0 0 O

4 4 0.22 4 4 0.26

5 13 0.71 5 13 0.85

6 14 0.77 6 14 0.92

7 19 1.04 7 19 1.24

8 52 2.86 8 52 3.41

9 61 3.35 9 61 3.99

10 185 10.16 10 185 12.12

11 150 8.24 11 150 9.82

12 286 15.71 12 286 18.73

13 274 15.05 13 274 17.94

14 224 12.31 14 224 14.67

15 168 9.23 15 168 11.00

16 41 2.25 16 41 2.69

17 21 1.15 17 21 1.38

18 2 0.11 18 2 0.13
Subtotal 1807 99.285 Subtotal 1514 99.149
Missing 13 0.71 Missing 13 0.8513

Total 1820 100 Total 1527 100.000

Mean 10.21 Mean 12.19
Median 12.00 Median 12.00

Mode 0.00 Mode 12.00

SD 4.96 SD 2.30

 

 

3 Age of 0 indicates never used a drug.

72

 

Frequency

too-l

 

 

 

 

 

 

 

 

 

l l
0.00 5.00 10.00
Age of ﬁrst use

Figure 3.1 Histogram of Age of First Use

73

l
15.00

I
20.00

Independent Variables

The independent variables used in the study were: family attention, externalizing
behavior, peer drug use, school climate, socioeconomic status, gender, and race. These
seven variables were chosen for the present study due to their salience in the literature as

signiﬁcant covariates of adolescent substance use.

Reliability estimates.

Convention in measurement typically requires a reliability of 0.70 or higher in
order to place conﬁdence in the results of a given scale. However, violations of the
assumptions underlying reliability estimates suggest caution when interpreting alpha.
Further, the appropriate degree of reliability is directly related to the intended use of an
instrument and its inherent dimensionality. For example, indexes, in contrast to scales,
are empirically derived composites of items that are purposefully selected to correlate to
some external criterion but not necessarily to each other (Reckase, 1996; Schmitt, 1996;
Streiner, 2003). Moreover, Streiner (2003) asserts that researchers violate the premise of
an index should they apply a reliability estimate that assumes interrelatedness (e. g.,
unidimensionality) among the items (e. g., coefﬁcient alpha).

Reliability differs from validity in that it does not assess what a test or set of items
attempts to measure, but only that something is being measured consistently. Cronbach’s
Alpha (coefﬁcient alpha) is a measure of the extent to which responses from a speciﬁc
sample of subjects are replicated or consistent across a set of test items, and is based on a
single administration of the measure or instrument. Alpha is, therefore, indicative of

interrelatedness, but not necessarily of homogeneity or unidimensionality of a construct

74

(Schmitt, 1996). As a result, an attempt to equate reliability to the degree to which a
given single construct has been measured is inappropriate.

F eldt and Charter (2003) recommend against using coefﬁcient alpha as a measure
of reliability for some types of scales, and demonstrate how violations of equality among
variances lead to biased estimates of alpha. They suggest that the reliability of both
parallel and tau equivalent scales can be adequately addressed with coefﬁcient alpha
estimates, but that the use of coefﬁcient alpha with congeneric scales will result in a
negative bias. Congeneric scales loosen the assumptions of classical test theory and do
not presume the equality of error variances in the measurement, nor of the scale of
measurement. Items (subsets) of a scale are allowed to make differential contributions to
the total-test true score. For example multidimensional scales, which are intended to
assess differing aspects of the same construct in a single scale or subscale, tend to result
in relatively lower alpha coefﬁcients for one or more the following reasons given by
Helms, Henze, Sass and Mifsud (2006): (a) unequal numbers of items reﬂecting the
various dimensions, (b) unequal item variances, and (c) clusters of respondents who do or
do not share similar attributes. In any of these circumstances, estimates of reliability
using alpha will be conservative.

In order to assess for whether the data ﬁt the previously described assumptions of
equality among variances, Feldt and Charter (2003), recommend the simple strategy of
examining the ratio of the largest item standard deviation (SD) to the smallest item
standard deviation (SDs). If the ratio (SDI/SDS) is between 1.00 and 1.30 (i.e., 30%),
then alpha will not be an excessively conservative estimate and would be an appropriate

reliability analysis. Likewise, they show that differences exceeding 30% indicate that the

75

data do not conform to the model of essentially tau equivalence on which alpha is based,
and the researcher should consider alternative analyses.

Following the above-mentioned recommendations, the covariates in the present
study are categorized as either indexes, or scales. In the event that the measure is an
index, no reliability is reported. Contrarily, if the measure is a scale the assumptions for
tau equivalence are assessed following the recommendations of Feldt and Charter (2003),
and the appropriate reliability estimates are provided. If the assumptions for equivalence
are met then Cronbach’s Alpha is provided. If the assumptions for tau equivalence seem
untenable, then the recommendations of F erketich (1990), who suggest the use of theta or
omega to estimate the reliability of item responses is followed. Omega employs the
following formula:

(2 =1—([k — Zhi] / [k + 2b]).
In the equation, k equals the number of items, h, is the commonality of the ith item, and b
is the sum of the correlations among the item responses comprising the scale.

Additionally, where appropriate, a conﬁrmatory factor analysis is provided to
assess the relationships among the items of the measures in relation to the latent construct
that they are assumed to measure.

Family attention.

Family Attention (FAM) in the originally PACARDO study was conceptualized
as an adaptation of the Capaldi and Dishion scale on parental monitoring (Capaldi &
Dishion, 1988). Although similar, FAM was expanded to include questions on affect and
communication between parents and the adolescent. Also, the concept of parental

monitoring was broadened to encompass other family members (e. g. grandmothers,

76

aunts, uncles) in the monitoring tasks in accordance with the extended family functions

common among the Latin family (Fussell & Palloni, 2004). The items for the FAM latent

trait are all yes/no responses. Family Attention is a level-1 covariate measured by the

following seven items from the PACARDO-V:

V14.

V15.

V16.

V17.

V18.

V20.

V25.

Are your parents or guardians aware of what you think or feel about things
that are important to you?

Are your parents or guardians aware of your likes and /dislikes?

I always ask my parents for permission when I go out and have fun.

Do you feel that your parents or guardians care about you?

Are your parents or guardians often aware of where you are and what you
are doing?

Sometimes young people come home after school and don’t ﬁnd anyone
home. Has your father, mother, or some other adult been home when you
returned from school or work during the last school year?

Do you frequently have discussions with your parents/guardian that end in

a shouting match?

Each item on the scale is scored as yes/no response (yes = 2) such that high scores

indicate increased FAM. For each observation, scores on the eight items were averaged

and then standardized to create a more readily interpretable factor composite (mean = 0,

SD = 1.0). Table 3.4 provides descriptive statistics for F AM and Figure 3.2 provides a

graphical representation of the frequency distribution.

77

Frequency

600 -

500 -

400 —

300 -

zoo-j

100-

Table 3.4 Descriptive Statistics of Family Attention

 

Mean

Std. Error of Mean
Median

Mode

Std. Deviation
Skewness

Std. Error of Skewness
Kurtosis

Std. Error of Kurtosis
Minimum

Maximum

 

F reguency
Valid 1 745

Missing 75
.0000000
.02393 879
.3409703
1.06622
1.0000000
-1 .055
.059

.860

.l 17
-4.01052
1.06622

 

 

 

 

.,_ r1

 

 

 

 

 

 

 

 

__.. 1 _
4 -3 -2 -1 o

 

 

 

Standardized F AM

Figure 3.2 Histogram of F AM Frequency

78

 

The test for tau equivalence (SDL/SDS = .499/ .238 = 2.0967) revealed that the assumption

for equality among the error variances was not met for FAM, and therefore the omega

method was used to assess for reliability. The obtained reliability coefﬁcient for the

present sample’s scores was .71. A conﬁrmatory factor analysis was employed to test for

the ﬁt of the items to the latent construct.

Conﬁrmatory factor analysis (CFA) results for F AM revealed a reasonably well

ﬁtting model. Even though the Chi Square statistic was signiﬁcant (x2 78.326, (if 19 p <

.001), this is not unusual for large sample sizes and is acceptable when corollary ﬁt

indexes are satisfactory (Kline, 2005). Bollen (1989) recommends using a normed chi-

square (xzmodel/dﬁnodel) for larger sample sizes, and advocates that values of 2.0 up to

5.0 indicate a reasonably well ﬁtting model. The normed chi-square for the present model

is 4.1 and falls within the realm suggested by Bollen. Additional ﬁt indexes such as the

RMSEA (the statistic least susceptible to sample size), the Comparative Fit Index, and

the Tucker-Lewis coefﬁcient demonstrated a well ﬁtting model (RMSEA .043, CFI .981,

and TLI .980 respectively). All factor loadings were highly signiﬁcant. Table 3.5 records

the factor loadings for the variables of interest.

Table 3.5 CFA Model Results

 

 

 

Latent V. Observed V. Estimates S.E. Est./S.E.
FAM BY V14 0.829 0.019 42.922
V15 0.768 0.023 33.492

V16 0.516 0.032 16.056

V17 0.854 0.025 34.507

V18 0.674 0.025 27.338

V20 0.215 0.038 5.675

V25 0.41 0.033 12.29

 

 

79

Peer drug use

Peer drug inﬂuence (PDRG) a level-1 covariate was measured by six items from

the PACARDO-V:

V30. Some of my friends smoke cigarettes.

V31. Many of my friends smoke cigarettes.

V33. Some of my friends have smoked marijuana.

V34. Have you had friends who like to sniff glue or gasoline?

V35. Some young people have started using coca base, crack, or cocaine. Do

you have a friend who has used coca base, crack, or cocaine?

V36. Do you have several friends who have used coca base, crack, or cocaine?
Each item on the scale was scored as yes/no response (yes = 2) such that higher scores
indicated increases in peer drug use. The scores for the six items were averaged for each
observation and then standardized (mean = 0, SD = 1.0) for interpretability. PDRG is
categorized as an index in that the measure is an empirically derived composite of items
intentionally selected to be related to the external criterion of potential peer drug
inﬂuence, but not necessarily to each other. Table 3.6 provides descriptive statistics for

PDRG, and Figure 3.3 provides a graph of the frequency distribution.

80

Frequency

Table 3.6 Descriptive Statistics of PRDG

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency
Valid 1783
Missing 37
Mean .0000000
Std. Error of Mean .02368232
Median .3428144
Mode 1.06285
Std. Deviation 1.00000000
Skewness -.870
Std. Error of Skewness .058
Kurtosis .336
Std. Error of Kurtosis .116
Minimum -3.25736
Maximum 1.06285
600—
F
500—
400— [- F
300—
200— F
100—
I- ’/'ﬁﬂ\_
0 I I
-2.00000 0.00000
Standardized PDRG

Figure 3.3 Histogram of PDRG Frequency

81

 

Externalizing behavior

Extemalizing behavior (EXTB) is a level-1 covariate, which on the original
PACARDO was adapted from the Drug Use Screening Inventory (Tarter & Hegedus,
1991) for use in research on non-clinical samples. The items for the EXTB latent trait are
all yes/no responses (yes = 2) such that higher scores indicate increases in EXTB.
Individual scores for the ﬁve items were averaged and standardized (mean = 0, SD = 1.0)
for interpretability. Tables 3.7 provide descriptive statistics for EXTB, and Figure 3.4

provides a histogram of the frequency distribution for EXTB.

Table 3.7 Descriptive Statistics of EXTB

 

 

 

Frequency

Valid 1803

Missing 17
Mean .0000000
Std. Error of Mean .02355061
Median .2351980
Mode .23520
Std. Deviation 1.00000000
Skewness -.911
Std. Error of Skewness .058
Kurtosis .492
Std. Error of Kurtosis .115
Minimum -3. 19850
Maximum 1.09362

 

82

Frequency

 

500—1

400 —

300-

200 -

100—

./

 

 

 

 

’-
a '\h

 

 

 

 

D

 

U 7 _

 

I

0.00000
Standardized EXTB

I
2.00000

Figure 3.4 Histogram of EXTB Frequency

83

The following items from the PACARDO-V measure EXTB:

V40.

V41 .
V42.
V43.

V48.
The test for tau equivalence (SDL/SDS = .495/ .323 = 1.532) revealed that the

assumption for equality among the error variances was not met for EXTB, and therefore

the omega method was used to estimate a measure of reliability. The obtained reliability

Have you intentionally damaged another person’s belongings during the

last school year?

Have you stolen anything during the last school year?

Have you done anything risky or dangerous during the last school year?

Is it true that the majority of the time you don’t do your homework?

Have you ever been suspended from school?

coefﬁcient for the present sample’s scores was .63.

A conﬁrmatory factor analysis was employed to test for the ﬁt of the items to the
latent construct. CFA results for EXTB revealed an excellent ﬁtting model (x2 5.345, df 5
p< .3753). Additional ﬁt indexes also suggested an excellent ﬁt RMSEA (.006), CFI

(.999), and TLI (.999). All factor loadings were highly signiﬁcant. Table 3.8 records the

factor loadings for EXTB.

Table 3.8 CF A Results for EXTB

 

Latent V. Observed V. Estimates S.E. Est./S.E.

 

EXTB V40
V41
V42
V43
V48

 

0.68 0.046
0.681 0.046
0.581 0.042
0.406 0.044
0.555 0.05

14.877
14.714
13.854

9.281
11.192

 

84

 

School Climate

School Climate (SCLM) is a level-1 covariate, which on the original PACARDO
was adapted from the Drug Use Screening Inventory (Tarter & Hegedus, 1991) for use in
research on non-clinical samples. The items for the SCLM latent trait are all yes/no
responses (yes = 2) such that higher scores indicate decreases in SCLM. Individual scores
for the four items were averaged and standardized (mean = 0, SD = 1.0) for
interpretability. The test for tau equivalence (SDL/SDS = .441/ .216 = 2.042) revealed that
the assumption for equality among the error variances was not met for SCLM, and
therefore the omega method was used to estimate a measure of reliability. The obtained
reliability coefﬁcient for the present sample’s scores was .57. Table 3.9 provides
descriptive statistics for SCLM and Figure 3.5 provides a histogram of the frequency

distribution for SCLM.

Table 3.9 Descriptive Statistics for SCLM

 

 

Freguency

Valid 1797

Missing 23
Mean .0000000
Std. Error of Mean .02358989
Median -.7251664
Mode -.72517
Std. Deviation 1.00000000
Skewness 1.365
Std. Error of Skewness .058
Kurtosis 1.440
Std. Error of Kurtosis .115
Minimum -.72517
Maximum 3.95391

 

 

85

Frequency

 

700 -

600 -'

500 -

400 —

300-

200-

100-

 

0...
4.00000

-3.00000

ﬂ

 

fﬁ\
\

 

 

 

 

l'l

 

l I
-2.00000 -1.00000

 

Dr

0.00000 1.00000

Standardized SCLM

Figure 3.5 Histogram of SCLM Frequency

86

The following items from the PACARDO-V measure SCLM:

V44. I have had excellent relations with the majority of my teachers.

V50. Some young people feel happy when they think of going to school. In

overall, have you felt happy when you think of going to school?

V51. I have thought about quitting school altogether?
V52. Sometimes young people say, “going to school is a waste of time.” For
you, has going to school been a waste of time during this last year?

A conﬁrmatory factor analysis was employed to test for the ﬁt of the items to the
latent construct. CFA results for SCLM revealed a good ﬁtting model (12 6.101, df 2 p<
.0473). Additional ﬁt indexes also suggested a good ﬁt RMSEA (.035), CFI (.983), and
TLI (.956). All factor loadings were highly signiﬁcant as can be seen on Table 3.10,

which records the factor loadings for SCLM.

Table 3.10 CFA Results for SCLM

 

 

Latent V. Observed V. Estimates S.E. Est/SE.
EXTB V44 0.51 0.051 9.981
V50 0.692 0.057 12.06

V51 0.639 0.058 10.991

V52 0.52 0.068 7.661

 

 

 

Socioeconomic status.

SES is a level-1 covariate and was measured by ﬁve items from the

PACARDO-V:
V6. What type of neighborhood do you live (ordinal variable scored 1-3).
V7. How many vehicles does your family have (ordinal variable scored 1-5)?
V9. How many bedrooms does your house have (ordinal variable scored 1-5)?

87

V12. What academic grade did your father (or the person who is like your
father) achieve (ordinal variable scored 1-5)?

V13. What academic grade did your mother (or the person who is like your
father) achieve (ordinal variable scored 1-5)?

The items that make up the SES scale were measured on a Likert type scale and
scored by summing across the ﬁve items for each observation. The composite created
from the sum was then standardized (mean = 0, SD = 1.0) for interpretability with a
positive score indicating above average SES. SES is categorized as an index in that the
measure is an empirically derived composite of items intentionally selected to be related
to the external criterion of socio-economic status, but not necessarily to each other. Table
3.11 provides descriptive statistics for SES and Figure 3.6 provides a histogram of the

frequency distribution for SES.

Table 3.11 Descriptive Statistics for SES

 

Frequency

Valid 1778

Missing 42
Mean .0000000
Std. Error of Mean .02371560
Median —.0473766
Mode -.04738
Std. Deviation 1.00000000
Skewness .115
Std. Error of Skewness .058
Kurtosis -.817
Std. Error of Kurtosis .116
Minimum -2.27878
Maximum 2.74188

 

 

 

88

Frequency

 

200 —

150—

100-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

F l 1
-2.00000 -1.00000 0.00000 1.00000

Standardized SES

1
2.00000

Figure 3.6 Histogram of SES Frequency

89

 

The .MAMBI

The MAMBI (Which stands for Guia de Observacion Medio Meme del
Salon, Colegio y Vecindario, or Observational Guide for the Classroom, School, and
Neighborhood Environment) was developed for use in a NIDA-funded grant “Cross-
National Research in Clusters of Drug Use” (Dormitzer, et al., 2004) and is an
observational guide to be ﬁlled out by administrators and teachers. The purpose of the
MAMBI is to assess for the environmental conditions in which the children are studying
(Are there enough desks and chairs for each student to have one? Is there barbed wire or
broken glass on the top of the walls that surround the school?) The MAMBI is an index
comprised of 40 items with a dichotomous response set. Each item was summed to create
a factor composite indicating the extent the school possessed a favorable environmental
condition. High values indicate less favorable conditions. No published studies have
reported on the validity of the MAMBI. The MAMBI is categorized as an index in that
the measure is an empirically derived composite of items intentionally selected to be
related to the external criterion of the school environment, but not necessarily to each
other. Table 3.12 provides descriptive statistics for the MAMBI and Figure 3.7 provides a

histogram of the frequency distribution for the MAMBI.

90

Frequency

Table 3.12 Descriptive Statistics for MAMBI

 

 

 

 

 

 

 

 

 

 

 

 

Frequency
Valid 39
Missing 14
Mean 60.1026
Std. Error of Mean .58524
Median 60.0000
Mode 59.00
Std. Deviation 3.65481
Skewness .496
Std. Error of Skewness .378
Kurtosis 1.455
Std. Error of Kurtosis .741
Minimum 53.00
Maximum 71.00
I—
6-
.- / it
2_ _
U 40.00 45l00 50.00 55i00 60.00

MAMBI

91

Figure 3.7. Histogram of the MAMBI

 

Data Analytic Plan
Modeling Approaches

The ﬁrst data exploration step involved descriptive analyses to characterize the
sample, examine the data for systematic patterns in missing values, and to assess the ﬁrst
initiation of use of all drugs (i.e., alcohol, tobacco, inhalants, cocaine, prescription pills,
ecstasy, heroin, and marij uana). Second, contingency table analyses and ANOVA were
used to further explore if drug use varied by individual characteristics.

Students were not selected randomly across a sampling frame of students. Rather
districts were selected, then schools, and ﬁnally students. Also given that schools often
are a homogenizing factor in the lives of youth, the non-independence of observations
must be accounted for in the statistical modeling approach. Therefore, the third step of
this analysis used a multilevel modeling approach that allowed for the control of the
variation in the outcome that may be attributable to the environments in which the
students interacted. To model the dependent variable, AF U, HLM 6.02a (Raudenbush,
Bryk, & Congdon, 2004), software was used under a general modeling strategy that
moved from modeling the level-1 variance to the level-2 variance adding covariates to
the model according to their theoretical importance.

The unconditional model or Null model (i.e., no explanatory variables) was
developed in order to gauge the degree of variability between schools in drug use. The
unconditional model established baseline effects for the coefﬁcients and the variance
components in order to ascertain the aggregate variance that might be explained by later

models.

92

Level-l conditional models (models 1-6) introduced sequentially the following
covariates FAM, EXTB, PDRG, SCLM, SES, and FEM (gender). Model-7 introduced a
random coefﬁcients model and determined whether the level-1 slopes should be ﬁxed,
allowed to vary randomly, or allowed to vary non-randomly. Model-8 introduced level-2

covariates to model the variance in the intercept and regression coefﬁcients.

Research Questions and Hypotheses
The specific research questions and their analytic procedures are as follows:
1. What percentage of youth used each of the following drugs: tobacco, alcohol,
marijuana, cocaine, crack, heroin, amphetamines, inhalants, ecstasy, or prescription?

Data analysis for question 1: Descriptive.
2. Did age of ﬁrst drug use vary by individual variables?
2.1. Did age of ﬁrst actual drug use vary by gender?
Hypothesis: The age of ﬁrst drug use will vary by gender.
2.2. Did age of ﬁrst actual drug use vary by race?
Hypothesis: The age of first drug use will vary by race.
2.3. Did age of ﬁrst drug use vary by SES?
Hypothesis: The age of ﬁrst drug use will vary by SES.
2.4. Did age of ﬁrst drug use vary by family attention?
Hypothesis: The age of ﬁrst drug use will vary by family attention.

2.5. Did age of ﬁrst actual drug use vary by externalizing behavior?

93

Hypothesis: The age of ﬁrst drug use will vary by externalizing behavior.
2.6. Did age of ﬁrst drug use vary by peer drug use?
Hypothesis: The age of first drug use will vary by peer drug use.
2.7. Did age of ﬁrst actual drug use vary by school climate?
Hypothesis: The age of ﬁrst drug use will vary by school climate.
Data analysis for questions 2.1- 2.7: Hierarchical Linear Modeling (HLM)
Were school characteristic related to the onset of drug use?
3.1. Did School Condition help to explain the variance in age of ﬁrst drug use?
Hypothesis: School Condition will be related to age of ﬁrst drug use.
3.2. Did Mean SES help to explain the variance in age of ﬁrst drug use?
Hypothesis: Mean SES will be related to age of ﬁrst drug use.
3.3. Did Mean School Climate help to explain the variance in age of ﬁrst drug use?
Hypothesis: Mean School Climate is related to age of ﬁrst drug use.

Data analysis for question 3.1-3.3: HLM.

94

CHAPTER IV: RESULTS
Descriptive Statistics

This chapter presents the results of the data analysis for age of ﬁrst drug use
(AF U) in regards to the speciﬁc hypotheses set forth in previous sections. This study
sought to shed light on different risk and protective factors that play a role in the
initiation of substance use among school-attending youth in Caracas, Venezuela.
Fourteen schools were surveyed, six of which were private institutions and eight were
public. Of the 1,820 students included in the analysis, 847 (46.5%) were from private
schools. High schools in Venezuela are made up of a total of ﬁve grades, 7th-11th. The
sample was equally distributed among the ﬁve grades with n=340 in 7th, n=345 in 8th,
n=379 in 9’“, n=3 75 in 10th, and 379 in 1 1m. The majority of participants lived in the
lowest housing area (n=1007, 55.7%), did not own a vehicle (n=723, 39.7%), lived in a
home with 2-3 bedrooms (n=11 16, 61.3%), and had 4-6 people living in their home
(n=1104, 60.9%). The majority of respondents reported educational levels of the father
and mother as having ﬁnished a post high school degree (n=542, 30.2% and n=543,
30.1% respectively). Only 35.5% of fathers and 35.3% of mothers were reported as not
having ﬁnished high school. A total of 987 (54.2%) participants reported belonging to the
Catholic religion, 322 (17.7%) reported belonging to a non-Catholic Christian religion, 4
(.2%) students reported being Muslim, and 110 (6.1%) students reported belonging to
some other religion. A total of 387 (21.3%) students reported belonging to no religion

(see Table 4.1 for a complete description of demographics).

95

Table 4.1 Demographics of Sample

 

 

 

 

 

 

 

Grade N %
7th 340 0.187
8th 345 0.19
9th 379 0.208
10th 376 0.207
11th 380 0.209
Total 1820 100
Housing
low-income 1007 0.553
Middle-income 761 0.418
Upper-income 41 0.023
Missing 1 1 0.006
Total 1820 100
Vehicles owned
None 723 0.397
One 611 0.336
Two 242 0.133
Three 92 0.051
Four or more 141 0.077
Missing 1 1 0.006
Total 1820 100
Number of bedrooms in
home
None (one room) 45 0.025
1 181 0.099
2-3 1116 0.613
4-5 378 0.208
6 or more 89 0.049
Missing 1 1 0.006
Total 1820 100
Number of people living
in home
1-3 327 0.18
4-6 1109 0.609
7-8 214 0.118
9-10 90 0.049
More than 10 73 0.04
Missing 7 0.004
Total 1820 100

96

 

Table 4.1 Demographics of Sample (Cont)

 

 

 

 

 

 

Religion

Catholic 987 0.5423

C hristian/not Catholic 322 0.1769

Muslim 4 0.0022

Other 1 10 0.0604

None 387 0.2126

Missing 10 0.0055

Total 1820 100

Parent Educational Level Father Mother
N % N %

Some Elementary Edu 239 0.1313 249 0.1368
Some Secondary Edu 399 0.2192 391 0.2148
Finished Secondary Edu 419 0.2302 416 0.2286
Some Higher Edu 198 0.1088 209 0.1148
Finished Higher Edu 547 0.3005 545 0.2995
Missing 18 0.0099 10 0.0055
Total 1820 100 1820 100

 

Overall, the sample was comprised of a high number of students that had
consumed alcohol (81.3%) and cigarettes (31.5%) on at least one occasion (see table 4.2
for details). Approximately 48% of the sample had used at least one drug, 28% had used
two drugs, and 7.5% had used three or more drugs (see table 4.3). As a result of the
higher rates of alcohol and cigarette consumption in the sample, a distinction was made
between legal and illegal drug use. While not technically legal for the majority of the
youth in the sample (legal age to purchase alcohol and cigarettes in Venezuela is 18)
alcohol and cigarette use are culturally sanctioned as evidenced by the fact that there is
virtually no police action taken against underage youth who consume alcohol or
cigarettes, nor against store owners who sell these drugs to them. However, codes for

drugs such as morphine or diazepam that require a prescription, or those that are

97

technically illegal are more readily enforced. Following this distinction, approximately
13% of the sample had consumed an illegal drug and 2.3% had consumed multiple illegal
drugs (see table 4.2). The average age of initiating drug use was 12.19 (SD=2.67) for any

drug and 12.52 (SD=2.68) for illegal substances (see table 4.4).

Table 4.2 Frequencies of Students Reporting Having Initiated Drug Use

 

 

 

 

Used (%) Never Used (%) Missing (%) Total (%)
Cigarette 572 31.5 1240 68.1 8 0.4 1820 100
Alcohol 1477 81.3 334 18.4 9 0.5 1820 100
Cocaine 5 0.3 1802 99.4 13 0.71820 100
Ecstasy 17 0.9 1790 98.7 13 0.3 1820 100
Inhalants 45 2.5 1762 97.2 13 0.3 1820 100
Heroin 8 0.4 1800 99.3 12 0.3 1820 100
Prescription 160 8.8 1644 90.7 16 0.91820 100
Marijuana 68 3.8 1719 94.5 33 1.81820 100
Any Druga 1513 83.5 287 15.8 12 0.7 1820 100
Illegal Drugsb 240 13.2 1521 83.9 51 2.81820 100
Multiple lllegalb 42 2.3 1715 94.6 55 3.01820 100

 

3Refers to at least one of the drugs listed on the PACARDO-V
b
Refers to any drug from the PA CA RDO-V except alcohol and cigarettes.

Table 4.3 Number of Different Drugs Consumed

 

 

 

 

# Drugs N %
0 287 15.77
1 873 47.97
2 509 27.97
3 101 5.55
4 25 1.37
5 7 0.38
6 2 0.11
7 1 0.05
Total 1805 99.18
Missing 15 0.82
Total 1820 100

 

98

 

Table 4.4 Ages of First Initiation of Substance Use
N Mean Std. Deviation Min. Age Max. Age

Illegal Drugs8 240 12.52083 2.679455 5 19
Any Drugb 151312.19167 2.301141 4 18
21Refers to at least one of the drugs listed on the PA CA RDO—V
bRefers to any drug from the PA CA RDO-V except alcohol and cigarettes.

 

 

 

 

 

Preliminary Analyses
Age of First Drug Use
Gender, race, SES, family attention, externalizing behavior, peer drug use, and

school climate were examined for main effects on the dependent variable of age of ﬁrst
drug use in order to inform later model building. To assess the effect of ordinal and
binary variables (i.e., gender, and race) on age of ﬁrst drug use a series of ANOVAs were
performed. To assess the effects of the continuous variables (i.e., SES, family attention,
externalizing behavior, peer drug use, and school climate) on the dependent variable an
OLS regression was performed with a simultaneous entry method. Simultaneous
regression is useful in exploratory research to determine the relative inﬂuence of each of
the variables studied, since it estimates the direct effects of each independent variable on
the dependent variable. It is important to note here that these ﬁndings are only
exploratory since they do not account for dependencies within the observations due to
nesting. A basic assumption of both ANOVA and regression analysis is the independence
of observations. Since the sample of students was not drawn randomly, and is nested

within schools, this assumption is violated and thus increases the tendency toward Type I

99

errors. Still, the ﬁndings are useful insofar as they provide information for subsequent

model building.

Gender
An examination of how gender inﬂuences initiation into drug use revealed
signiﬁcant differences between males and females in mean age of ﬁrst drug use. The
mean age for females (12.4 years) was found to differ signiﬁcantly from the mean age of
males (11.9 years) at F (1, 1498) = 13.818,p < .001. These ﬁndings suggest that males
and females do differ in regards to age of ﬁrst drug use, and that gender should be

included in subsequent models to control for this variability.

Race
No signiﬁcant differences for age of ﬁrst drug use were found for race (F = 1.235,
df 1 , 1496, p = .295). However, given the unequal groups for the black and indigenous
categories the four groups were re-categorized by skin tone (i.e., light and dark) creating
a dichotomous variable (light = 1) and the analyses rerun. Again, no differences were

found on the response variable by race.

SES, Family Attention, Externalizing Behavior, Peer Drug, and School Climate
Age of ﬁrst drug use was regressed on SES, family attention, externalizing
behavior, peer drug, and school climate in order to assess for potential main effects. The
results indicated that the ﬁve variables had signiﬁcant regression coefﬁcients (see table

4.5). However, the proportion of explained variance in age of ﬁrst drug use is a relatively

100

small 6% (adjusted R2 = .05 8). Additionally, diagnostics were inspected to assess for
collinearity. The variance inﬂation statistics (VIF) and the tolerance of variables were all
close to 1, which would indicate independence. These results suggest that all ﬁve

variables should be considered for subsequent modeling.

Table 4.5 Regression: Age First Use

 

 

 

Unstandardized Standardized

Coefﬁcients Coefﬁcients Collinearity Statistics

B S.E. Beta t Sig. Tolerance VIF
(Constant) 12.1029 0.0887 136.379 0
SES -0.2694 0.0603 -0.1180 -4.4672 0.0000 0.9779 1.0226
FAM 0.2641 0.0647 0.1181 4.0810 0.0000 0.8154 1.2264
SCLM 0.1473 0.0633 0.0659 2.3254 0.0202 0.8489 1.1780
PDRG -0.1478 0.0648 -0.0651 -2.2811 0.0227 0.8384 1.1928
EXTB -0.2254 0.0698 -0.0980 -3.2309 0.0013 0.7418 1.3481
FEMALE 0.3603 0.1224 0.0790 2.9427 0.0033 0.9474 1.0555

 

 

 

Correlations were run to examine the strength of the relationship between the
level-1 covariates and age of ﬁrst drug use, and to assess again for potential confounds
caused by multicollinearity. Results showed Pearson correlation coefﬁcients for family
attention (.126), externalizing behavior (-.151), school climate (.125), and SES (—.147),
were signiﬁcant at the p < 0.01 level (2-tailed) and Spearman’s Rho for female (.095)
was signiﬁcant at p < .001. The only covariate that was not signiﬁcant was peer drug
inﬂuence (.022). The correlations between covariates did not raise any concern for
collinearity (see Table 4.6). The highest correlation was between externalizing behavior
and peer drug use (-.382).

Correlations between level-2 covariates and age of ﬁrst drug use were also

calculated to assess their relative strength and to assess for potential collinearity (see

101

Table 4.7). Results found that all the covariates were signiﬁcantly correlated with age of

ﬁrst drug use at the p < 0.01 level (2-tailed). MSES had the strongest relationship (-.248)

followed by MSCLM (-.212), MCOND (.207), and MPDRG (.084). The correlations

between the covariates suggested that MSES and MSCLM were highly related (.810), as

were MSES with MCOND (-.700).

Table 4.6 Correlations Level-1 Covariates

 

SES 0.025 0.024 .054** -.075** 1

 

FAM EXTB PDRG SCLM SES

FAM 1 -.355** .318** .268** 0.025
EXTB -.355** 1 -.382** -.349** 0.024
PDRG .318** -.382** 1 .207** .054**
SCLM .268** -.349** .207** 1 -.075**

FEM -.095** -.127** 0.03 .056** _.124**

EM
-.070**
-.127**
0.03
.056**
-.124**
1

 

 

“Correlation is signiﬁcant at the 0.01 level.
*Correlation is signiﬁcant at the 0.05 level.

FAM = family attention; EXTB = externalizing behavior; PDRG =
peer drug use; SCLM = school climate; SES = socioeconomic status;

FEM = female.

Table 4.7 Corrleations Level-2 Covariates

 

AFU 1 -.248** .084** -.212**
MSES -.248** 1 -.293** .810**
MPDRG .084** -.293** 1 -.163**
MSCLM -.212** .810** -.163** 1

 

MCOND .207** -.700** .373** -.577**

ﬂ MSES MPDRG MSCLM MCOND
.207**
-.700**
.373**
-.577**

 

l

 

**Correlation is signiﬁcant at the 0.01 level.
* Correlation is signiﬁcant at the 0.01 level.

AF U = age of ﬁrst drug use; MSES = mean socioeconomic

status; MPDRG = mean peer drug use; MSCLM = mean school

climate; MCOND = mean school condition.

102

Missing Data

As was mentioned in chapter 3, eleven students were removed from the data set
for missing data or marking the non-response option. Eleven students declined to assent
to participate and the parent/ guardian of twenty-four students requested that their students
not participate, and were allowed to leave before data collection began. Thus, the survey
participation was approximately 98% of the targeted sample of school-attending youths
(i.e., l — [Total non-participation/(total nonparticipation + Total valid observations)] =
percent participation, or 1 — [46/(46+1820)] = .9753). Student-level non-participation also
might have occurred if parents instructed their children to stay home from school on the
day of the assessment, or if students chose not to come to school in order to avoid
participation. This practice, if it occurred, would have been minimal since the exact day
of the assessment was not announced to either the parents or the students, and teachers
reported normal rates of absenteeism.

Values were considered missing under the following conditions: (a) if there was
no response (left blank), (b) if the “no response” option was marked, or (c) if more than
one response was marked for any given item. Missing values were very low ranging from
0 (0%) to a maximum of 125 (6.8%) for any given variable. The highest rate of
missingness was for the demographic variable Religion (11 = 125, 6.8%). There was a
mean of 38.28 (2.01%) missing values (SD = 29.33) across the entire data set. As a result
of low levels of missingness no systematic patterns were detectable, and missing values
were considered missing completely at random and treated using listwise deletion when

they occurred.

In conclusion, preliminary analyses are useful to determine the relative inﬂuence
of each of the independent variables on the dependent variable in order to guide
subsequent modeling. In multilevel modeling omitting a relevant independent variable
may lead to model misspeciﬁcation, while including an excessive amount of explanatory
variables may create instability in the model. Instability means that small changes to the
model may lead to large changes in the results due to, for instance, multicollinearity
(Kreft & de Leeuw, 1998; Raudenbush & Bryk, 2002). These preliminary analyses
suggest the inclusion of gender, SES, family attention, externalizing behavior, peer drug
use, and school climate as level-1 covariates, and MSES, MSCLM, MCOND, MPDRG
as level-2 covariates in the development of a multilevel model of age of ﬁrst drug use.

The next section will address the development of these models.

Multilevel Models
Age of First Drug Use: The Null Model

Several models were ﬁt in order to determine the effects of SES, family attention
(FAM), school climate (SCLM), gender (FEM), peer drug inﬂuence (PDRG), and
externalizing behavior (EXTB) on Age of First Drug Use (AFU). The analysis began by
ﬁtting a one-way Random-Effects ANOVA model. This model is referred to as the null
model or the unconditional model in that there are no covariates at either level-1 or level-
2. The null model is used as a baseline in order to determine the total amount of
variability in the outcome within and between schools and as a comparison for

subsequent conditional models. Speciﬁcally the model was:

AFUij = 1301' + rij

104

BOj = 1'00 + qu
Using full maximum likelihood estimation, the model converged in four iterations. The
average school mean, 700, (intercept) was estimated as 12.27. The estimated between
school variance, 1'00 was 0.3407 and the estimated within variance, 0'2, was 4.95704.
Based on the covariance estimates, the intra-class correlation (ICC): 0.3407/(0.3407 +
4.95704) = .06431. This indicates that the portion of the total variance that occurs
between schools is a small 6.4%, leaving 93.6% of the variance to be explained (1-.064)
within schools. The 95% conﬁdence intervals for the magnitude of variation among
schools means was: 12.274 1.96*(0.3407)“2 = (11.126, 13.414). Again, this indicates that

there is a relatively small amount of variation in age of ﬁrst drug use among schools. The

magnitude of the variation among schools can be formally tested (Ho: Too = 0), and is

distributed using a large—sample x2 with J — 1 degrees of freedom under the null

hypothesis. The present model takes on a value of 119.42025 with 13 degrees of freedom
(I = 14 schools), and is highly signiﬁcant (p < 0.001). Taken together, what the null ’
model shows is that there is variance to be explained in AFU, and that the variance to be
explained seems to be primarily within schools, that is at level-1. One explanation for a
small proportion of between school variance may be the structure of employment in
Venezuelan schools. Very few teachers in Venezuela are so called “resident teachers.”
Unlike the system in US, where teachers are employed by a school and do all of their
work at one location, the Venezuelan teachers are primarily hired by the hour and may
teach in numerous schools during a given week. This practice of teachers moving from
school to school undoubtedly reduces the heterogeneity among schools. Additionally, the

sample of schools having come from the western section of Caracas, that is a singular

105

geographic location within the city, certainly contributes to the lack of variability found
between schools.

Even though the variance that remains to be modeled at level-2 is small, it is still
important to model. Due to the nested design of the data, a multilevel modeling approach
is still recommended for several reasons (Raudenbush & Bryk, 2002): First, Instead of
erroneously assuming that each observation adds a piece of independent information, a
multilevel model correctly accounts for the dependence in the data providing appropriate
unbiased and efﬁcient estimates of ﬁxed effects and standard error estimates. Second, a
class-level analysis would render low power to ﬁnd signiﬁcant results at the student level,
whereas a multilevel approach provides efﬁcient parameter estimates in a nested design
without sacriﬁcing power. Third, a multilevel model allows for a test of homogeneity of
regression and thus provides information regarding whether a given covariate should be

allowed to vary randomly or to remain ﬁxed.

Age of first drug use — Model-I
In order to address research question 2 outlined in chapter 3, and its subsequent
hypotheses, a series of models were ﬁt that would represent age of ﬁrst drug use (AFU)
in each of the J schools. A stepwise strategy was used that added variables according to
theoretical importance. All variables at level-1 were entered into the equation as raw
score variables (i.e., not mean centered). An uncentered approach was selected since
theoretically there is no reason to remove the between school variation as occurs in

centering. Moreover, this approach is a better ﬁt to the purposes of this study given the

106

focus on individual onset of substance use and to the data given the small proportion of
variance in the outcome attributed to level-2 (Kreft & de Leeuw, 1998).

The models began at level-1, speciﬁcally AFU for student i in school j was
regressed on the primary variable of interest for this study Family Attention (FAM).
Preliminary analysis showed a signiﬁcant positive correlation between FAM and AFU

(.156). The model is represented formally,

AFUij = l30j + Bu (FAM) ij + rij
Bot = Yoo + qu

Br j : 1’10
where BOj is the average age of ﬁrst drug use in school j when FAM is 0. Since all the

variables with the exception of those that are binary were standardized with a mean of 0
and a standard deviation of 1, the intercept becomes the average age of ﬁrst drug use in
school j for those individuals who have the mean scores on X. The relationship between
regression coefﬁcients and the outcome, then, can be understood in terms of a change of
one standard deviation in X produces a change corresponding to the coefﬁcient in Y. For
example, a coefﬁcient of .33 can be understood as an increase of one standard deviation
in X will result in a .33 of a year increase in AFU, or approximately 4 months.

The model with FAM converged in 4 iterations allowing the deviance of this model
to be compared with the deviance of the null model. Adding FAM to the model created a

better ﬁtting model as can be seen by the substantial reduction in deviance between the
two models with only one extra parameter estimated (x2 = 293.859, df 1 p< .001). Under

full maximum likelihood estimation, differences between deviances in two (nested)

models have a chi-square distribution, and these differences, compared to the degrees of

107

freedom lost, can show if one model is a signiﬁcant improvement over the other. A
likelihood ratio test can be used to test the signiﬁcance of magnitude of the improvement.
The 1 ratio for FAM is a highly signiﬁcant positive value (5.140) indicating that FAM is
an important predictor for AFU. These ﬁndings support hypothesis 2.4 “The age of first
actual drug use will vary by family attention, ” and indicate that on average
approximately 0.2934 of an increase in AFU (i.e., z 4 months) occurs for every standard
deviation increase in FAM (AF U= 12.3334j + .293375j*FAM ij + rij) in school j. The

level-2 slope for FAM remained constrained to zero.

Age of ﬁrst drug use — Model-2
The literature indicates that peer drug use (PDRG) is a strong correlate of
adolescent substance use. However, preliminary analyses revealed a non-signiﬁcant
correlation between PDRG and AFU (.022). Still, given the importance of this variable in
the literature it was entered into the level-1 model to represent formally its relationship

with AFU. Speciﬁcally,
AFUij = BOj + 151 j (PDRG) ij + l32j (FAM) ij + rij
13sz Yoo + qu
B1 j = Y 10
l32j 2 “1’20
where BOj is the average age of ﬁrst drug use in school j when FAM and PDRG are both

zero. Again, the model converged in 4 iterations. Since this model is nested within the

previous one, the deviance scores can be compared to assess model ﬁt. The change in

108

deviances was an impressive 126.180 with 1 additional parameter estimated, which
indicates that the model containing PDRG does seem to reduce the variance in AF U.
However, the t-ratio was not signiﬁcant (t= -.347, p= 0.728). Additionally the regression
coefﬁcient was very small (-0.02l736) and the standard error relatively high (0.062661).
Together these indicate that the variance explained by PDRG in the model is meager and
non-signiﬁcant, and therefore does not support hypothesis 2.6 “The age of ﬁrst actual
drug use will vary by peer drug use. This is a very surprising ﬁnding given the
prevalence of peer inﬂuences in the literature on adolescent substance use in the US. and
will be addressed in-depth in the next chapter. Given that PDRG does not add any
substantial information to the prediction of AFU, it was dropped from the model. F AM

remained in the model with a randomly varying intercept and ﬁxed slopes.

Age of ﬁrst drug use — Model-3
A third variable strongly supported by the literature as predictive of adolescent
drug use is externalizing behavior (EXTB). Preliminary analyses revealed a signiﬁcant
negative correlation between peer EXTB and AFU (-.151). In order to assess how EXBT
inﬂuences AF U holding FAM constant, it was entered into the model at level-1.

Speciﬁcally,
AFUiJ- = Boj- + 01,- (EXTB) ij 7‘ sz (FAM) ij + riJ'
130,: = 100 + qu
B11: Ylo

l32j = 120

109

where BOj is the average age of ﬁrst drug use in school j when FAM and EXTB are both

zero. Again, the model converged in 4 iterations. Since this model is nested within the
previous one, the deviance scores can be compared to assess model ﬁt. The change in
deviances was 48.20078 with 1 additional parameter estimated, which indicates that the
model containing EXTB does reduce the variance in AF U. The t-ratio is signiﬁcant

(t= -4.610, p< 0.001) and the regression coefﬁcient shows a strong effect (-0.281867, s.e.
= 0.061147). Together these indicate that the variance explained by EXTB in the model is
signiﬁcant, and supports hypothesis 2.5 “The age of ﬁrst actual drug use will vary by
externalizing behavior. ” The negative regression coefﬁcient indicates that on average

there is a 0.281867 of a year decrease in AFU for every standard deviation increase in
EXTB (AFU= 12.360149j'l' .209997J*FAM ij - 0.281867 j*EXTB ij + rij) in SChOOlj after

controlling for the effects of FAM. Therefore, FAM and EXTB remained in the model

with a randomly varying intercept and ﬁxed slopes.

Age of ﬁrst drug use — Model-4
A fourth variable that has been shown to play a role in adolescent substance use is
the sense of acceptance or bonding that the student feels with the institution. A
preliminary analysis also showed student climate (SCLM) had a signiﬁcant positive

correlation with AFU. Student Climate is entered to the level-1 model speciﬁcally,
AFUij = BOj + 1311' (EXTB) ij + By (FAM) ij + B3j (SCLM) ij + rij
Boy = Yoo + uoj'
1311': Y 10

132i: Y20

110

B3j = Y30
where BOj is the average age of ﬁrst drug use in school j when F AM, EXTB, and SCLM
are zero. The model converged in 6 iterations. Since the two models are nested, the
results of this model are compared to the previous yielding a change in deviance score of
53.61441 with one additional parameter estimated, which is signiﬁcant at p < .001.
However, the t-ratio for SCLM was not signiﬁcant (1.271 p= 0.204), the regression
coefﬁcient was small (0.079075), while the standard error relatively large (0.062205).
Together these indicate that little variance in AFU was explained by SCLM in the model.
Hypothesis 2.7 “The age of ﬁrst actual drug use will vary by student climate” was,
therefore, not supported and SCLM was dropped from the model. FAM and EXTB
remained highly signiﬁcant and were left in the level-1 model with random intercept and

ﬁxed slopes.

Age of ﬁrst drug use — Model-5
Preliminary analysis indicated that the gender variable, FEMALE (FEM), had a
small but signiﬁcant correlation with AFU (.096). In order to assess how FEM inﬂuences
AF U holding F AM and EXTB constant, it was entered into the model at level-1.

Speciﬁcally,
AFUij = Boj- + 01; (EXTB) ij + 132; (FAM) ij + 1331' (FEM) ij + rii
BOj = 700 + qu
1313': 1’10

[321' = 1’20

111

133; = 1'30
where POj is the average age of ﬁrst drug use in school j for males when FAM, EXTB, are
zero. The model converged in four iterations and the change in deviance was signiﬁcant
(x2 = 41.91004, dfl p < 0.001) as was the t-ratio (2.511, p < .001). Given that females
were scored 1 and males 0, the regression coefﬁcient can be understood as an increase in
AF U if the student is female. That is, I311 is the adjusted mean difference between males

and females in school j while controlling for the effects of FAM and EXTB. This

suggests that females, in an average school, start drug use almost 4 months after males

do, other conditions being equal (INTERCEPTj + 0.295063 j*FEMALEiJ-+

0.239002j*FAM,J-+ -0.247136j*EXTBiJ- + rij). With FEM added to the model the mean

difference in AF U explained by F AM and EXTB remained signiﬁcant (t = 3.965, p <
.001 and t = -3.978, p < .001 respectively). Therefore, FAM, FEM, and EXTB were left
in the model with a randomly varying intercept and ﬁxed slopes.

A ﬁnal step in ﬁtting the level-1 model added SES. However, SES did not explain
any signiﬁcant variance in AFU and was dropped from the model. As a result hypothesis
2.3 “The age of ﬁrst actual drug use will vary by SES " was not supported. Complete

level-1 estimates can be seen in Table 4.8.

112

Table 4.8 Estimates for Models

 

Fixed Effects
INTERCPT
FAM (Bl)
EXTB (82)
PDRG (B3)
STDCL (B4)
FEMALE (B5)
SES (B6)
MSES (G01)

Variance
Component

 

Intercpt V(UO)
FAM V(UI)
level -1 V(R)
ICC

Model Fit

 

 

Reliability (B0)
Reliability (Bl)
Deviance
Deviance Change

df

NULL Model (s.e.) Model-1 (s.e.)

12.2799**(0.1615)

0.31704
4.95943
0.0601
0.869
6749.60168

3

12.3344** (0.1579)
0.2934**(0.0571)

0.30023
4.83112
0.0585
0.862
6455.7427

293.86
4

Model-2 (s.e.)
12.3601**(0.1572)
O.2100**(0.0600 )
-0.2819**(0.0611)

0.29774
4.73022
0.0592
0.863
6407.5419

48.2
5

 

113

 

Table 4.8 Estimates for Models (continued)

 

 

 

 

 

 

 

Fixed Effects Model-3 (s.e.) Model-4 (s.e.)
INTERCPT 12.3547 **(0.1580) 12.3625**(0.1542)
FAM (Bl) 0.2400**(0.0705) 0.1973*(0.0817)
EXTB (B2) -0.3138**(0.0654) -0.2637**(0.0678)
PDRG (B3) -0.1092 (0.0731)

STDCL (B4) 0.0791 (0.0568)
FEMALE (B5)

SES (B6)

MSES (G01)

Variance

Component

Intercpt V(UO) 0.28338 0.27497

FAM V(Ul)

level -1 V(R) 4.74899 4.73912

ICC 0.0563 0.0548

Model Fit

Reliability (B0) 0.854 0.852
Reliability (B1)

Deviance 6280.543244a 6353.927484a
Deviance Change 127 53.61

df 6 6

Model-5 (s.e.)
12.2044**(0.1665)
0.2390**(0.060275)
-0.2471**(0.0621)

0.2951**(0.1175)

0.28495
4.69343
0.0572

0.858

6360631849"

46.91
6

 

114

 

Table 4.8 Estimates for Models (continued)

 

 

 

 

 

 

 

 

 

Fixed Effects Model-6 (s.e.) Model-7 (s.e.) Model-8 (s.e.)
DITERCPT 12.2049**(0.1619) 12.1967**(0.1655) 12.2032**(0.0960)
FAM (B1) 0.2350**(0.0607) 0.2541**(0.0824) 0.2497*(0.0833)
EXTB (B2) -0.2499**(0.0624) -0.2297**(0.0622) -0.2216**(0.0621)
PDRG (B3)

STDCL (B4)

FEMALE (B5) 0.2993*(0.1186) 0.3031*(0.1172) 0.2928*(0.1168)
SES (B6) -0.0461 (0.0692)

MSES (G01) -0.2525**(00.0350)
Variance

Component

Intercpt V(UO) 0.2626 0.28087 0.02653

FAM V(U 1) 0.0434 0.04596

level -1 V(R) 4.67309 4.65072 4.64926

ICC 0.0532

Model Fit

Reliability (B0) 0.846 0.855 0.362

Reliability (Bl) 0.489 0.503

Deviance 6254.34961 6356.868 6335,7095
Deviance Change 106.282242 3.76388 21.15843

df 7 8 9

 

Note: Coefﬁcients and standard errors have been rounded to nearest ten thousandth

21Change in deviance calculated from model-3
** p<.01 * p<.05

Before leaving the level-1 models, one other interesting point is that FEM, SES,
and SCLM were all signiﬁcantly related to the AF U in the OLS regression model, but
only FEM maintains the signiﬁcant relationship in the multilevel model. Raudenbush and
Bryk (2002) attribute the differences between estimates as coming from three common
errors: aggregation bias, misestimated standard errors, and heterogeneity of regression.

An in-depth discussion of these three common errors is beyond the scope of this

115

dissertation. However, these ﬁndings illustrate the importance of using a model that
accounts for the nested design of the data, even when the between school variability is

small. More will be said about this in the discussion section of this dissertation.

Age ofﬁrst drug use — Model-6
Next, a random coefﬁcients model was speciﬁed in order to guide the ﬁnal
development of the level-1 equation and to provide statistics for subsequent level-2
model building. As a ﬁrst step in this process, all three slopes were allowed to vary

randomly. The model converged after 450 iterations. The model comparison test of the

variance covariance components revealed a poor overall model ﬁt (x2 = 0.91545, df 9 p >
.500). The univariate x2 tests of homogeneity of variance for the Pqi coefﬁcients showed

that only F AM had signiﬁcant variation among schools (x2 = 24.14005, df 13 p < 0.03),

although the variance component indicated that after freeing the slope to vary across
schools, little variance remained to be explained (Var = 0.03678). The ﬁxed effects for
EXTB, FAM, and FEM continued to be signiﬁcant (t = -3.340, p = 0.006, t = 3.210 p =
0.007, and t = 2.548 p = 0.025 respectively). Since the x2 tests of homogeneity of
variance for FEM and EXTB indicated non signiﬁcant variability, the random effects for
FEM and EXTB were again set to 0 and the model was rerun with only a random
intercept and a randomly varying slope for FAM. Speciﬁcally the new model is
represented as,

AFUij = [30,- + B]; (EXTB) ij + 132; (FAM) ij + 1331 (FEM) ij + rij

I30) 2 “1’00 + uoj'

116

131F710
132) = 720+U2j

B3 j = 730

This new model converged in only 8 iterations. The t-ratios for the predictors
were all signiﬁcant at the p < .001 level, however, the test for model comparison yielded
a non-signiﬁcant chi square (x2 = 2.94049, df 2 p < 0.228). Tau (as correlations) were
unimpressionable (-0.028) indicating that the lack of signiﬁcant variance indicated by the
likelihood ratio test was not a result of multicollinearity.

The model reliability statistics provide additional guidance on appropriate
speciﬁcation of level-1 regression coefﬁcients (i.e., as ﬁxed, random, or non-randomly
varying) by indicating how much of the observed variation in the estimated slope is
potentially explainable. Raudenbush and Bryk (2002) suggest that whenever the
reliability of a regression coefﬁcient drops below .05 it is recommended to specify it as
either ﬁxed or non-randomly varying. In the present model, the reliability of estimated [32,-
is a relatively robust .489, lending support to a model with a random coefﬁcient. One
other insight that can be drawn from these results is in regard to the ability of the model
to detect a structural effect in the data. The strong reliability estimate for the intercept
(0.855) suggests good power to detect the effects of school characteristics on AFU.
Likewise, the moderate reliability estimate of FAM (.489) suggest that inferring how
school characteristics might inﬂuence AFU for students with differing levels of family
attention is also reasonable, although this must be viewed with some caution.

Models that allow a regression coefﬁcient to vary randomly often suffer from

instability, which is generally reﬂected in a decrease in the precision of the individual

117

parameters (Kreft & de Leeuw, 1998). However, in the present model the coefﬁcients and
their standard errors remain relatively unchanged (compare models 5 and 6 in Table 4.8).
This also lends support for leaving the random coefﬁcient in the model.

A range of plausible values with 95% conﬁdence intervals can be calculated for

the random slope estimate. Thus for this data school means (Boj) would be expected in the

range of(1 1.1592, 13.23541), and the differentiation effect of FAM ([31,) in the range of

(-0.15421, 0.662432). These results suggest moderate variation among schools on the

FAM effect, and we would expect to ﬁnd some schools where family attention effects are
negligible since values near zero are plausible for ng. Visual inspection of the graph of

AF U regressed on F AM shows that slopes vary considerable among schools with FAM
having negligible effects in some schools and very strong positive effects in others (see

ﬁgure 4.1).

118

 

 

I - - MEANSES: lower
’ ’ — MEANSES: mid 50%
"""" MEANSES: upper

13.41 i ’

 

  
 

AGEFIRST

11.40 ..

 

r I

 

Y I I I r I ﬁ’ I T 1

11147 -020 1.07

FAM

r r I r
-2.74

 

Figure 4.1. FAM random slopes by AFU identiﬁed by MSES.

119

 

Together these statistics suggest that while the speciﬁcation of [32,- as random
cannot be ﬁrmly determined empirically, there is sufﬁcient evidence to leave it in the
model unless theory indicates otherwise. Moreover, Raudenbush and Bryk (2002) point
out that the likelihood ratio statistic for the null hypothesis that one or more variance
components is null tends to be conservative when the number of groups is small, and thus
decreases the chances of rejecting a false null hypothesis. Now a brief discussion will be
given regarding the theoretical rationale for allowing the FAM coefﬁcient to vary
randomly.

Theoretically, we would expect to see family interaction vary across schools for
several reasons. First, in Venezuela, students must apply (compete) for openings “cupos”
at both private and public institutions from kindergarten through high school. Schools
with better academic reputations are often in a position to deny admittance to students
whose families do not respond to the administration’s requirements, or whose guardians
are not diligent in seeking out these “cupos” in a timely fashion. This would have the
effect of creating a school speciﬁc characteristic that has the potential to inﬂuence
aspects of family attention. Second, at least three of our schools were religious, and
another was a nonreligious private institution with foreign funding that strongly
encourages communication between the school and guardians of the minor students. With
the addition of these characteristics, it is reasonable to expect that family attention could
vary differentially among schools. Third, Hirchi’s theory of institutional bonding (1969)
suggests that students do attach to institutions and that this attachment has the potential to

moderate the effects of F AM.

120

In conclusion, while there is not enough between school variance in [32,- to model
how differing school characteristics might affect the way FAM varies among schools,

both empirically and theoretically it seems that allowing [33,- to vary randomly is plausible,
if not warranted. The evidence provided by t ratios, the too point estimates, the x2 test of

homogeneity, and the reliability estimates indicate that there is enough variation among
schools in [32; to treat it, at least initially, as random. Therefore, it was decided to leave the

random coefﬁcient in the model and proceed to model the variability in the intercept.

Age ofﬁrst drug use — Model-7

In preparation for adding predictors to level-2, an analysis was run in HLM 6.02
to determine potential level-2 predictors. These results found that MSES (t = -6.348),
PUBLIC (t = 3.118), MSCHCON (t=3.328), and MSCLM (t=-3.885) held good potential
to explain level-2 variance in the intercept. Clearly, MSES is the strongest candidate to
explain the between school variance in the intercept. Theoretically, MSES contains more
precise information than PUBLIC since the difference between public and private schools
is usually a proxy for some other characteristic such as SES. MCOND is a composite
score derived from the MAMBI that describes the physical condition of the school. A
descriptive analysis revealed that students with very low SES are at times in very well
maintained schools due to external funding sources. This is the case with one non-
religious private school in the sample that was founded by interests from the US. as part
of an international outreach effort to beneﬁt underprivileged children in several third-
world nations. Therefore, both empirically and theoretically MSES seemed to be the best

choice to begin modeling at level-2.

121

MSES was added to the second level intercept grand mean centered with the
FAM slope varying randomly and EXTB and FEM constrained to 0. Speciﬁcally the new
model is,

AFUU = BOj + [31; (EXTB) ij + 132,) (FAM) ij + 1331' 0:13th + rii

130,: “/00 + 701 (MSES, —MSEs..),- + Uoj'

Br) = 710
132) = 1’20 . Uzj
133)” 7" 1’30

where 130‘, is the intercept, yo] is the effect of MSES on [30} While uoj had been the
deviation of school j’s mean from the grand mean, it now represent the residual BOj - 700 -

701 (MSES; —MSES..)j, and T00 is the variance in I30)“ after controlling for MSES. The
combined model can be expressed as,
AFUij = 1’00 + 1’01 (MSES) —MSES--)j + 110(13XTB)1)'+ 120 (FAM) ij + 1’30 (FEM) ij + qu +
U21 + Ft

The model converged after only 12 iterations, which in itself is an indication of a
well ﬁtting model. We ﬁrst see that the ﬁxed effects show MSES is negatively related to
AF U (estimated 701 = 0252506, I = -7.219). The x2 likelihood ratio test shows that the
change in deviance with one additional parameter estimated is signiﬁcant (x2 = 21 .15843,
p < .001) and the model is a good ﬁt. The variance component has been reduced from
0.28087 in the previous model to 0.02653 (x2 = 21 .32558,p < .045) indicating that
MSES has accounted for approximately 90% of the available variance to be explained in

the intercept. The negative coefﬁcient of MSES indicates that a one standard deviation

122

increase in the average SES of school j decreases AF U by approximately .25 years, or 3

months after accounting for the effects of EXTB, FAM, and FEM [AF Uij = 12.203181 +
-0.252506* (MSESJ- —MSES..)j + -0.221641*(EXTB) ,j + 0.249663*(FAM) ,J- +
0.292807*(FEM) ii]- Since the level-l predictors are standardized with a mean of 0 and a

standard deviation of 1, and the gender variable is binary with males having a score of 0,
one standard deviation increase MSES lowers the age of onset (the intercept) in an
average school by approximately 4 months for male students in a family with average
attention and manifesting average externalizing behavior. Based on this equation we can
conclude that the earliest predicted age of ﬁrst drug use is for a male student with low
family attention and high externalizing behavior in a school where students, on average,
are high SES.

The following graphs illustrate this relationship. In ﬁgure 4.2, the slopes represent
the positive linear relationship between AFU and FAM with increases in FAM related to
increases in AFU. The ﬁrst line (from the bottom up) represents males with high
externalizing behavior. The next lines represents females with high externalizing
behavior; followed again by males with low externalizing behavior; and ﬁnally by
females with low extemalizing behavior. We see that while holding the effects of FAM
constant, EXTB seems to inﬂuence AFU equally for males and females. Likewise, this
graph illustrates that students with high EXTB initiate drug use earlier than students with

low EXTB regardless of gender.

123

 

12.99: ,’

   

 

 

12.46‘
I-
(D
E
11. 11.93‘
LLI O"
3 — EXTB = -1.028,FEMALE = 0
" .o’" "' EXTB = -1.028,FEMALE = 1
11.40+ .0". ’o.’
,a "" EXTB = 1.471,FEMALE = o
’ ,o" """ EXTB = 1.471,FEMALE = 1
9’.
1088401 -2.74 -147 0.20 1.07

FAM

 

 

 

Figure 4.2. The joint relationship of EXTB and FEM on AF U holding F AM

constant.

124

The following graph (Figure 4.3) offers roughly the same picture as the previous
one with approximately the same interpretation. Here the slopes represent the negative
linear relationship between AF U and EXTB with increases in EXTB related to decreases
in AFU. The ﬁrst line (from the bottom up) represents males with low family attention.
The next lines represents females with low family attention; followed again by males
with high FAM; and ﬁnally by females with high FAM. Again, we see that while holding
the effects of EXTB constant, FAM seems to inﬂuence AF U in the same way for males
and females. That is, students with low FAM initiate drug use earlier than students with

high FAM, regardless of gender.

125

 

 

AGEFIRST

I
O
O
O
O
O
. O
.

 

— FAM = -1607, FEM = 0
——. FAM = -1.607, FEM =1

_.. . . FAM = 1.039, FEM = 0
FAM =1.039, FEM = 1

 

 

 

Figure 4.3. The joint relationship of FAM and FEM on AFU holding EXTB

constant.

126

 

The next graph (Figure 4.4) offers a slightly different picture by comparing the
effects of F AM and EXTB together on males and females. Here again the data seem to
suggest that the effects F AM and EXTB have on AFU are nearly the same for both
groups, albeit with differing intercepts. Starting from left to right for males (0.00), the
ﬁrst bar indicates that low F AM and low EXTB are about the same as the high F AM and
high EXTB (last bar in 0.00 group). And, the proportion of this relationship seems to hold
among females but with higher intercept values (see ﬁrst bar compared to last bar in 1.00
group). For both males and females the earliest age of ﬁrst onset is when there exists low
F AM and high EXTB (second bar in both groups), the latest age of ﬁrst onset is when
there is high F AM and low EXTB (third bar in both groups). The relative effects of FAM
and EXTB seem to be equal across gender. This graph also suggests an additive effect
between F AM and EXTB. For both males and females, the effects of FAM and EXTB
seem to either cancel each other out creating a mid-range effect (i.e., high FAM — high
EXTB, low FAM — low EXTB), or they accentuate each other creating either strong
negative or strong positive movement in AF U (i.e., high FAM — low EXTB, or low FAM

— high EXTB). And this effect is the same across genders.

127

 

 

AGEFIRST

13.36‘

12.79‘

12.23‘

11.66‘

 

11.10:

A

 

    

   
   

 

  
   
    

 

 
   
   

 

 

 

 

 

    
   

     
  
      
    
  
   
  

   
 
   
  
  
 

 

 

 

W
w
9"
«N»
N
F‘ «v.»
\- * 01“»
"s 5: V‘V‘:
., 5. V‘s/‘1
:: a $‘"
5. L‘ J”
t: :1 WV“
:: n as.“
~« ~: . w»:
:2 :‘ P O 9 3.».
is 5‘ . . N
.. a; ..... we»:
:1 :11 k... 3' ‘
N ::I V.’ ’
~ no -~
:! . . “w
. a: O O 0 $3
9 O ,V.
no ,N.
09 «Mb
0 o o 3.,
4 V‘

       

.FAM = -1.607,EXTB = -1.028
m FAM = -1.607,EXTB = 1.471

Illlllll FAM = 1.039.EXTB = -1.028
FAM =1.039,EXTB = 1,471

 

Figure 4.4. The joint relationship of FAM and EXTB on AFU by gender.

128

 

In ﬁgure 4.5 the second level effects are introduced. Here the ﬁrst line (from the
bottom up) represents students with high externalizing behavior that attend high SES
schools while holding family attention constant. However, the next line represents low
externalizing behavior and high SES schools holding F AM constant. The third line in the
graph again represents high externalizing behavior and low school-level SES, but has a
considerably later age of onset then the low externalizing behavior/high SES relationship.
The fourth line represents low EXTB and low MSES while holding FAM constant. This
graph shows a cross-level interaction where externalizing behavior is being moderated by
school-level SES. That is, the impact of the relationship of EXTB on AF U depends on
the normative environment in a school as characterized by MSES. This is a particularly
interesting ﬁnding since individual-level or level-1 SES was not a signiﬁcant predictor of
age of ﬁrst drug use (see table 4.8). That is, the disposable resources available to a
student on a personal level do not seem to be related to her or his age of ﬁrst drug use.
However, the normative environment as characterized by a school level mean SES is

negatively related to age of ﬁrst drug use, and moderates the effects of EXTB.

 

 

AGEFIRST

— EXTB = -1.028,MSES = -1.996
-- EXTB = -1.028,MSES = 3.061
V ..... — - - EXTB =1.471,MSES = -1.996

. ........... EXTB =1.471,MSES = 3.061

 

 

 

f

1281” 1154+ 028 “0.199

FAM

 

Figure 4.5. The cross-level interaction effects of EXTB by MSES on AFU

holding FAM constant.

 

As in the previous graph, ﬁgure 4.6 illustrates a cross level moderation of MSES
on FAM in school J. Again, the ﬁrst line (from the bottom up) represents low F AM at
level-1 in school J, and high MSES measured at the school level while holding EXTB
constant. However, one would expect that the next line will be low-FAM, low-MSES if
the family variable were exacting the stronger effect on AFU. Instead, the second line is
high FAM — high MSES, which again indicates the cross-level interaction that moderates

the relationship between FAM and AFU.

131

 

 

AGEFIRST

— FAM = -1.607,MSES = -1.996
__. FAM = -1.607,MSES = 3.061

    

 

 

13.42'\ . .
‘ \ . . ooooo FAM =1.039,MSES = 3.061
\ .
, ' \ , — - 'FAM =1.039,MSES=-1.996
. \
12.69: . ' \
. . \
11.96“ .............
+ .............
<h~~ ...........
11.23: “a--- .........
1 ‘~~~
i ~~~~~~
I ~‘~~~~
10.49 ' r I r r 177 r r v v I r . v v v I 7‘
-0.65 0.46 1.58 2.70
EXTB

 

 

Figure 4.6. The cross-level interaction effects of FAM by MSES on AFU holding

EXTB constant.

Figure 4.7 demonstrates the effect of MSES on gender. Here the ﬁrst line (from
the bottom up) as we have seen represents males from schools with high average SES
scores holding constant FAM. The next line represents females from high average SES.
However, the male and female lines in the high MSES are relatively close together, as are
the lines in the low MSES schools. This indicates a grouping effect in AFU that tends to
erase the differences between males and females. This becomes increasingly obvious by
the larger jump in AFU that seems related to difference between high and low MSES
(i.e., compare the difference between lines 1 and 2, to the difference between lines 2 and
3). Thus, it seems that school-level SES seems to have a strong equalizing effect for

males and females in determining age of ﬁrst substance use.

 

AGEFIRST

 

 

 

 

11.12 ,x“ ......... — - - FEM = 0. MSES = -1995
1 xx ------------ FEM = 0, MSES = 3.061
Ix” ........ — FEM = 1, MSES = -1996
. ........ --- FEM =1, MSES = 3.061

10.41 .I. I I I I I I I I l

401 -274 -147 —0.20 1.07
FAM

 

 

Figure 4.7. The cross-level interaction effects of FEM by MSES on AFU holding FAM

constant.

4

.)

Having added MSES as a level-2 covariate, the explainable variance in the
intercept seems to be depleted to a negligible amount and since MSES was determined
theoretically to be the most salient level-2 covariate, no other level-2 variables are added
to the model. Likewise, the meager remaining variance to be explained in F AM suggests
that the model is at peak balance between parsimony and ﬁt. Therefore, the study now

turns to diagnostics in order to assess the adequacy of the model.

Age of First Drug Use — Model Diagnostics

I With a tentative model developed, an essential part of model building is
determining the adequacy of the model by assessing if the assumptions of the model
appear valid for the data. Two of the more important assumptions are: (a) that the level-1
errors are independent and normally distributed with a mean of zero; and (b) that the
random effects are normally distributed with a mean of zero, and are independent across
groups. These assumptions were tested empirically using the level- 1 and level-2
residuals produced by HLM 6 during the modeling process. Level-1 residuals are
examined ﬁrst.

A boxplot of the within school residuals can be used to determine if the residuals
are centered at 0, and that the variances are constant across groups. Figure 4.8 shows that
the residuals seem to be centered at 0, albeit with a small amount of variability, and that
the variability appears to be fairly constant across schools. The several data points that
appear to be outliers are students who reported very early ages of ﬁrst drug use (e.g., 4
years of age). This was discussed in chapter 3, and the theoretical rationale for leaving

these observations in the data set was given. Increasingly early onset of drug use is

becoming a phenomena common to many countries (United Nations General Assembly
Special Session on the World Drug Problem, 1998) and therefore argues more for their

inclusion than to viewing them as outliers.

 

10.000‘

5.000

0.000“

Level-1 residuals

-5.000‘

-10.000‘

 

 

 

 

 

 

o
O 0300-0
00 000 0*
o o
o o O

 

 

rﬁrrrr 11111111
0102 03 04 05 06 07 08 091011 121314

Schools

Figure 4.8. Box plot of Level-1 residuals by each of the 14 schools.

137

A scatterplot of the residuals against the ﬁtted values is used to assess for
problems with heteroscedasticity. Figure 4.9 shows that there are no recognizable
patterns, which indicates that the assumption for heteroscedasticity is reasonably met.

Finally, a normal P-P plot of the level-l residual is used to assess the normality
assumption of the data. If data are normally distributed they will be arrayed along a
straight line in the P-P plot. Figure 4.10 shows that the data for age of ﬁrst drug use
appear to be quite normal.

Additionally, a test of homogeneity of level-1 variance was performed to assess if
the variances depend systematically as a function of level-1 or level-2 predictors. A test is
provided in the HLM 6.02 software. Formally the test of homogeneity of level-1 variance
is

H0 = Z d2,
where d is standardized measure of dispersion for each group j. This statistic has a large
sample x2 distribution with J — 1 degrees of freedom under the homogeneity hypothesis
(Raudenbush & Bryk, 2002). A rejection of the null would indicate heterogeneity of the
level-1 variances and may indicate a mis-speciﬁcation of the level-1 model, which holds
the potential to bias estimates of the level-2 coefﬁcients. The test was not signiﬁcant (x2
= 19.74179, df13, p = 0.102), so the null is retained and it is concluded that the level-1
variance does not depend on measured predictors. Having shown that the level-1

assumptions are reasonably met, level-2 assumptions are now addressed.

138

 

0
©
5000— O O 00 00 00 0
Q) 0

o Cno
00
0.000— 000m 00
m
o 00

l1resid

 

 

 

0 00° 0
-5.00 o Co 0 a)
00 00 000
c
Q) 0 00
o
0
-10000—
1 r 1 r 1
10.000 11.000 12.000 13.000 14.000
ﬁtval

Figure 4.9. Scatterplot of level-1 residuals against the ﬁtted values.

 

1.0“

0.8“

0.6“

0.4“

Expected Cum Prob

0.2“

 

 

 

 

00‘ I I I I I
0.0 0.2 0.4 0.6 0.8 1.0

Observed Cum Prob

Figure 4.10. P-P plot of the level-1 residuals

140

A visual inspection of the Q-Q plot found that the Mahalanobis distances were
approximately )8 distributed (see ﬁgure 4.1 1). Additionally, Mahalanobis distances were
plotted against the expected values of the order statistics for a sample of size J schools.

The constructed Q-Q plot shows that the random effects are distributed approximately

v-variate normal (see ﬁgure 4.12). A Pearson correlation rp between the expected and

observed Mahalanobis distances was statistically signiﬁcant (rp = 0.94, p < 0.001). HLM

is robust against violations of normality (Raudenbush & Bryk, 2002), and this assumption
seems to be reasonably met. The mild divergence of several of the observations can be
attributed to differing sample sizes among the 14 schools and should not be taken as
evidence of a non-normal distribution. For example the sample from school 10 had 11 =
212, where school 9 had n = 104, clearly more than twice as large. Table 4.9 provides a

list of the sample sizes within each school.

Table 4.9 Within school sample size
School 1 2 3 4 5 6 7 8 9 10 11 12 13 14
N 135 119 142 150 104 109 104 124 104 212 109 149 132 125

 

 

 

 

141

 

 

 

 

4.000“
3.000— 0 o
o
H
.52
E 2.000“
o o
o
o
1.000— O O
o
0
0.0001
1 I I I I
0.000 1.000 2.000 3.000 4.000
chipct

Figure 4.11. Normal Q-Q plot of Mahalanobis’ Distance

142

 

4—4

N co
1 1

Expected Normal Value
T

 

 

 

 

I I I
0 1 2

Observed Value

w—
#—

Figure 4.12 Mahalanobis’ Distance by the expected values of the order statistics.

143

Finally, the empirical Bayes estimates of the random intercept were plotted
against MSES to assess for a linear relationship between the intercept and MSES. Again,
a visual inspection of the scatterplots suggests that the residuals do not follow a
systematic pattern and the assumption of a linear relationship is reasonably met.

Together these tests offer evidence that the level—1 and level-2 assumptions have
been reasonably met and provide conﬁdence in the ﬁtted model. The next chapter will
provide a more in-depth discussion of the ﬁndings reported in this chapter and their

implications.

144

 

 

 

 

0.200 -
o o
o
0 .1 00 -
0
IL
8 o o
*5 0.000 —
'3 o
o o
o o o
o
-01 oo —
o
o
0200 -4
I I I I I I I I
-3000 -2000 -1 .000 0.000 1 .000 2.000 3.000 4.000
meanses

Figure 4.13. Empirical Bayes intercept estimate by MSES.

CHAPTER V: DISCUSSION

This study looked at how known risk and protective factors for adolescent
substance abuse in the US. are present in a sample of Venezuelan school-attending
adolescents. Results of a hierarchical linear model showed that family attention,
externalizing behavior, and gender were signiﬁcant level-1 covariates, and that mean
socioeconomic status was a signiﬁcant level-2 covariate. This chapter provides a

discussion of the results, implications, limitations, and future research directions.

Discussion of Results
General Description of Drug Use

Overall, the sample was comprised of a relatively high number of students that
had consumed alcohol (81.5%). The next highest drug consumed was cigarettes with a
relatively moderate 31.5% having at least tried cigarettes. However, incidence of illicit
drug use was very low ranging from .3% (crack), .4% (heroin), .9% (ecstasy), 2.5%
inhalants, 3.7% (marijuana) to 8.8% (prescription pills). In comparison to their
Venezuelan counterparts, adolescents age 11-19 in the US. report very high alcohol
(98.7%), cigarette (96.2%), and marijuana (96.8%) use (SAMHSA, 2005). Additionally,
according to the National Survey on Drug Use and Health (N SDUH) report the contrast
of the percentages of youth who have tried other illicit drugs in the US. as compared to
Venezuela is dramatic: crack (22.6%), heroin (9.8%), ecstasy (64.9%), and inhalants

(24.1%).

146

Although alarming, the discordance in percentage of drug use among youth in the
US. compared to Latin America is not surprising. Epidemiological studies have
consistently reported very low levels of adolescent illicit drug use, even in areas where
drug crops are commonly cultivated, produced, and subsequently exported (Vega & Gil,
1998). Other studies have found a signiﬁcant positive relationship between rates of drug
use for foreign-bom Hispanics living in the US. and length of time living in the US.
(Vega & Gil, 1998; Warheit, Vega, Khoury, Gil, & Elfenbein, 1996).

The low prevalence of illicit substance use in the sample is contrary to the general
perception of many of the teachers, therapists, and parents interviewed in focus groups
conducted by the author. Teachers, therapists, and parents alike decried the ubiquity of
illegal substances in Caracas neighborhoods. Two possible explanations for the
discrepancies between the data of the present study and qualitative reports come to mind.
First, the perception of increases in any behavior is contrasted to what had been the norm
and can be distressing to the observer. In a city such as Caracas where not only is the
population voluminous (approximately 5,000,000 in the metropolitan area), it is also
densely packed in a series of valleys among the mountains. Therefore, even slight
increases may be highly visible and perceived as alarming. Second, since the sample was
drawn from school-attending youth, school may function as a protective factor with the
great majority of youth who engage in the consumption or trade of illicit substances
dropping out of school and therefore excluded from the current measure. If the latter
proves to be true, then the low prevalence of substance use among Venezuelan youth is,
at least in part, an artifact of the survey methodology employed. Future research should

incorporate survey methods that capture non school-attending youth in order to

147

understand more fully both the prevalence of substance use and the risk and protective
factors that may be Operating among Venezuelan youth in regards to their decisions to
consume illicit substances.

Compared to use of illicit substances, higher rates of subjects reported having
consumed alcohol (81 .3%) and/or smoked a cigarette (31.5%) at some point. Some
experimentation with licit substances such as alcohol or tobacco is considered by most to
be normative and not a cause for concern (e.g., Zucker, Fitzgerald, & Moses, 1994).
Moreover, in Venezuela parents often include alcohol in family functions and
celebrations such as weddings, and approve of children drinking under their supervision.
However, since the survey did not specify where drinking took place, nor whether or not
parents were present, this ﬁnding may be an artifact of the survey design more than an
indication of a cultural phenomenon. The ﬁndings also revealed that 42.9% of youth
reported not having drank any alcohol in the past year, and approximately 28% reported
drinking only sporadically (i.e., “every once in a great while”) throughout the year. Still,
almost 15% reported drinking monthly, 7.8% reported drinking weekly, and 2.4%
reported drinking daily. These ﬁndings suggest that alcohol consumption may be a
concern for approximately 25% of the sample. Furthermore, for the 7.8% who reported
weekly drinking and the 2.4% who reported daily drinking, their level of use may have
already passed into the realm of abuse, which may place them on the pathway to alcohol

dependence.

148

Independent Variables

Family Attention

The predicted relationship between family attention (FAM) and age of ﬁrst drug
use (AF U) received solid support in the data analysis. Family life characterized by a lack
of parental support, poor communication, increases in negative reinforcement, and poor
boundary setting and monitoring has been shown to be one of the stronger covariates with
the onset of a series of child misconduct problems including substance use (Belcher &
Shinitzky, 1998; Gorman-Smith, Tolan, Zelli, & Huesmann, 1996; King & Chassin,
2004). In the present study increases in FAM (i.e., positive family functioning) was
correlated with signiﬁcant increases in age of ﬁrst drug use (AFU), and this relationship
was maintained even after accounting for the effects of externalizing behavior and
genden

Descriptive statistics showed that FAM (standardized to have M=0 and SD=1)
was negatively skewed with a median of .34 and a mode of 1.07. The pronounced skew
in FAM created a ceiling effect with the majority of participants (25%) scoring a full
standard deviation above the mean. This ceiling effect is suggestive of strong positive
family relationships for a signiﬁcant portion of the sample, and that these relationships
are associated with increased AF U for school attending youth in Venezuela. Interestingly,
though, this effect varied across schools. The random slopes model showed a signiﬁcant
effect for FAM, and Figure 4.1 illustrated that the effects of F AM were negligible in
some schools and highly signiﬁcant in others. A decomposition of the graph in ﬁgure 4.1
shows that the random effect roughly follows the pattern of public vs. private with the top

ﬁve most signiﬁcant slopes belonging to public institutions and ﬁve of the six least

149

signiﬁcant slopes belonging to private schools. Unfortunately, an empirical explanation
for these differences is beyond the scope of this study since the restricted number of
schools in the sample limited the between school variance and thus, the ability to model
the F AM slope. Nevertheless, this is an important preliminary ﬁnding because it supports
a contextual approach to the development of treatment and prevention strategies. For
example, there may be characteristics inherent in public institutions that indicate
increased emphasis on parental monitoring, or that by introducing parental monitoring
earlier as opposed to later in the intervention will lead to quicker results and thus
encourage families to continue with the treatment. Or, perhaps because parents are paying
for their child’s education in private institutions, they are more inclined to be involved
with their children. While these speculations are not able to be tested using the present
data, the results argue for an ecological approach that hypothesizes differential effects
associated with school level variables that interact with families to inﬂuence AF U at the
individual level.

The literature is replete with studies that place parental monitoring and parental
support as the two primary family variables that are associated with adolescent substance
use. The original PACARDO study conducted by Anthony and colleagues (2001),
recognizing this pattern, combined items that assess for parental monitoring and parental
involvement to create the F AM variable. The present study followed the lead of Anthony
and colleagues and utilized the same approach. However, in an attempt to understand
more fully the dynamics of this variable, the two scales were recreated as separate
variables and entered into the multilevel regression. Interestingly, the ﬁxed effect for

parental support was highly signiﬁcant while the effect for parental monitoring was not

150

even marginally signiﬁcant (t = 3.721, p < .001 & t = 1.196, p = 0.232, respectively).
These results do not necessarily suggest that parental monitoring is unimportant for
Venezuelan youth, but only that parental monitoring may not be signiﬁcantly associated
with AFU in a cultural context that is permissive toward alcohol use by minors. Given the
high incidence of alcohol use by school attending youth, and the very low occurrence of
illicit drug use, when AF U is regressed on parental monitoring the majority of the
variance to be modeled is among youth who have consumed alcohol. Therefore, it may be
that since modest alcohol consumption is condoned by many parents, parental monitoring
loses its power of association to vary signiﬁcantly with AF U. However, this needs to be
further explored in future studies.

Another potential explanation for this ﬁnding is a cultural mis-speciﬁcation of the
measurement of the parental monitoring construct. In a large urban, Latin American city
such as Caracas, what parents and youth consider as monitoring may vary signiﬁcantly
from established criteria in the US. Additionally, other mechanisms employed by parents
to monitor their children may have been developed in a response to a distinct context, and
are not tapped by the items that make up the parental monitoring measures typical of the
US. context. These mis-speciﬁcations would be reﬂected in the relationship between
parental monitoring and AF U.

F urtherrnore, the lack of signiﬁcant association between parental monitoring and
AFU may be due to measurement error. Important to note is that the items used to assess
for parental monitoring were used in the original PACARDO study as a component of the
family attention variable and not as a stand alone scale for parental monitoring. Of the

seven items that comprise the family attention variable only three of those are indicative

151

of parental monitoring (V16, V18, & V20). For the present data the Omega procedure
yielded a reliability estimate of .56. and the Pearson correlation coefﬁcient for the
relationship between parental monitoring and parental support was a modest but
signiﬁcant .265 (p < .01). Obviously, the three items used to capture monitoring behavior
among parents are not comparable to a fully developed scale that has withstood rigorous
psychometric evaluation. Therefore, while intriguing, conclusions drawn from this
ﬁnding would be inappropriate without further research.

Finally, Steinberg and colleagues (1995) found that the pathway of parental
monitoring to adolescent drug use was mediated by peer groups, with parental monitoring
inﬂuencing delinquent behavior through the choice of peers by youth. If in Venezuela
peer inﬂuence is not an important covariate of AF U, a subject that will be taken up
shortly, then parental monitoring would lose its signiﬁcant relationship to AFU.

In conclusion, F AM covaries signiﬁcantly with AFU. However, the relationship
in this study seems to be through parental support more than through parental monitoring,
is bound to a cultural context that is permissive to alcohol consumption among youth, and
is moderated by a school context. Again, these ﬁndings support the importance of
culturally adapted ecological approaches to treatment and prevention interventions in
order to increase the effectiveness of the interventions and meet the differing needs of
families in distinct contexts. These ﬁndings are important areas of investigation in future
studies.

Externalizing Behavior
Research has continuously shown a strong link between EXTB and substance use

(e.g., Clark & Winters, 2002). The results of the present study involving Venezuelan

152

adolescents agree with previous research and support the hypothesis that externalizing
behavior (EXTB) is signiﬁcantly related to AF U, such that as EXTB increased AFU
decreased. Additionally, research on youth from the US. suggests that EXTB is
signiﬁcantly correlated with negative family relationships (Capaldi, DeGarmo, Patterson,
& Forgatch, 2002). Parents that engage in higher levels of antisocial behavior, tend to
also manifest aggression toward each other, to be less consistent in their discipline, and
use more coercive behavior with their offspring, which in turn leads to a failure to inhibit
youthful EXTB (Capaldi, et al., 2002). The present study on Venezuelan youth also
found a signiﬁcant negative correlation between FAM and EXTB (r = -.362, p < .001).
Youth with higher levels of EXTB are in families who exhibit lower levels of FAM.
While causality cannot be inferred from the present ﬁndings, they do corroborate results
from US. studies on substance using youth and lend initial support for the
implementation of family-based treatment and prevention strategies in Venezuela.
Beyond the association between EXTB and initiation of substance use, if left
unaddressed, EXTB may develop into a conduct disorder (CD). Furthermore, in a
longitudinal study by Sartor and colleagues (Sartor, Lynskey, Heath, Jacob, & True,
2007), CD was shown to be the only risk factor assessed that covaried both with alcohol
initiation and progression from ﬁrst drink to a later alcohol disorder. These researchers go
on to report that the role of CD in the development of alcohol disorder is most probably
through involvement with other deviant peers, who reinforce drinking and other
antisocial behaviors. Peer inﬂuences were also signiﬁcantly correlated with EXTB and
FAM in the present study (r = .383, p < .001, and r = -.315,p < .001 respectively). These

ﬁndings again echo what Steinberg (1995) and others (e.g., Capaldi, et al., 2002; Liddle

& Dakof, 1995; Szapocznik & Coatsworth, 1999; Szapocznik & Williams, 2000) who
draw from human ecology theory have proposed that microsystemic interactions more
than a single causal agent should be considered when developing strategies to interrupt
problem behavior.

While the form of these strategies, their timing in delivery, and other factors may
differ across cultures, it appears that the interaction between family, peers, and antisocial
behavior, and the inﬂuence these play in the onset of substance use are constant from
US. to Venezuelan culture.

Peer Drug Inﬂuence

Although research in the US. has consistently shown that youth whose peers use
drugs will also tend to use drugs, this relationship was not supported in the current study.
The multilevel model revealed a mild negative ﬁxed effect for peer drug inﬂuence
(PDRG) on AFU that was non-signiﬁcant. Several potential explanations may shed light
on these ﬁndings.

First, in Venezuela, as is true in other Latin American countries (Vega & Gil,
1998), there is a strong societal prohibition against illicit drug use. This cultural attitude
toward illicit drug use may create a protective factor that helps youth ﬁnd greater
amounts of peers that do not use drugs by making open declarations of drug use among
peers who do use less common. In US. schools multiple subcultures are openly formed
and youth attach themselves to one or another (e. g., the “jocks,” the “burnouts,” the
“druggies,” or the “brains”). This process is often inﬂuenced by environmental
mechanisms. For example, the study by Steinberg and colleagues (1995) mentioned

earlier in this dissertation, pointed out how parental involvement interacted with

154

adolescent subcultures formed on the lines of ethnicity. This study revealed, for instance,
that even though African American parents scored consistently high in parental
involvement, it was difﬁcult for African American youth to break into the “brains”
subgroup. In Venezuela, it may be that the peer subculture that uses illicit drugs is
ostracized and leave school, and therefore was not represented in the sample.
Interestingly, this same pressure does not seem to occur with cigarettes or alcohol.
Second, as a result of the societal pressure not to become involved with illicit
drugs, it may be that in Latin America peer inﬂuences are more of a protective factor that
buffer against the opportunities to become involved with illicit drugs. Differential effects
of peer groups by culture have been found by at least one other research team. In a
longitudinal study, Apospori and colleagues (Apospori, Vega, Zimmerman, Warheit, &
Gil, 1995) found that while peer group associations were linked with early drug
experimentation for White non-Hispanics, they functioned as a protective factor against
early experimentation for African American youth. Equally intriguing is that they found
that this effect waned over time due to changing perceptions and levels of tolerance
toward deviant peers. Interestingly, comparing the correlations between PDRG with age
of ﬁrst illicit drug use and PDRG with AF U, reveals that there is a signiﬁcant negative
correlation between PDRG and age of ﬁrst illicit drug use (r = -.357, p < .001), and only
a meager non-signiﬁcant correlation between PDRG and AF U (r = .017). If one accepts
the premise that alcohol is a culturally accepted substance then this suggests a cross-level
moderation effect between type of drug (i.e., culturally accepted) and how peer inﬂuence

functions (i.e., as a protective factor or a risk factor).

Third, as was mentioned earlier, the inﬂuence of deviant peers on AFU may be
reduced in Venezuela simply due to attrition. Approximately 68% of youth who start
school in Venezuela do not ﬁnish high school (Mundo, 2003). While no empirical data
that linked drop-out rates to substance use for Venezuela was found, it seems plausible to
assume that youth who become involved with the consumption or trafﬁcking of illicit
drugs may drop out of school at increasingly greater rates than those who choose not to
become involved.

The absence of a signiﬁcant link between PDRG and AF U among this sample of
Venezuelan youth supports the line of reasoning that cultural inﬂuences are a vital
consideration when developing treatment and prevention interventions. For school
attending youth in Venezuela, PDRG may not be a central concern and energies may be
focused on other areas. Still, further research is needed to validate this ﬁnding, and if
replicated, to ferret out the mechanism that may be at work that insulate youth against the
inﬂuence of peers that use illicit substances.

School Climate

The predicted relationship between school climate (SCLM) and AF U was not
signiﬁcant in the model. School climate that allows youth to form an attachment to the
institution has been shown to serve a protective function for youth in US. schools.
School attachment has been deﬁned in the literature as a sense of affection toward and
enjoyment of school (Hill & Werner, 2006) or as a basic expression of the human need to
ﬁt in (Anderrnan, 2002). Most schools in the US. have clubs, sports activities, and other
extracurricular functions that allow students to become involved, and which in turn create

a sense of belonging and acceptance. However, this was not the case in the present study.

156

Only half of the schools were reported to have any extracurricular activities. Just under
half of the students (48.6%) felt that their teachers were genuinely interested in their well
being; 56% reported that the majority of the time they do not do their homework; 30%
reported skipping school two or more times in a month; 15% reported being afraid to go
to school; and 39% reported that their grades were worse than the previous year.
Nevertheless, 75% were happy when they thought about going to school, and only 5%
felt that school was a waste of time.

One possible reason for this ﬁnding may be that the majority of the schools in the
sample suffered from a general lack of resources, which may have detrimentally affected
the school climate. Of the teachers and administrative personal who responded to items
on the MAMBI 83% reported that there are not sufﬁcient desks and chairs for the
students in the classrooms; 86% stated that there was not chalk for the blackboards;
56.6% reported that students didn’t have their own textbooks; 60% reported that there
were broken windows in the building that went without repair; and 68% reported that
doors in building were damaged or off their hinges. What's more, as was previously
mentioned, conversations with administrators and teachers revealed that most of the
teachers in the schools surveyed functioned on an itinerant basis; that is they traveled
from school to school teaching their subject matter as hourly paid workers. One such
teacher reported teaching 10 different courses at 4 distinct schools. It may be that these
factors combine to inhibit the formation of a strong school attachment.

While not included in the current study, another factor that may have inﬂuenced
the school climate variable is the reported adversarial relationship between teachers and

parents (Lodo-Platone, 2004). According to Bronfenbrenner’s coo-developmental theory

157

a mesosystem effect occurs when teachers and parents have poor communication. If
parents perceive themselves in an adversarial relationship with teachers these feelings
will be reﬂected in their interactions with their children, which in turn may exert an
inﬂuence on the student’s attachment with the institution in general, and therefore, inhibit
the formation of an important protective factor.

A third possible reason for not ﬁnding a signiﬁcant link between SCLM and AFU
is the design of the study. The ceiling effect found in SCLM indicates that most students
had formed a bond with the institution as measured by the PACARDO-V. This ceiling
effect could be the result of: (a) the measure may not be valid for Venezuela, or (b) that
the youth who did not form a signiﬁcant bond with the institution were not in the sample
(i.e., had dropped out of school).

The ﬁndings of this study indicate an important disparity between schools as a
context in the US. compared to schools as a context in sampled population in Venezuela.
The combination of poor school conditions, teachers who may not form an institutional
bond themselves, or identify any professional satisfaction from their work at a given
school, and an adversarial relationship between teachers and parents, may deter the
formation of a strong school attachment on the part of students, or at least not in the same
manner in which it is formed in the US. Given that SCLM is an important protective
factor in the US. but did not result as one in the present study, does not necessarily imply
neglecting SCLM as a focus for intervention. The ﬁndings of this study suggest that
school attendance may be a protective factor in and of itself by virtue of the low rates of
illicit substance use among school-attending youth. This being the case, helping students

form an institutional bond may hold the potential to reduce desertion rates, which in turn,

158

would allow them to have contact with an important protective factor. More research is
needed to determine the extent of the effect of SCLM on Venezuelan youth, and if this is
an area to be targeted for intervention.

Gender

The results of the present study agree with previous research and support the
hypothesis that gender (FEM) is signiﬁcantly related to AFU, and that males have earlier
onset of drug use than do females. However, independent sample t tests revealed that the
only drug where there was a signiﬁcant mean difference in age of onset was alcohol with
males beginning use at an average 12.24 (SD = 2.388) years, and females at an average
12.70 (SD = 2.167) years (I = -3.87, p < .001). This is a mean difference of .459, or
approximately 5.5 months. Additionally, a contingency table analysis revealed that males
are not any more likely to use any of the other drugs mentioned in the study than are
females.

While inferences regarding trends cannot be made from the present data, these
ﬁndings would appear to suggest agreement with other research that shows a growing
convergence in rates of use and of age of initiation between males and females (e.g.,
Khoury, Warheit, Zimmerman, Vega, & Gil, 1996). This convergence in drug use
between genders may be related to increasingly fewer restrictions on females in
Venezuela. In the present study, independent t tests showed no signiﬁcant differences
between males and females on the parental monitoring scale. Historically in Latin
America, female children have been monitored by parents more closely than males.

These results suggest that this historic trend may be shifting, and that other factors apart

159

from parental monitoring may be operating to create the small but signiﬁcant gender
difference found in the data. More research is needed to test this hypothesis.
Socioeconomic Status

Studies that have examined the effects of socioeconomic status (SES) on
adolescent substance use have had mixed results. Some researchers have found a positive
relationship between SES and adolescent substance use (e. g., Luthar & Becker, 2002;
Luthar & D’Avanzo, 1999), while others have reported a negative relationship between
SES and adolescent substance use (e. g., Chassin, Presson, Sherman, & Edwards, 1992;
Droomers, Schrijvers, Casswell, & Mackenback, 2003). The present study found no
signiﬁcant effect of SES on AFU and thus the hypothesis that AFU will vary with SES
was not supported. However, MSES, a school level average modeled in the random
intercept, was found to signiﬁcantly covary with AFU (t = -7.2l9, p < .001). These
ﬁndings show that while SES as a characteristic of the individual is not related to AF U,
MSES as characteristic of the school is strongly related to AFU, such that higher levels of
MSES at the school level are related to earlier ages of drug use. Moreover, the effects of
MSES moderated the effects of F AM, EXTB, and FEM creating a cross-level interaction.

Family therapists theoretically look at substance use from a systemic perspective
that includes multiple systems, but in practice they typically limit their interventions to
the conﬁnes of the family. What these results suggest is the need to enlarge the scope of
interventions to include other subsystems that inﬂuence the decision on the part of youth
to engage in substance use. MSES is a variable which, independent of the individual’s
personal resources or of those which the family may possess, inﬂuences AF U through a

sub-systemic effect. While widely recognized by theoreticians as existing (e.g., Becvar &

160

Becvar, 2003) these sub-systemic effects are rarely taken into account in the development
of treatment or prevention strategies. The ﬁndings from this study suggest that such an
effect is not only present, but may exert a more powerful inﬂuence over the decision of
the youth to initiate drug use than that of the family. This point is illustrated in the
graphical representation of the data in ﬁgure 4.6. Here, while holding constant the effects
of externalizing behavior, the line indicating earliest age of onset is low-FAM with high-
MSES (ﬁrst line from the bottom up). If the inﬂuence from the family were the more
prominent covariate of AFU, then one would expect that the next line to be low-FAM
with low-MSES. However, this is not the case. The next line, from the bottom up, is
high-FAM with high-MSES indicating that the effects of MSES may supersede the
inﬂuence of FAM on the adolescent’s decision to initiate drug use. This implies that a
sub-system or culture exists in a given school that is differentially associated with
substance initiation. When assessing for individual risk and protective factors, it is
important to determine to what extent factors associated with the individual are
confounded by environmental conditions. That is, putative individual differences
assumed to be associated with initiation of drug use may actually reﬂect the cultural
inﬂuences of a subsystem in which the individual is nested. This has obvious implications
for the development of prevention and treatment strategies. While not sufﬁcient to make
causal inference, the data analysis from this study do indicate that there is an ecological
effect that covaries with AF U at what Bronfenbrenner denominated the mesosystem, or at
the intersection of the family, peer, and school microsystems. MSES while not rending
non-signiﬁcant the effects of FAM, EXTB, and FEM, does appear to moderate their

relationship with AFU.

161

Discussion of Methodologies: Limitations
Survey Population

An important consideration of this dissertation involves the speciﬁcation of the
survey population in terms of school-attending youths. Youths not in school were
excluded from the sampling frames. The sample did include both private and public
school students, but youth who had already dropped out of school to work, to care for a
child, or for other reasons were not represented. Consequently, it is possible that youths
most seriously affected by drug use are not included in the sample (i.e., those who
dropped out of school because of their drug use, or who began drug use after dropping
out of school). Additionally, schools were selected from only two districts in the western
region of the capital city of Venezuela, Caracas. These limitations are the result of
logistic decisions, cost considerations, and the general purpose of the study being to
collect preliminary data on which to base future projects. As a result, there is a need to
limit inferences to the school-attending youth of the study population. Generalizing the
present ﬁndings to all Venezuelan youth, or even to non-school-attending youth from the

area from which the sample was drawn would be inappropriate.

Measurement Strategies: Use of Self-Reported Data
The self-report assessment of drug use has limitations such as possible over-
reporting or under-reporting by various subpopulations (Beauvais & Oetting, 2002; Vega,
Zimmerman, Warheit, Apospori, & Gil, 1993). For example, in the United States,

increased prevalence of drug use has been described among adolescents that participated

in the Monitoring The Future (MTF) survey compared to the household data collected in
the National Household Survey on Drug Abuse (N HSDA) (Gfroerer, Wright, &
Kopstein, 1997; Wright & Davis, 2001). This discrepancy may be the result of an
unwillingness to admit drug use due to the close proximity of adolescents to their parents
in the NHSDA. However, it is equally possible that the MTF survey approach may have
resulted in inﬂated estimates.

In the present study, in order to ensure that accurate information was collected, a
protocol was developed to engage the youth in the process of data collection, through
creating excitement, a sense of meaning, an interactive mode of delivery, an assurance of
conﬁdentiality and voluntary participation, and a sense of respect for the wishes of their
respective parent or guardian. Given the previously reported low rate of non-response, it
seems that a high level trust was achieved. In addition, efforts were made to identify and
exclude youth who were over-stating their drug use. This was accomplished by including
two items that refer to the fake drug “Cadrina.” Again, the low rates of youth indicating
that they had consumed Cadrina suggest a level of truthfulness in the responses, at least
in the case of possible over-statement of their use.

It is also possible that due a tendency to give socially desirable responses, youth
inaccurately reported their levels on the dependent measure and the covariates of interest.
An effort was made to reorganize and, or separate desks to allow students a greater sense
of privacy. However, this was not always possible given the small physical size of the
classroom and the large number of students present on the day of the survey. To the

extent that students felt that they lacked adequate privacy there may have been under-

163

reporting of their levels of drug involvement and, or their levels on covariates such as
FAM and EXTB.

Confounding is a threat to the validity of this study, and there is a potential for the
mis-speciﬁcation of the model through the exclusion of important covariates
(Raudenbush & Bryk, 2002). Attempts were made to avoid this problem by the inclusion
of the most prominent covariates identiﬁed in the literature. At the same time, there may
be unmeasured confounders that are not accounted for in the models.

Another limitation of this study is that there is not information on other aspects of
parenting practices such as consistency of discipline, use of positive reinforcement, or
history of familial substance uses such alcohol disorders. These constructs may be useful
when considering how important covariates interact to inﬂuence youthful drug
involvement (Ialongo, et al, 2001).

A further concern pertains to the statistical distribution of the covariates. All of
the covariates were either negatively or positively skewed creating strong “ceiling” or
“ﬂoor” effects with large percentages of the sample giving responses that tend toward the
highest or lowest possible scores for that construct. Consequently, it is difﬁcult to
distinguish youths whose scores are in the more extreme end of the spectrum. For
example, FAM was negatively skewed and had a strong “ceiling effect” whereby
approximately 25% of the sample gave responses that produced the highest score possible
on this construct. As such, it is difﬁcult to distinguish youths whose parents have more
coercive parenting styles. To the extent that these limitations are associated with higher
levels of drug involvement, the ﬁndings would be skewed to null values and might tend

to underestimate the association between these covariates and youthful drug involvement.

164

Lastly, it is possible that the measures are not properly capturing the constructs
they are intended to measure. The original PACARDO project questionnaire from which
the PACARDO-V was adapted for use in Venezuela is based on the translation of a North
American, English language instrument; it is possible that some items have not
“translated” well or do not capture adequately the cultural reality of this population. At
the same time, a great deal of effort was made to develop an instrument that was not only
language appropriate but culturally appropriate as well. The input of collaborators from
each country in the original PACARDO project was incorporated into the development of
the instrument, and both the PACARDO and the PACARDO-V were piloted prior to

assessment.

Model Specification and Temporal Sequencing

The ﬁndings of this study are based upon cross-sectional survey data and a
limited set of assumed covariates of AFU. Necessarily, not all pertinent constructs could
be measured within the conﬁnes of a dissertation study with the previously mentioned
logistical and ﬁnancial constraints. Therefore, some of the omitted constructs might be
confounding the observed associations. Furthermore, the small sample size at level-2 in
the multilevel model potentially reduced the amount of variance between schools and
thus placed constraints on the degree to which covariates could be modeled at this level.

Additionally, the cross-sectional nature of the data precludes the temporal
sequencing necessary for causal inferences. For example, it is possible that parents have
“given up” or retreated on their involvement with their child as a result of their drug use

instead of the inverse.

165

Discussion of Methodologies: Strengths

Notwithstanding limitations such as the ones mentioned, the study has a number
of counter-balancing strengths. First this study employs a rigorous survey and associated
methodology that was previously used in seven other Latin American countries, and adds
to the body of existing evidence provided by organizations such as Monitoring the Future
(MTF), the European School Survey Project on Alcohol and Other Drugs (ESPAD), and
Global Youth Tobacco Survey (GYTS).

Second, the study uses an ecological framework that provides empirical data in
support of the need for cultural adaptations in the areas of testing and measurement and
treatment and prevention intervention development. Speciﬁcally, the ecological
framework allows for the understanding of the data from the perspective of a highly
segregated society whose clustering effect will inherently produce different contexts and
proﬁles of risk and protective factors among its inhabitants. These observations offer
valuable insights to both researchers and practitioners interested in culturally sensitive
treatment of youth across the spectrum of diverse ethnicities.

Third, the present study employed a multilevel modeling approach that accounted
for the clustering effect of students nested within schools. It is important to note that
while the preliminary analysis found SES, peer drug inﬂuences (PRDG), and school
climate (SCLM) to be signiﬁcant covariates of age of ﬁrst drug use (AF U), the
hierarchical linear model found just the opposite. SES, PRDG and SCLM did not
signiﬁcantly covary with AFU at the individual level. However, the relationship between

SES and AFU was signiﬁcant as a second level school characteristic in the form of

166

MSES (Mean SES). This underscores the importance of accounting for the nesting in the
data when planning for an appropriate statistical tool. While it is beyond the scope of this
dissertation to challenge the ﬁndings of other studies that did not account for the nesting
of the data, other researchers have. For example, Baldwin, Murray, and Shadish (2005)
reexamined the ﬁndings of 33 studies reporting on treatments that had been identiﬁed as
empirically supported treatments (EST) by an American Psychological Association
commissioned task force. They found that, depending on what assumptions were made
about how large the dependencies among the observations truly were, from 6 tol9 of the
studies no longer had any signiﬁcant results after correcting for the violation of the
assumption of independence of observations. As was mentioned in chapter four, even
though the variance between schools was only approximately 6% in the present study, it
is still important to model that variance.

Another related point, and a strength of the methodology, is that the multilevel
approach used in this study allowed the slope of FAM to vary randomly. This provided
information regarding how the effect of FAM on AFU varied from negligible in some
contexts to highly signiﬁcant in others. While an empirical explanation for this variance
was not obtained given the small sample size, this discovery provides evidence for a
mesosystemic effect described in Bronfenbrenner’s coo-developmental theory, and holds
important implications for treatment and prevention strategists to look for family by
context interactions when planning interventions.

Fourth, the PACARDO-V survey is one of the ﬁrst studies to look at risk and
protective factors in Caracas, Venezuela and provides preliminary data on which to

develop future research and to guide prevention and treatment strategies. The study has a

167

relatively large sample size that allows detectable differences to be found at the school
level. It was conducted with extensive involvement of Venezuelan nationals who all had
prior experience working with adolescents in a school setting, and who were very

involved in assessing that the instrument was culturally appropriate.

Implications for Treatment and Prevention

Identifying risk and protective factors of adolescent substance use to guide
prevention and treatment efforts as well as research has become the primary research
paradigm in the ﬁeld (Vega & Gil, 1998). Many factors that constitute either risk or
protection for youth have been established empirically in the substance abuse literature.
The risk and protective factors approach provides researchers and human service
professionals with clear targets for the focus of their efforts as well as identiﬁes areas that
will not produce fruitful results. An important rationale for undertaking the present study
has been the identiﬁcation of communalities and differences among Venezuelan
adolescents in order to inform the process of cultural adaptation of family-based
prevention and treatment models. The results of this study have the following
implications.

First, the ﬁndings of the present study indicate that the mean age of ﬁrst drug use
is approximately 12 years for both males and females. Research in the US. suggests that
before middle school (approximately 12 years of age) parents seem to hold the greatest
inﬂuence over youth, with peers exerting an increasingly greater inﬂuence as the
individual moves through adolescence (Cummings, 1995). Since F AM was signiﬁcantly

related to AFU in the present study, this suggests that prevention strategies aimed at

168

parents would need to target youth and their families at or before 12 years of age to be
most effective. Additionally, strategies aimed at peer groups might be most effective if
conducted in mid to late adolescence.

Second, the relationship between F AM and AF U suggests that interventions
targeting parenting behavior will be effective in delaying the onset of substance use. If
future research supports that parental support more than parental monitoring delays onset
of substance use among Venezuelan youth, then interventions that add an affective
component to behavioral oriented interventions may yield greater results for this
population. Additionally, the variation among schools in the relationship between FAM
and AF U suggests that context must be an important consideration in the development
and testing of any family-based intervention.

Third, as the adolescent begins to struggle with separation from parents and
individualization, conforming to peer group norms becomes increasingly important
(Lerner, Petersen, & Brooks-Gunn, 1991). Since peer inﬂuences were not signiﬁcantly
related to AFU, they may not be technically considered to exercise risk or protective
functions in the sample. However, given the numerous studies that have identiﬁed peer
inﬂuences as being predictive of substance abusing behavior in adolescents, it is
interesting that in this sample they were not. In fact, some research has found that peer
inﬂuences may exert a protective function on early use of drugs (Apospori, et al, 1995)
among certain groups. This may also be the case with the present sample. In an item that
asked youth if many of their friends think it is a bad idea to use drugs, 85.2% responded
in the afﬁrrnative. This suggests that the relationship within peer groups in Venezuela

may constitute a “neutral” area that could be swayed toward exerting a protective

169

function with appropriate prevention interventions. This potential area for intervention is
further supported by the ﬁnding that schools with higher MSES had earlier age of ﬁrst
drug use, suggesting that a school culture exists that inﬂuences youth beyond individual
characteristics. Although the present study did not test for an interaction between peer
inﬂuences and MSES, it is logical to assume that a mechanism through which the culture
created by MSES is communicated is through peer groups.

Fourth, the ﬁnding that MSES is inversely related to AF U points to a hypothesis
that as schools become more exclusive, AFU decreases and substance use increases.
Future research that includes greater variability between schools will have to conﬁrm or
reject this hypothesis. However, if true, MSES will be an important identiﬁer for high-
risk populations for which speciﬁc interventions may be developed. Similarly, EXTB
was identiﬁed as a strong covariate of AF U and can be used to identify at risk
populations for more speciﬁc and intensive intervention techniques.

Fifth, school climate did not show a signiﬁcant relationship with AF U in the
analysis. However, as was previously mentioned, the ceiling effect witnessed in the
distribution of SCLM may have resulted in a tendency toward the null, in that the vast
majority of schools was equally lacking in resources and maintenance of the physical
plant. If this is the case, then these results may still hold suggestions for interventions.
Given the overall lack of extracurricular activities, a community level intervention may
be to ﬁnd ways to involve community members to develop and implement extra
curricular activities for youth, such as sports, business clubs, drama clubs, etc.
Additionally, students could be motivated to participate in school beautiﬁcation activities

and creations of artwork that would stimulate a sense of pride and belonging.

170

Finally, given that evidenced-based treatment and prevention models exist, an
important concern for human service professionals and researchers working with ethnic
minorities or international populations is whether these models need to be adapted, and if
so, to what extent. A primary purpose of this research was to test for the presence of
known risk and protective factors in an international population to lay the groundwork for
the subsequent implementation of prevention and treatment models. The ﬁndings suggest
there are similarities and differences between the sample and what is typically found
among adolescents in the US. The absence of a relationship between peer group
inﬂuences and AFU is a ﬁnding that clearly deviates from the majority of research in the
ﬁeld and one that suggests substantial changes in the design of interventions. The
contextual variations that moderate the relationship between FAM and AFU is another
difference from how this variable has been shown to work among youth in the US. and is
also suggestive of adaptations to interventions. The role of MSES in AF U has not
received much attention in the published literature in the US. and therefore introduces
the possibility of a further area for cultural adaptation to established interventions. The
low level of illicit substance use among Venezuelan youth suggests the presence of
naturally occurring protective factors that must be accounted for and exploited in
prevention and treatment models.

One family-based model that has adopted an ecological approach and accounts for
cultural differences in the treatment of adolescent behavior problems including substance
abusing behavior is “Familias Unidas” (Tapia, Schawartz, Prado, Lopez, & Patin, 2006).
F amilias Unidas is a multilevel approach that strategically targets risk and protective

factors such as parent involvement, peer groups, and school bonding to prevent the onset

171

of youthful behavior problems. The current study offers empirical evidence that informs
the adaptation of models such as F amilias Unidas to the Venezuelan culture.

Together these ﬁndings suggest the need for a culturally adapted approach in the
planning and implementation of treatment and prevention strategies for Venezuelan
youth. The complexities introduced by the differing nuances of meaning created by
culture and other environmental factors necessitate an ecological approach to adolescent
substance use. Researchers and practitioners that ignore these nuances run the risk of
missing the mark and wasting precious resources that will ultimately result in the

perpetuation of human suffering.

Implications for Further Research

While this study provides a useful description of the level of youthful drug
involvement in this particular region of Caracas and its demographic characteristics, the
cross-sectional design provides only a “snapshot” in time. In order to obtain information
on how the identiﬁed covariates interact to effect the development of youthful drug
involvement over time a longitudinal design is necessary.

Alternative sampling strategies that allow for inclusion of non-school attending
youth would allow for a more deﬁnitive picture of youthful drug involvement in the
country. However, samples of school-attending youth are still important. The great
likelihood that these youth willomature into the private sector and public sector leaders of

the future necessitates a comprehensive understanding of the factors that inﬂuence their

development. Moreover, comparing and contrasting the dynamics between school

172

attending and non school-attending youth holds the potential to elucidate individual, and
environmental factors that inﬂuence developmental processes.

Future research in this area should also seek to increase the number of participants
and the number of schools in the sample and to diversify the sample across
socioeconomic and geographic strata. Given the low incidence of illicit drug use, a larger
sample size at level-1 is needed in order to model illicit drug use and compare the
characteristics of youth who use illicit drugs to those who only consume alcohol and
cigarettes. Additionally, increased amounts of schools from differing levels of SES and
from rural and urban environment would provide the ability to model second level
characteristics that may exert an inﬂuence over level-1 covariates as was seen in the case
of MSES and FAM in the present study. Quantitative estimates gained from the present
study provide evidence of a clustering effect such that the occurrence of drug experiences
in one youth within a school is increased when other youths in the same school are
involved with drugs. Findings of this type conﬁrm the need to identify school-level
factors that contribute to student drug use and in order to develop school-based
prevention efforts (Delva Bobashev, Gonzalez, Cedeno, & Anthony, 2000).

Qualitative research or mixed-method designs that included the use of focus
groups and consultation with psychologists, parents, youth, and educators from differing
regions of the country would aide in the reﬁnement of a more robust family attention
construct. This could include a more vigorous multidimensional measure that includes
items tapping differing facets of parenting such as consistency of discipline, positive
reinforcement, family cohesion, and child rejection. A more robust multidimensional

construct would allow for a more complete understanding of the mechanisms involved in

173

the inﬂuence that family attention might be playing in relation to youthful drug
involvement. The same method could also be used to gain increased understanding
regarding the school climate variable and the apparent adversarial relationship between
teachers and parents.

Additionally, research is needed to establish the potential interaction between
parental monitoring and parental support. It may be that there is a curvilinear relationship
between monitoring and support such that monitoring is only signiﬁcant up to a point and
then becomes non signiﬁcant or detrimental if not accompanied by support. Therefore, in
regards to the role of parental monitoring, more research is needed to ferret out the
idiosyncrasies of how this variable works in the Venezuelan culture.

Another area of research is the nature of the adversarial roles between teachers
and parents. This area may be of particular concern as it relates to school desertion rates.
Research has established that in the US, students who drop out are more likely to use
drugs (Wallack & Corbett, 1990). In fact, school drop-outs have been shown to have rates
of tobacco use 79% higher than their school-attending peers (Pirie, Murray, & Luepker,
1988). If school is shown to be a protective factor buffering against the early onset of
substance use then cultivating the relationship between teachers and parents may be an
important area of intervention.

Finally, the hypothesized naturally occurring protective factors that buffer against
the early onset of substance use among school-attending youth is an intriguing area for
future research, and one that holds considerable potential to inform both theories
regarding adolescent substance use as well as models that seek to reduce the number of

youth who engage in this behavior.

174

Concluding Remarks

F arnily therapy practitioners and prevention specialists rely on sound empirical
data to formulate and deliver their interventions. While the ﬁeld of adolescent substance
use has made many gains over the past years, much is still left to be learned. The present
study is illustrative of the promise that cross-national research holds to reveal aspects of
environmental inﬂuences that otherwise may be difﬁcult to discover by those embedded
in the same system they seek to observe (Keeney, 1983). The ﬁndings provided by this
study underscore the importance of culturally adapted interventions that carefully assess
for intergroup and intragroup variations created by social contexts. Individuals form
different cultures experience their process of socialization in systematically different
ways that should be the focus of research and intervention. Family interaction patterns,
structure, and social contexts work together to form an ecosystem that creates meaning
for its members, and which in turn governs behavior through recursive feedback loops.
Risk and protective factors that inﬂuence adolescent substance use are abstract constructs
that researchers use to attempt to capture and measure these interactional sequences and
social environments.

This study identiﬁed family attention, externalizing behavior, gender, and mean
socioeconomic status as salient covariates of age of ﬁrst drug use. It also revealed
characteristics of Venezuelan society as they pertain to adolescent substance use. While
far from comprehensive, this study advances the literature on adolescent substance use in
several important ways: (a) by being one of only a handful of studies that have assessed

substance use in the country of Venezuela; (b) by providing empirical data that shed light

175

on the importance of culture in the design and delivery of treatment and prevention
strategies; and (c) by suggesting new areas of research for both domestic and
international researchers. Finally, this study has shed light on what public health planners
and human services providers call for most desperately: clearly identiﬁed targets that are

amenable to prevention and treatment activities.

176

APPENDIX A

PACARDO-V 2007

ESTE CUESTIONARIO ES ANONIMO, POR ESO NO DEBES ESCRIBIR TU NOMBRE EN
NINGUNA PARTE.

IN STRUCCIONES GENERALES:

Lo que vas a realizar es facil y no necesitas haber estudiado para contestar las preguntas. No es
un examen y por lo tanto no hay respuestas buenas ni malas, solo tus percepciones. Lo
importante es que las repuestas que des, sean verdaderas.

Si te surge alguna duda o si quieres hacer algr’rn comentario, nosotros con gusto atenderemos tus
inquietudes, solo alza la mano. Por favor, no dudes en preguntamos si 10 necesitas.

La informacion que provees en este cuestionario se manejara anonimamente, dado que su
ﬁnalidad es conocer opiniones y caracteristicas de grupos de jévenes estudiantes para el
desarrollo y aplicacion de programas preventivos.

Hay cinco secciones con un total de 112 preguntas. Por favor, responde a todas las preguntas.
Aunque las respuestas no se ajusten exactamente a tu experiencia, marca la respuesta que te
parece mas acertada, o que es mas cercana a tu experiencia. Por ejemplo, marca <SI>, si la
mayoria de las veces es cierto, o <NO>, si la mayoria de las veces es falso. Si alguna pregunta te
hace sentir incomodo(a), puedes marcar la opcion <Sin Responder> (SR).

Vamos a estar leyendo en voz alta las preguntas, asi que nadie debe adelantarse o ir a la préxima
pagina hasta que se indique. Esto nos ayudara a terminar mas rapido y estar seguros que nadie se
confunda 0 se pierda.

oAlguien tiene alguna pregunta?

INSTRUCCIONES PARA EMPEZAR:

1. Saca la hoja de respuesta (una hoja con muchos circulos).

Usa el lapiz que esta dentro del sobre para llenar la hoja de respuestas.

Asegt’rrate de llenar los circulos completamente y de no hacer marcas fuera de circulo. Si

te equivocas, asegurate de borrar la marca equivocada completamente. La maquina no

puede leer dos marcas.

4. Coloca el numero que esta en tu sobre en el cuadro llarnado PID (en la parte inferior
izquierda de la hoja). Pon los numeros en los cuadros y llena el circulo correspondiente
debajo de cada numero.

5. Donde dice nombre (last name) coloca las primeras 5 letras del nombre del colegio sin
dejar espacios. Por ejemplo, 51 e1 colegio se llama “San Agustin”, coloca en los cuadros
“sanag”. Y llena los circulos correspondientes debajo de cada letra.

DJN

177

6.
7.
8.

No pongas tu nombre en ninguna parte de la hoja.
Donde dice fecha (date) anota tu edad.
Ve al numero 1 para comenzar.

Seccién 1: Preguntas qenerales

1.
2.

10.

11.

12.

13.

Tu posicién en el colegio (A) Estudiante

Tipo de escuela: (A) Privada (B) PL'Iinca

Tu Edad: (A) 11-12 (B) 13-14 (C) 15-16 (D) 17-18 (E) 19+

Aﬁo Escolar: (A) Séptimo/ 1er 000 (B) Octavo/Zdo ar’io (C) Novena/3er aﬁo
(D) 1° Diver/4‘10 aﬁo (E) 2° Diver/510 ar’io

éCon qué raza te identificas mds? (A) Negro (B) Morena (C) Blanca
D) Indigena

éDénde vives? (A) barrio/bloques/casa (B) urbanizacién/apartamento,
(C) quinta

éCuéntos vehiculos tiene tu familia? (A) O (B) 1 (C) 2 (D) 3 (E) M65 de 3

éCuc’mtas personas viven en tu casa? (A) 1—3 (B) 4-6 (C) 7-8
(D) 9-10 (E) 10+
éCudntas dormitorios tiene tu casa?
(A) 0 es un solo ambiente (B) 1 (C) 2-3 (D) 4-5
(E) 6+
éCudl es tu sexo? (A) Varén (B) Hembra
clA cucil religio’n perteneces? (A) Catélica, (B) Cristiano no Catélica (C)
Musulmana (D) Otra (E) Ninguna

éQué grado académico tiene tu padre (6 la persona que es como tu padre)? (A)
Primaria (B) No terminé Secundaria (C) Termino’ Secundaria (D) No
terminé estudios superiores (E) Terminé estudios superiores

éQue’ grado académico tiene tu madre (6 la persona que es como tu madre)? (A)

Primaria (B) No termino Secundaria (C) Terminé Secundaria (D) No
termihé estudios superiores (E) Terminé estudios superiores

178

Seccio’n 2: Prequntas sobre diferentes aspectos de la vida del joven

venezolono

Por favor, responde todas las preguntas, aunque las respuestas no se ajusten
exactamente a tu experiencia. Marca <SI>, si la mayoria de las veces es cierto, o <NO>,
si la mayoria de las veces es falso. Marca <SR> si la pregunta te incomoda y no quieres
responder. No debes dejar ninguna pregunta sin responder. Marca la respuesta
correspondiente llenando completamente el circulo en la hoja de respuestas.

 

 

 

 

 

 

14. c'Tus padres o representantes saben lo que piensas o sientes Si No SR
sobre las cosas importantes para ti? (A) (B) (C)
15. (Tus padres o representantes, hon estado conscientes de lo Si No SR
que te gusto o no te gusto? (A) (B) (C)
16. c'Siempre pides permiso a tus padres cuando sales de la Si No SR
casa a divertirte? (A) (B) (C)
17. (Has sentido que eres importante para tus padres a Si No SR
representantes? (A) (B) (C)
18. (Generalmente tus padres o representantes hon estado Si No SR
enterados de dénde estds y qué haces? (A) (B) (C)
19. A veces, los padres les dicen a sus hijos que no se junten con
personas que se meten en problemas. cTe hon dicho tus Si No SR
padres o representantes que no te juntes con personas que (A) (B) (C)

puedon meterte en problemas durante el tiltimo aﬁo?

20. A veces los jévenes regreson a casa después de lo escuelo y
no encuentron a nadie. c'Hon estado en casa tu popd, tu mamc’r Si No SR
0 algun otro odulto cuando has regresodo a casa después de (A) (B) (C)
closes o trabojo, durante el tiltimo aﬁo escolor?

21. c'AlgL’rn miembro de tu fomilio que vive en casa (Madre,
Padre, Hermano(a)), ho fumado cigarrillos durante el ultimo
oﬁo?

22. (LAlgtin miembro de tu familio que vive en casa (Madre,
Padre, Hermano(a)) ha consumido bebidos alcohélicas
durante el ultimo oﬁo?

23. (LAlgLin miembro de tu familia que vive en casa ho tornado
alcohol hasta el punto de cousor problemas en la casa, en el
trabajo a con los amiqo(o)s durante el liltimo 060?

24. cAlgdn miembro de tu familia que vive en casa (Madre,
Padre, Hermono(a)) ha consumido alguna drogo ilegol como la

 

 

Si No SR
(A) (B) (C)

 

Si No SR
(A) (B) (C)

 

Si No SR
(A) (B) (C)

 

Si No SR

 

 

 

 

 

 

 

 

morihuona, cocoina, etc. durante el tiltimo an'o? (A) (B) (C)
25. c'Con frecuencia has tenido discusiones con tus padres que Si No SR
hon terminado a gritos? (A) (B) (C)
26. Mis padres siempre me estdn hablando sobre lo daﬁino que Si No SR
son las drogas. (A) (B) (C)
27. Mis padres siempre me estcin hablando sobre la doﬁino que Si No SR
son el alcohol y los cigarrillos. (A) (B) (C)

 

179

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

28. éAlguno(a) de tus omigo(a)s se ha metido en problemas con Si No SR
la policia? (A) (B) (C)
29. cTus amigo(a)s han robado, o hon causado daﬁo a propésito a Si No SR
las cosas de otras personas? (A) (B) (C)
. . . . Si No SR
30. Alguno(a)s de mis amigo(a)s fuman Cigarrillos. (A) (B) (C)
. . . . Si No SR
31. Mucho(a)s de mis amigo(a)s fuman crgarrillos. (A) (B) (C)
32. Algunos jévenes piensan que es una buena idea usar drogos y
otros piensan que es uno mola idea usar drogos. cTienes Si No SR
mucho(a)s amigo(a)s que piensan que usar drogos es una mola (A) (B) (C)
idea?
. . . Si No SR
33. Alguno(a)s de mis amrgo(a)s han fumado marrhuana. (A) (B) (C)
34. (Has tenido amigo(a)s a quienes Ies gusto inhalar pegamento Si No SR
0 gasolina? (A) (B) (C)
35. Algunos jévenes hon comenzado a usar bozuco, cocaina o .
crack. c'Tienes algt’m amigo que haya usado bozuco, cocaina, 5' NO 5R
0 crack? (A) (B) (C)
36. cTienes varios(a)s amigo(a)s que usan bozuco, cocaina, 0 Si No SR
crack? (A) (B) (C)
37. (Se hon "jubilado" del colegio/liceo mucho(a)s de tus Si No SR
amigo(a)s? (A) (B) (C)
. . . Si No SR
38. cHas pertenecrdo a alguna banda o pandilla? (A) (B) (C)
. . . 9 Si No SR
39. cTe en0jas con frecuencra. (A) (B) (C)
40. Has daﬁado intencionalmente las cosas de otras personas Si No SR
durante el tiltimo aﬁo escolar. (A) (B) (C)
. , . .. Si No SR
41. cHas robado algo durante el ultimo ano escolor? (A) (B) (C)
. . . , . ., Si No SR
42. cHas hecho algo rresgoso o peligroso durante el ultimo ono? (A) (B) (C)
43. (LES cierto que la mayoria de las veces no haces las tareos Si No SR
del colegio? (A) (B) (C)
44. He tenido excelentes relaciones con la mayoria de mis Si No SR
profesores. (A) (B) (C)
45. (Has sentido temor o miedo al ir al colegio/liceo durante el Si No SR
Ultimo aﬁ'o escolor? (A) (B) (C)
46. Siento que la mayoria de mis profesores se interesan en mi Si No SR
sinceromente. (A) (B) (C)
47. éTe has "jubilado" de la escuelo dos dias 0 mos en un solo Si No SR
mes durante el dltimo aﬁo escolor? (A) (B) (C)

 

 

 

 

180

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. . . . . 9 Si No SR

48. cHos sndo suspendrdo(a) del colegio/liceo. (A) (B) (C)

. .. 9 Si No SR

49. cHon empeorado tus notas este ano. (A) (B) (C)
50. Algunos jévenes se sienten contento(o)s cuando piensan en ir .

. . . . Sr No SR

al colegio/liceo. cEn general, te has sentido contento(a) al (A) (B) (C)

pensar en ir al colegio/liceo, durante el L’iltimo aﬁo escolor?

. . Si No SR

51. He pensado en abandonar el colegio/liceo completamente. (A) (B) (C)
52. A veces la gente joven dice que "ir al colegio/liceo es una .

, . . ,, . . . , . . . Sr No SR
perdrda de trempo . cPara t1 ha srdo uno perdrda de tiempo Ir (A) (B) (C)
a| colegio o liceo durante el ultimo aﬁo escolor?

53. En forma general, cse ayudan entre si las personas en tu Si No SR
vecindario o urbanizacién? (A) (B) (C)
54. Cuando un joven hace algo malo, a veces los vecinos le
cuentan a su representante. cEn tu vecindario 0 Si No SR
urbanizacién Ios vecinos le cuentan a los representantes (A) (B) (C)
cuando un joven hace algo malo?
55. (Es comun oir que alguna persona fue agredida por la Si No SR
delincuencia en tu vecindario o urbanizacién? (A) (B) (C)
56. Es comt’rn ver a personas usando o vendiendo drogos en tu Si No SR
vecindario o urbanizacién. (A) (B) (C)
57. Frecuentemente veo personas borrachas o drogadas en las Si No SR
calles de mi vecindario o urbanizacién. (A) (B) (C)
58. Me siento seguro cuando camino solo(a) en mi vecindario 0 Si No SR
urbanizacién. (A) (B) (C)
59. Las personas que viven en tu vecindario o urbanizacién Si No SR
frecuentemente dai‘ian o roban la propiedad de otros. (A) (B) (C)

 

Seccio’n 3: Preguntas sobre el efecto y accesibilidad de las drogos

Por favor, responde a todas las preguntas, aunque las respuestas no se ajusten
exactamente a tu experiencia. Marca solo una de las letras <A, B, C, 0 D>, segun se
ajuste mejor a tu experiencia. Marca <(E) N0 Sé> solo si la pregunta te incomoda y no
quieres responder. No debes dejar ninguna pregunta sin contestar. Marca Ia respuesta
correspondiente llenando completamente el circulo en la hoja de respuestas.

 

 

 

60. <1 ue’ ries o corre la ente a
8 . 9 . . 9 (A) (B) (C) (b) (E)
per judrcarse (fisrcamente 0 de . , .

. Nrngun Leve Mediano Gran No
otras maneras),sr fuma alrededor ries o ries o ries o ries o sé
de una caja de cigarrillos por dia? g g 9 9

61. Para ti y tus amigo(a)s, cqué ton (A) (C) (D) (E)
fcicil o dificil es conseguir , . (8) chil . , . Muy No

. . Muy facrl DIfICil . , . .
crgarrillos? dIfICII se

 

 

 

 

 

 

 

181

 

62. {Qué riesgo corre la gente a

 

 

 

 

 

 

 

 

 

 

 

perjudicarse (fisicomente 0 de (A) (B) (C) (D) (E)
otras maneras), 31 de vez en Ningun Leve Mediano Gran No
cuando consume una o dos bebidos riesgo riesgo riesgo riesgo sé
alcohélicas?
63. éQué riesgo corre la gente a
perjudicarse (fisicamente a de (A) (B) (C) (D) (E)
otras maneras), si consume Ningrin Leve Mediano Gran No
bebidos alcohélicas riesgo riesgo riesgo riesgo sé
frecuentemente?
64. Para ti y tus amigo(a)s, équé tan (A) (B) ch” (C) ”(\D) S:
fdcil o dificil es conseguir alcohol? Muy fdcil Dificil . 9y. ,
drficrl se
65. c ué ries o corre la entea
Q . 9 . . 9 (A) (B) (e) (D) (E)
per judrcarse (frsrcamente 0 de . , .

. Nrngun Leve Mediano Gran No
otras maneras), sr consume crack 0 ries o ries o ries o ries o sé
bazuco de vez en cuando? 9 9 9 g

66. (L ué ries o corre Ia entea
9 . 9 , . 9 (A) (B) (c) (0) (E)
perjudrcarse (fiSicamente a de . , .

. Nrngun Leve Mediano Gran No
otras maneras), sr consume crack 0 ries o ries o ries o ries o sé
bazuco frecuentemente? g g 9 9

67. Para ti y tus amigo(a)s, équé ton (A) (C) (D) (E)
fc’rcil o dificil es conseguir crack a , . (B) Fc’rcil . , . Muy No
Muy facrl Difrcrl . , . ,
bazuco? drfrcrl se
68. Para ti y tus amigo(a)s, cQué tan (D) (E)
, . . . . . (A) . . (C)
focrl o drfrcrl es conseguir Mu fcicil (B) Facrl Dificil Muy No
Ecstasy? y difr’cil sé
69. c ué ries o corre la entea
peerudicarsg (fisicomengte 0 de (A) (B) (C) (D) (E)

. Ningrin Leve Mediano Gran No
otras maneras), sr consume ries o ries o ries o ries o sé
Ecstasy de vez en cuando? 9 9 9 9

70. c ué ries o corre Ia entea
perjc'irdicarsz (fisicomengte 0 de (A) (B) (C) (D) (E)

. Ningun Leve Mediano Gran No
otras maneras), sr consume ries o ries o ries o ries 0 se’
Ecstasy frecuentemente? 9 g g 9

71. Para ti y tus amigo(a)s, cqué ton (A) (C) (D) (E)
facil o dificil es conseguir Mu fécil (B) chil Dificil Muy No
inhalantes como pegamento, etc? y dificil sé

72. cQué riesgo corre la gente a
perjudicarse (fisicamente 0 de (A) (B) (C) (D) (E)
otras maneras), si consume Ningun Leve Mediano Gran No
inhalantes de este tipo de vez en riesgo riesgo riesgo riesgo sé

cuando?

 

 

 

 

 

 

182

 

 

 

73. 6Que' riesgo corre la gente a

per judicarse (fisicamente 0 de (A) (B) (C) (D) (E)
otras maneras), si consume Ningdn Leve Mediano Gran No
inhalantes de este tipo riesgo riesgo riesgo riesgo se’
frecuentemente?

 

74. CQUé riesgo corre la gente a
per judicarse (fisicamente 0 de
otras maneras), si fumo marihuana
de vez en cuando?

(A) (B) (C) (D) (E)
Ningtin Leve Mediano Gran No
riesgo riesgo riesgo riesgo se

 

75. c'Qué riesgo corre la gente a
perjudicarse (fisicamente 0 de

(A) (B) (C) (D) (E)

 

 

 

 

 

 

 

. . Ningtin Leve Mediano Gran No
otras maneras), sr fuma marrhuana ries o ries o ries o ries o sé
frecuentemente? g g g g

76. Para ti y tus amigo(a)s, c'qué tan (A) (C) (D) (E)
fdcil o dificil es conseguir , . (B) Fcicil . . . Muy No

_ Muy facrl DIfICIl . , . .
marrhuana? dIfICII se

 

 

Seccién 4: Preguntas sobre diferentes aspectos de la vida del joven

venezolano en te’rminos de frecuencia.
Por favor, responde a todas las preguntas, aunque las respuestas no se ajusten
exactamente a tu experiencia. Marca solo las letras <A, B, C, D, o E>, segun se ajusta
mejor a tu experiencia. N0 debes dejar ninguna pregunta sin contestar. Marca la
respuesta correspondiente llenando completamente el circulo en la hoja de respuestas.

 

Muy De

 

. . Varias

, , NI Una Vez en Varias . .

Las srguientes preguntas se Veces A Diarro
, , , ~ Solo Vez Cuando Veces al ,
refieren ol ultimo ano. N COdCl O M05
Este Ano Durante el Mes
.. Semano
Ano
77. Ver televisién o 'u or
J 9 A B C D E

video juegos en casa.

 

78. Practicar algun deporte
como ft'rtbol, béisbol, A B C D E
escalar la montah‘a, etc.

 

79. Tocor un instrumento

 

 

 

 

 

. A B C D E
musrcal

80. Traba jar para ganar A B C D E
drnero

81. Ir de citas o cortejor A B C D E

82. Fumar cigarrillos A B C D E

83. Pasar trempo con mi A B C D E
fomrlra

84. Apostar par dinero A B C D E

 

 

 

 

 

(dados,cabaHos,

 

 

cartas, etc.)

 

85. Rezar, oror o leer la

 

 

 

 

 

Biblia A B C D E
86. Hacer tareos en la casa
como cocinar, limpiar, A B C D E
etc.
87. Fumar marihuana A B C D E
88. Ir a actividades de la
iglesra como closes, A B C D E
actrvrdades socrales,
retiros, etc.
89. Ir a conciertos de
mtisica rock, reguetén, A B C D E
etc.
90. Consumir alcohol, A B C D E

cerveza, anis, etc.

 

91. Hacer tareos y estudiar
para los excimenes del A B C D E
colegio/liceo

92. Consumir crack, bozuco,

 

 

 

 

, A B C D E
o cocarna
93. Salir a bailar A B C D E
94. Salir con amistades al
cine o a pasear en los A B C D E
centros comerciales
95. Consumir heroina A B C D E

 

 

 

 

 

 

 

 

Seccién 5: Preguntas sobre la edad en que hubo la primera oportunidad de

consumo y la edad del primer consumo.

Voltea de nuevo la hoja de respuestas hasta el [ado 1 (SIDE I). A La mano derecha
superior verds unas cajas verticales con circulos adentro. Estos circulos solo contienen
numeros. En esta seccio'n todas las respuestas sertin edades. La primera columna de
numeros dentro de los circulos, corresponde a la posicién de las decenas y la segunda
columna a la posicién de las unidades. Por ejemplo si tu respuestafuese 7, colocarz'as 0
en la primera columna y 7 en la segunda; si tu respuestafuese 16, colocarias 1 en la
primera columna y 6 en la segunda. Si aun no has tenido oportunidad de hacer lo que
dice la pregunta, colocarias 0 en la primera columna y tambie'n 0 en la segunda
columna. Marca la respuesta correspondiente llenando completamente el circulo en la
hoja de respuestas. Por favor, responde a todas las preguntas, contestando lo que mds se
ajusten a tu experiencia.

 

 

 

1. Acerco del cigarrillo. (LA qué edad tuviste tu primera oportunidad
. . Edad:
de fumar crgarrillo?
2. (A qué edad probaste cigarrillo por primera vez? Edad:

 

 

 

184

 

3. Acerca de las bebidos alcohélicas. (LA que’ edad tuviste tu primera

 

 

 

 

 

oportunidad de consumir bebidos alcohélicos? Edad:
4. (LA qué edad probaste alcohol por primera vez? Edad:
5. Acerco del crack 0 bozuco. (LA que’ edad tuviste tu primera
. . Edad:
oportunidad de consumir crack 0 bozuco?
6. (LA qué edad probaste crack 0 bozuco por primera vez? Edad:
7. Acerca de Ecstasy, (LA que’ edad tuviste tu primera oportunidad de E do d'
consumir Ecstaﬂ? .
8. (LA que’ edad probaste Ecstasy por primera vez? Edad:

 

9. Acerco del pegamento del zapatero, gasolina, éter, u otras
sustancias inhalantes. (LA que’ edad tuviste tu primera oportunidad Edad:
de consumir inhalantes de este tipo?

 

10. (1A qué edad probaste inhalantes por primera vez? Edad:

 

11. Acerca de la drogo Cadrina (LA qué edad tuviste tu primera

 

 

 

oportunidad de consumir Ia drogo Cadrina? Edad:
12. (A que’ edad probaste la drogo Cadrina por primera vez? Edad:
13. Acerca de las drogos como la heroina. (LA qué edad tuviste tu
. . . , Edad:
primera oportunidad de consumir la heroma?
14. (LA qué edad probaste por primera vez heroina? Edad:

 

15. (LA que’ edad probaste por primera vez pastillas ‘por razones no-
me’dicas" que fueron prescritas por un me’dico a otra persona tales Edad:
como el valium, etc?

16. Acerca de la marihuona. c'A qué edad tuviste tu primera
oportunidad de fumar marihuana?

 

Edad:

 

17. c'A qué edad probaste marihuana por primera vez? Edad:

 

 

 

 

Muchisimas gracias por tu colaboracion. Por favor, devuelve la hoja de respuestas, el

PACARDO, y el lapiz de nuevo a1 sobre y colécalo en la caj a. Te estaremos obsequiando

un boligrafo como seﬁal de nuestro agradecimiento por haber participado en este estudio.
oHay a1 guna otra pregunta antes de terrninar?

185

APPENDIX B

PACARDO-V 2007

THIS SURVEY IS ANONYMOUS AND THEREFORE YOU SHOULD NOT WRITE YOUR NAME
ANYWHERE ON ANY OF THE FORMS.

GENERAL INSTRUCTIONS:

What you are going to do is not difﬁcult and it doesn’t require any study to answer the questions. It is not
an exam and there are no "right" and "wrong" answers. What is important is that the answers you give are
true.

If you have a question or comment, we are happy to help you, just raise your hand.

The information in this survey is secret. The purpose is to understand the opinions and characteristics of
young people like you in order to develop and implement prevention programs.

There are ﬁve sections and a total of l 12 questions on the survey. Try to respond to all of them. Even if
the answers don't coincide exactly with your experience, mark the answer that best describes your
experience.

For example, mark "Yes" if the majority of the time the answer is correct, or "No" if the majority of the
time it is false. If a question makes you feel uncomfortable you can mark the option, "No Response" (NR),
or simply leave it blank.

We will be reading the questions out loud, and we ask that you do not work ahead or go to the next page
until you are instructed to do so. This will help us ﬁnish more quickly and ensure that no one gets confused
or lost.

Do you have any questions before we start?

INSTRUCTION TO BEGIN:

1. Take out the answer sheet (the sheet with a lot of little circles on it).

Use the pencil that is in the envelope to ﬁll out the answer sheet.

Be sure to ﬁll the circles in completely and not to make marks out side of the circle. If you make a

mistake, be sure to erase your previous answer completely. The scoring machine will not read two

answers.

4. Write the number that is on your envelope in the box labeled PID on the answer sheet (in the
lower left hand comer of the answer sheet). Write the number in the boxes and then ﬁll in the
corresponding circle below each number.

5. In the box that says “last name” write the ﬁrst ﬁve letters of the name of your school without

leaving any spaces. For example if your school name is “San Agustin”, write “sanag,” and ﬁll in

the corresponding circle below each letter.
DO NOT put your name on the answer sheet.
Where it says “date”, write your age.

8. Go to number 1 to begin.

wk)

>15"

186

Section 1: General questions

10.

11.

12.

13.

Your position in the school (A) Student

Type of school: (A) Private (B) Public

Your age: (A) 11-12 (B) 13-14 (C) 15-16
(D) 17-18 (E) 19+

What grade are you in: (A) Seventh (B) Eighth (C) Ninth (D) Tenth
(E) Eleventh

Which race do you most identify with? (A) Block (B) Brown (C) White
D) Indian

Where do you live? (A) housing Project (B) Apartment (C) House

How many cars does your family have? (A) 0 (B) 1 (C) 2 (D) 3 (E)

More than 3

How many people live in your house?

i. (A) 1-3 (B) 4—6 (C) 7-8 (D) 9-10 (E) 10+
How many bedrooms are there where you live?
(A) 0 it only has one room (B) 1 (C) 2-3 (D) 4-5 (E) 6+
What is your sex? (A) Male (8) Female

What is your religion?
(A) Catholic (B) Christian non-Catholic (C) Muslim (D) Other
(E) None

How many years of education does your father (or the person who is like your
father) have?
i. (A) Elementary school (B) Didn‘t finish high school (C) Finished
high school (D) Didn't finish college(or technological training)
(E) Finished college(or technological training)

How many years of education does your mother (or the person who is like your
mother) have?
i. (A) Elementary school (B) Didn't finish high school (C) Finished
high school (D) Didn't finish college(or technological training)
(E) Finished college(or technological training)

187

Section 2: Questions about different aspects of Venezuelan youth

Please answer all of the questions even though the answers do not exactly fit your
experience. Mark “Yes ” if the majority of the time the answer is true or “No " If the
majority of the time the answer is false. Mark “NR ” if the question makes you feel

uncomfortable and you don’t want to respond. You should not leave any question without

a response. Indicate your answer by ﬁlling in the corresponding circle on your answer
sheet

 

14.

Your parents or guardians know how you think or feel
regarding the things that are really important to you?

Yes(A)

No (3)

NR (C)

 

15.

Your parents or guardians know what you like and
don't like?

Yes(A)

No (B)

NR (C)

 

16.

Do you always ask permission when you go out to have
a good time?

Yes(A)

No (B)

NR (C)

 

17.

Do you feel that you are important to your
parents/guardians?

Yes(A)

No (B)

NR (C)

 

18.

Generally speaking, your parents or guardians know
where you are and what you are doing?

Yes(A)

No (B)

NR (C)

 

19.

Some times, adults tell their children not to hang
around other young people who get into trouble.
During the last year, have your parents or guardians
told you not to hang around friends that could get you
into trouble?

Yes(A)

No (B)

NR (C)

 

20.

Sometimes young people come home from school and
no adult is there. During the last year have you come
home from school or work and no adult has been
there?

Yes(A)

No (B)

NR (e)

 

21.

Has a member of your family that lives with you such
as mother, father, sibling, etc. smoked cigarettes
during the last year?

Yes(A)

No (8)

NR (C)

 

22.

Has a member of your family that lives with you such
as mother, father, sibling, etc. drank alcohol during
the last year?

Yes(A)

No (B)

NR (C)

 

23.

Has a member of your family that lives with you such
as mother, father, sibling, etc. drank alcohol to the
point of causing problems during the last year?

Yes(A)

No (B)

NR (C)

 

24.

Has a member of your family that lives with you such
as mother, father, sibling, etc. used an illegal drug
like morihuona, crack, etc. during the last year?

Yes(A)

No (8)

NR (C)

 

25.

Do you often have arguments with your parents or
guardians that end in fights?

Yes(A)

No (8)

NR (C)

 

26.

My parents or guardians are always talking to me
about how dangerous drugs are.

Yes(A)

No (8)

NR (C)

 

 

27.

My parents or guardians are always talking to me
about how dangerous alcohol and cigarettes are.

 

Yes(A)

 

No (B)

 

NR (C)

 

188

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

28. Have some of your friends been In trouble wrth the Yes(A) No (8) NR (C)
pohce?
29. Have some of your friends stolen, or damaged Yes(A) No (8) NR (C)
another person s belongings on purpose?
30. Some of my friends smoke cigarettes. Yes(A) No (B) NR (C)
31. Many of my friends smoke cigarettes. Yes(A) No (B) NR (C)
32. Some young people think that it is a good idea to use
drugs and some think that it is a bad idea to use
drugs. Do you have many friends that think using Yes(A) No (3) NR (C)
drugs is a ppq idea?
33. Some of my friends have smoked marihuana Yes(A) No (B) NR (C)
34. Have'you had friends that like to sniff glue or Yes(A) No (8) NR (C)
gasoline, etc?
35. Some young people have begun to use cocaine, crack
or coca base. Do you have a friend that has used Yes(A) No (B) NR (C)
cocaine, crack or coca base?
36. Do you have several friends that use cocaine, crack or Yes (A) No (8) NR (C)
coca base?
37. Have many of your friends skipped school? Yes(A) No (8) NR (C)
38. Have you ever belonged to a gang? Yes(A) No (8) NR (C)
39. Do you get angry frequently? Yes(A) No (B) NR (C)
40. Have you intentionally damage another person's
belongings during the last year? Yes(A) No (B) NR (C)
41. Have you stolen anything during the last year? Yes(A) No (8) NR (C)
42. Have you done anything risky or dangerous during the Yes( A) No (B) NR (C)
last year?
43. Is it true that the usually don't do your homework
from school? Yes(A) No (8) NR (C)
44. I have you had a good relationship wrth the majority Yes( A) No (B) NR (C)
of my teachers.
45. Have you been afraid to go to school during the last Yes (A) No (B) NR (C)
year?
46. I feel that the majority of my teachers are truly
interested in me and my well being. Yes(A) No (B) NR (C)
47. Have you skipped school two or more days in a Single Yes(A) No (8) NR (C)

month during the last school year?

 

 

 

 

189

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

48. Have you been suspended from school? Yes(A) No (B) NR (C)
49. Have your grades gotten worse during this past year? Yes(A) No (B) NR (C)
50. Some young people are happy when they think of going

to school. Generally speaking, have you felt happy

when you thought about going to school during this Yes(A) No (8) NR (C)

past year?
51. I have seriously thought about dropping out of school. Yes(A) No (B) NR (C)
52. Some young people think that going to school is a

waste of time. For you, has going to school been a Yes(A) No (8) NR (C)

waste of time during this past school year?
53. Generally speaking, do people in your neighborhood

help each other? Yes(A) No (B) NR (C)
54. When a young person does something wrong,

sometimes the people who live in his/her

neighborhood tell the child‘s parents about it. In your

neighborhood when a young person does something Yes(A) No (8) NR (C)

wrong, do the neighbors tell the child's

gents/guardians about what they did?
55. Is it common to hear of someone being hurt or

assaulted by delinquents in you neighborhood. Yes(A) No (B) NR (C)
56. Is it common to see people using or selling drugs in

your neighborhood? Yes(A) No (3) NR (C)
57. I frequently see people who are drunk or drugged in

the streets of my neighborhood. Yes(A) No (8) NR (C)
58. I feel safe when I walk alone in my neighborhood. Yes(A) No (8) NR (C)
59. Do the people who live in your neighborhood steel or Yes (A) No (8) NR (C)

damage the belongings of others?

 

 

 

190

 

Section 3: Questions about the risk and accessibility of drugs

Please answer all of the questions even though the answers do not exactly ﬁt your
experience. Mark "A, B, C, or D" according to your experience. Mark ‘E” only if you

don’t have any idea. Try not to leave any answer blank. Indicate your answer by filling in

the corresponding circle on your answer sheet.

 

 

 

 

 

 

 

 

 

 

60. How much risk of harmin themselves
(physically or otherwise)gdoes a (A) (B) (C) (D) (E)
. No Slight Medium Great Don't
person run If they smoke a pack of . . . .
. . risk risk risk risk know
Cigarettes daily?
(A) (D) (E)
61. For you and your friends how easy or Ve (B) (C) Very Don't
difficult is it to get cigarettes? FY Easy Difficult Difficu
easy It know
62. How much r'sk of harm'n themselves
(physically or otherwisel)gdoes a (A) (B) (C) (D) (E)
. . No Slight Medium Great Don't
person run If they have a few drinks . . . .
. . risk risk risk risk know
every once In a while?
63. Ho m ch r'sk of horm'n themsel es
(phzsictblly olr otherwise|)gdoes a V (A) (B) (C) (D) (E)
. . No Slight Medium Great Don't
person run If they drink alcohol . . . .
risk risk risk risk know
frequently?
(A) (0) (E)
64. For you and your friends how easy or Ver (8) (C) Very Don‘ t
difficult is it to get alcohol? y Easy Difficult Difficu
easy It know
65. Ho m ch r'sk of harm'n themsel es
(phtsicLblly dr otherwise')gdoes a V (A) (B) (C) (D) (E)
. No Slight Medium Great Don't
person run if they consume crack or . . . .
. . risk risk risk risk know
coca base every once In a while?
66. Ho m ch r'sk f h rm' th msel es
(phtsiclblly dr ofherjviseggdoe: a V (A) (B) (C) (D) (E)
. No Slight Medium Great Don't
person run If they consume crack or . . . .
risk risk risk risk know
coca base frequently?
67. For you and your friends how easy or (A) (D) (E)
. . . . (B) (C) Very .
difficult 15 It to get crack 0 coca Very . . . . Dont
Easy Difficult Difficu
base? easy I t know
(A) (0) (E)
68. For you and your friends how easy or Ver (8) (C) Very Don’ t
difficult is it to get Ecstasy? y Easy Difficult Difficu
easy It know
69. How much risk of harming themselves (A) (B) (C) (D) (E)
(physically or otherwise) does a No Slight Medium Great Don't
person run if they take Ecstasy every risk risk risk risk know

 

 

 

 

 

 

 

191

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

once in a while?
70. How much risk of harming themselves
(physically or otherwise) does a $30) Slight M531) m Gi‘Deb‘l’ D(0En)'t
person run if they take Ecstasy . .9 . u .
frequently? risk risk risk risk know
71. For you and your friends how easy or (A) (B) (C) V(E) (E)
difficult is it to get inhalants like Very . . . ry Don't
. Easy Difficult Difficu
glue, gasoline, etc? easy It know
72. How much risk of harming themselves
(physically or otherwise) does a S: SIiBh t Midi) rn GI‘D) t Diff
person run if they sniff glue, gasoline, . k .gk . It: .6: kn
etc. everyonce in a while? r15 r15 r15 r15 ow
73. How much risk of harming themselves
(physically or otherwise) does a No) 5??” Meg? to 6:2; t 0(5),?
person run if they sniff inhalants like . .9 . u .
this frequently? risk risk risk risk know
74. How much risk of harming themselves
(physically or otherwise) does a $30) SE82“ Meg) m G921 t D(oEn)'t
person run if they smoke marihuana . .9 . u .
once in a while? risk risk risk risk know
75. How much risk of harming themselves
(physically or otherwise) does a (132 sf?“ Me(cCli)um GI‘Deb t D(0En)'t
person run if they smoke marihuana . .9 . .
frequently? risk risk risk risk know
(A) (D) (E)
76. For you and your friends how easy or (B) (C) Very ,
Very Don t
difficult is it to get marihuana? Easy Difficult Difficu
easy It know

 

Section 4: Questions over the frequency of different aspects in the life of a

Venezuelan youth.

Please answer all of the questions even though the answers do not exactly ﬁt your

 

experience. Mark “A, B, C, D, or E” according to your experience. Try not to leave any
answer blank. Indicate your answer by ﬁlling in the corresponding circle on your answer
sheet

 

 

 

 

 

 

 

 

Every once Several Once
. . Not even in a great Several times a do
The followrng questions refer to once this while times a each orY
the my year. last year during the month
week more
year

 

192

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

77. Watch teleVISion or play Video A B C D E
games at home.
78. Play a sport like soccer,
baseball, mountain clima, etc. A B C D E
79. Play a musical instrument A B C D E
80. Work for money A B C D E
81. Go out on dates A B C D E
82. Smoke cigarettes A B C D E
83. Spend time with my family A B C D E
84. Gamble for money (dice, A B C D E
horses, cards, etc.)
85. Pray or read the Bible A B C D E
86. Do housework like cook, clean, A B C D E
etc.
87. Smoke marihuana A B C D E
88. Go to religious activities like
classes, social activities, A B C D E
retreats, etc.
89. Go to concerts (rock, A B C D E
regueton,etc)
90.D'kl h|.b ,'.
rm 0 co 0 ( eer anis A B C D E
etc.)
91. Do homework and study for A B C D E
tests
92. Do crack, coca base, or A B C D E
cocaine
93. Go dancing A B C D E
94. Go With friends to shopping A B C D E
centers, mowes
95. Take heroin A B C D E

 

 

 

 

 

 

 

 

Section 5: Questions about the age you first had an opportunity to try, or

that you did try a substance

Write the age you were when you ﬁrst had an opportunity to try the drug mentioned in the
question, or the age you were the ﬁrst time you tried the drug mentioned in the question.
Opportunity means that you could have tried the drug had you wanted to, but you may have
decided not to at that time. For example if you were 7 the ﬁrst time you an opportunity try
cigarettes but you didn’t try them, then write “7 ” in the space provided. T he ﬁrst time you tried a
drug refers to the ﬁrst time that you actually consumed the drug. For example if you were 10
years old the ﬁrst time you smoked a cigarette then write “10” in the space provided. It may also
be that the ﬁrst time you had the opportunity to smoke a cigarette you did. Then place the same
age in both spaces. If you never had an opportunity to smoke a cigarette then write “0 ” in the
space provided. Likewise, if you never have smoked a cigarette, even though you have many
opportunities, then write “0 ” in the space provided. For example, perhaps your ﬁrst opportunity
to smoke a cigarette was when you were 9 years old, but you. to this day have never smoked a

193

cigarette, then you would write ”9 " in the space provided for the question that asks about
opportunity and a “0 " in the question that asks about your age when you ﬁrst smoked a cigarette.

 

1. Regarding cigarettes: How old were you when you first had an

 

 

 

opportunity to smoke a cigarette? Age:
2. How old were you when you first smoked a cigarette? Age:
3. Regarding alcohol: How old were you when you first had an A963
opportunity to drink alcohol?
4. How old were you when you first drank alcohol? A93:

 

5. Regarding crack, cocaine, or coca base: How old were you when A981
you first had an opportunity to crack, cocaine, or coca base?

 

 

 

 

 

6. How old were you when you first tried crack, cocaine, or coca Age:
base?

7. Regarding Ecstasy: How old were you when you first had an Age:
opportunity to try Ecstasy?

8. How old were you when you first tried ecstasy? Age:

9. Regarding sniffing glue, gasoline, or other inhalants: How old Age:
were you when you first had an opportunity to sniff inhalants
like these?

10. How old were you when you first sniffed an inhalant like the Age:

ones mentioned?

 

11. Regarding the drug Cadrina: How old were you when you first Age:
had an opportunity to try Cadrina?

 

 

 

 

12. How old were you when you first tried Cadrina? Age:

13. Regarding heroin: How old were you when you first had an Age:
opportunity to try heroin?

14. How old were you when you first tried heroin? Age:

15. How old were you when you first took pills (that were not Age:

prescribed to you by a doctor) like valium, etc.?

16. Regarding marihuana: How old were you when you first had an Age:
opportunity to smoke marihuana?

 

 

A e:
17. How old were you when you first smoked marihuana? g

 

 

 

 

Thank you very much for participating in this survey. Please place the answer sheet, the
PACARDO questionnaire, and the pencil back into the envelope and place it in the box as
we pass by to collect it. We will be giving you a ballpoint pen as a gesture of our
appreciation for having participated in this study today.

Are there any questions before we ﬁnish for the day?

194

APPENDIX C

MAMBI SURVEY

Guide for the Observation of the school, classroom environment

In every school, the principal and teacher whose classroom is participating in the survey will fill out this
evaluation (if there are 2 classrooms participating, both teachers should fill out the forms).

The entire assessment group should also fill out the evaluation forms.

There are questions that pertain to solely one individual in the group. The others can simply leave those
questions blank.

The principal assessor should fill out the top part of the form regarding the identifiers before
distributing.

 

This evaluation should be completely quickly and simply to give a ﬁrst impression.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l Name of Researcher l l Today's T y I ~ 1 Scheduled time I l
' date: I l '
School Number . ‘ ‘ l '
. ; . i l - l l
l Job title of the person completing this survey: I
l
TO BE COMPLETED BY THE PRINCIPAL ONLY:
This school is (mark all that apply)?:
1. Dthc 4. DCompletely Private
2. [Religious (which religion) 5. [:lSemi-private
3. DBihngual
FOR THE PRINCIPAL, TEACHER, AND INTERVIEWER:
Description of the classroom LYes 2.No 3.Can’t tell
(Mark an X in the correct box)
. 1. Are there desks and chairs for every student?
, 2. Do the students appear comfortable in their desks?
3. Is there a place where students can store books and food?
. 4. Is the classroom decorated with the student's work, drawings, maps and
educational posters?
; 5. Is there a blackboard in every classroom?
; 6. Is there chalk or a writing instrument to write on the board?
7. Is the classroom well ventilated?
8. Does the classroom appear clean?
l 9. Do students have there own textbooks?
10. Are there extra curricular activities available (e.g., sports, clubs, etc)? 1 J
.. ___ l

 

 

195

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Does the school have the following? LYas 2.No 3.Can’t tell
(Mark an X in the correct box)
. 11. Does it have a playing field?
12. Does it have a gymnasium?
'L 13. Does it have a library?
L14. Does it have a cafeteria or dining room?
i 15. Does it have a patio or courtyard?
16. Is there graffiti on the walls?
‘ 17. Is there barbed wire on top of the fence or wall surrounding the school?
L 18. Is there broken glass on top of the fence surrounding the school?
i 19. Are there computers with Internet access in the principal's office?
20. Are there computers with Internet access for the students to use?
. 22. Is there a security system in the school?
L B. Is there a guard on school grounds?
L 24. Are access doors kept closed?
‘ 25. Are there bulletin boards on the walls with up—to-date armouncements?
26. Is the roof in good condition (signs of water damage)?
l 27. Are there places that showcase student achievements?
28. Are there broken windows?
l 29. Are there broken walls or those having holes? a
30. Is the paint peeling on the walls, doors, or window frames?
1 31. Are there doors that are off the hinges or that are broken?
Discribe the area around the school 1.Yes 2N0 3.Can’t tell
(Mark an X in the correct box)
32. Is there rubbish on the streets or around the buildings and houses in the
neighborhood?
3. Are there sidewalks in the streets around the school?
34. Do the houses surrounding the school appear well kept?
35. Is there a recreational park nearby?
36. Are there abandoned cars or cars being repaired in the streets?
37. Are there factories and warehouses around the school?
38. Are there stores and businesses around the school?
39. Are there billboards or public advertisements for tobacco around the
school?
40. Are there billboards or public advertisements for alcohol around the
school?

 

 

 

 

 

 

196

 

 

 

GUIA DE OBSERVACION DEL MEDIO AMBIENTE DEL SALON, COLEGIO Y
VECINADARIO (MAMBI)

En cada colegio, eI director y maestro de del salon que participa en la encuesta va llenar esta
evaluacién (si hay dos salones de clases en la encuesta entonces ambos maestros pueden llenar la
evaluacion).

Todo el personal que esta en el equipo de asesoramiento tambie’n lo llena al terminar el
asesoramiento.

Hay preguntas, al principio que Ie toca a solamente un individuo del grupo, (por ejemplo las
preguntas administrativas solamente el director quien los contesta). Los demds lo pueden dejar
en blanco.

El asesor principal llena la parte sobre identificacion de cada cuestionario antes de darles al
personal que lo va llenar.

 

Esta evaluacidn 5e completa en poco nempo El propOSito dar su primera impresiOn del medio ambiente del colegio.

 

L Nombre del Investigador L‘ Fecha I L I L Trempo: )

l l ll)

 

 

Numero delColegio 3 T T L— )7

l
i [All

 

 

 

 

Titulo de la persona completando la encuesta:

l

 

 

Para Ser Completado Por El Director.

Esta Escuela Es (marca todos que apliquen)?:

 

A. [:JPublica D. [:JCompletamente Privada
B. [:JReligiosa(cual religién) E. DSemi—privada
C. [:lBilingue

PARA EL DIRECTOR FOR THE PRINCIPAL, TEACHER, AND INTERVIEWER:

 

Descripcién del salén de clase SI (A) No (B) No Sé (C)

 

 

1. éHay sillas y pupitres para cada estudiante?

 

2. {L05 estudiantes se ven comodos en sus pupitres?

 

 

 

 

 

 

197

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l 3. éHay lugares donde los estudiantes pueden guardar sus libros y comida?
4. (El salon esta decorado con trabajos de los estudiantes, dibujos, mapas o
carteles de informacion educational?
( 5. gl-Iay pizarra en cada clase?
6. éHay tiza 0 con qué escribir en las pizarras?
L 7. gLa clase esta ventilada?
L 8. gLa clase se ve limpia?
( 9. (L05 estudiantes tienen sus propios libros de texto?
[ l Tiene el colegio lo siguiente? 51 (A) No (B) No Sé (C)
L 10. , gI-lay actividades después de las clases para los estudiantes (deportes, club,
etc.)?
L 11. [Ilene campo o parque para recreacién?
L '12. iTiene gimnasio?
13. [Ilene biblioteca?
14. [Ilene cafeteria o comedor?
L 15. [Ilene patio?
L '16. Hay graffiti en las paredes y muros?
'17. C'I-Iay alambre de puas encima de los muros que rodean el colegio?
i 18. C'Hay vidrios rotos encima de los muros que rodean e1 colegio?
19. éI-Iay computadoras con conexion al Internet en la oficina del director?
20. (Hay com putadoras con conexién al Internet para uso de los
estudiantes?
L 22. gHay sistema de seguridad en el colegio?
23. éI-Iay guarda en el colegio?
‘ 24. gMantienen los portones de acceso cerrados?
LT25. éI-lay un lugar para anuncjos?
26. £151 techo o cieIo raso esta en mal estado? (seftalcs de filtraciones)?
‘ 27. gHay un lugar que demuestra los logros de los estudiantes?
28. (Hay ventanas rotas?
29. gHay paredes rotas 0 con huecos?
l 30. éFaltan rejas o estan deterioradas (faltan pintar)?
L 31. éHay puertas desmontadas o danadas? J
Describe el area alrededor e1 colegio SI (A) No (B) No Sé (C)
32. C'Hay basura en las calles o alrededor de los edificios/ casas en el vecindario?
36. C'Hay aceras en las calles que rodean el colegio?
34. C'Las casas que rodean el colegio se ven bien cuidadas?

 

 

198

 

 

 

 

 

 

 

 

 

 

 

 

35. (Hay algun parque recreativo?

36. (Hay carros abandonados o srendo reparados en las calles?

37. (Hay fabricas o almacenes en los alrededores del colegio?

38. (Hay tiendas o negocios cerca?

39. ("Hay vallas pubhcxtarias o anuncios para tabaco cerca del colegio?
40. (Hay vallas publicitarias o animcios para alcohol cerca del colegio?

 

199

 

 

 

 

T

OFFICE OF
REGULATORY
AFFAIRS

Human Research
Protection Programs

BIOMEDICAL 8: HEALTH
INSTITUTIONAL REVIEW
BOARD (BIRB)

COMMUNITY RESEARCH
INSTITUTIONAL REVIEW
BOARD (CRIRB)

SOCIAL SCIENCEI
BEHAVIORAL i EDUCATION
INSTITUTIONAL REVIEW
BOARD (SIRB)

202 Olds Hall

East Lansing, Michigan
48824-1046
517-355-2180

Fax: 517-432-4503

wwwhumanresearchmsuedu
SIRB & BIRB: IRB@msu.edu
CRIRB: crirb@msu.edu

{ii-55%
{'2’ I.Ful’l'ngfLi‘h-
t «RAJ RIP." ,
hv-ff‘f‘ .5!
\‘t...

:V‘
“we-V

MSU is an aﬂirman’ve—acrion
equal-opportunity institution.

   
 
 

APPENDIX D

MICHIGAN STATE

 

UNIVERSITY

 

Initial IRB
Application
Approval

April 27, 2007

To: Adrian BLOW
36 Human Ecology
FCE
East Lansing. MI 48824

Re: IRB # 07-320 Category: EXPEDITED 2-7
Approval Date: April 27, 2007
Expiration Date: Apr“ 26, 2008

Title: TOWARD AN ECO-DEVELOPMENTAL THEORY OF ADOLESCENT SUBSTANCE ABUSE

The Institutional Review Board has completed their review of your project. I am pleased to advise you that
your project has been approved.

The committee has found that your research project is appropriate in deslgn. protects the rights and welfare of
human subjects, and meets the requirements of MSU's Federal Wide Assurance and the Federal Guidelines
(45 CFR 46 and 21 CFR Part 50). The protection of human subjects In research Is a partnership between the
IRB and the investigators. We look forward to working with you as we both fulﬁll our responsibilities.

Renewals: IRB approval is valid until the expiration date listed above. If you are continuing your project. you
must submit an Application for Renewal application at least one month before expiration. If the project is
completed, please submit an Application for Permanent Closure.

Revisions: The IRB must review any changes in the project, prior to initiation of the Change. Please submit an
Application for Revision to have your changes reviewed. If changes are made at the time of renewal. please
include an Appllcatlon for Revision with the renewal application.

Problems: If issues should arise during the conduct of the research, such as unanticipated problems. adverse

events, or any problem that may increase the risk to the human subjects. notify the IRB ofﬁce promptly. Forms
are available to report these issues.

Please use the IRB number listed above on any forms submitted which relate to this project. or on any
correspondence with the IRB ofﬁce.

Good luck In your research. if we can be of further assistance. please contact us at 517-355-2180 or via email
at IRB@m§g.gdu. Thank you for your cooperation.

Sincerely,

1441—3—7—

Peter Vasilenko, PhD.
SIRB Chair

C: Ronald Cox
2500 Teel Avenue
Lansing MI, 48910

200

Page: 1 of2

Michigan State University
Department of Family and Child Ecology

Project Title: Toward an Eco-Developmental Theory of
Adolescent Substance Abuse.

Participant Informed Consent Form

PURPOSE: This research aims to learn more about how parents, teachers, and communities
work together to protect their children from getting involved with drugs. The results from this
research will be used to deve10p new prevention and treatment programs in Caracas for
adolescents who have behavior problems or who get involved with drugs. These programs has
received scientiﬁc support in other countries, but have never been developed for use in
Venezuela. Your participation will shed light on how to best adapt these programs so that they
can be implemented in Venezuela.

BENEFITS: Your participation in this study has some potential beneﬁts to you. You may
beneﬁt by being valued as a credible and valuable resource in a project that has the potential to
help teens and families as well as the school environment. Hopefully this will be a source of
satisfaction and self-esteem for those who participate. You may beneﬁt directly in the ﬁiture by
seeing a decrease in drug activity in your school, work, or neighborhood, and an increase in the
quality of your relationships. Additionally, by lending your voice to this study, government
agencies and other organizations, which support families and schools, may become more aware
of and responsive to the needs of parents, teens, and schools.

RISKS: In any research study there are risks involved with participation. Participation in this
study may lead you to think about issues related to your own family, or other relationships that
make you feel uncomfortable, or bring back unpleasant memories. Likewise, some questions
may make you think about past behavior that you are not particularly proud of. Ifyou do
experience some adverse effects, we encourage you to speak to a psychologist or counselor that
you may know or to contact the psychologist whose information is provided below.

PARTICIPATION: Participation in this study consists of responding to a questionnaire that
has been used with other schools in Panama, Costa Rica, Nicaragua, El Salvador, Honduras,
Guatemala, and The Dominican Republic. The questionnaire asks you to respond to items
concerning different aspects of the school environment and your students. Because of the above-
mentioned risks, we want to emphasize that your participation in this study is voluntary. You
have the right to not participate in this study at all, to refuse to answer any questions, or end your
participation at any time, without penalty. It should take about 45 minutes of your time to
complete the questionnaire. Your participation in this research project will not involve any

Subject Initials

Date
This consent form was approved by the Social SdencelBehavioral/Eduwﬁon Institutional Review Board (SIRB) at Mchigan State
University. Approved 04/27/07 -— valid through 0426/08. This wrsion supersedes all previous versions. IRB ll 07-320.

 

201

Page: 2 of 2

additional costs to you beyond your time, and a ballpoint pen will be given to you in appreciation
of your participation when you turn in the questionnaire. Additionally, we are giving the school
a new laser printer in appreciation for participating in this study.

PRIVACY AND CONFIDENTIALITY: Information collected will be kept strictly
conﬁdential. This means that neither your name, the name of the school where you work, nor
any other information that could be used to identify you will appear on any of the documents
prepared as result of this study. A copy of this form is provided for your convenience in the
event you need it as a reference for later questions or concerns.

QUESTIONS AND CONTACTS: If at any time you have questions or concerns about this
study, you may contact Mr. Ronald Cox at 011-517-282-7152-3328, by email:
coxrona1@msu.edu. by mail: 107 Human Ecology, Michigan State University East Lansing, MI
48824, USA,. You may also contact Dr. Adrian Blow at 011-517-432-7092, by e-mail at
blow@msu.edu. or by mail at 3B Human Ecology, Michigan State University, East Lansing, MI
48824. Ifyou have questions or concerns regarding your rights as a study participant, or are
dissatisﬁed at any time with any aspect of this study, please feel free to contact Dr. Peter
Vasilenko by phone: 011-517-355-2180, e-mail: ucn'hs@msuedu or mail: 202 Olds Hall East
Lansing, MI 48 824, USA. If you would like to speak to someone about any discomfort you may
feel as a result of participating in this study, Lic. Mariela Rodriguez is a competent inexpensive
psychotherapist who practices in Caracas. Her number is 0414-257-9777.

If you agree to participate please sign and date this form on the line below. Your signature below
indicates your voluntary agreement to participate in this study. '

Name of Participant:

 

Signature: Date:

 

Subject Initials

Date
This consent form was approved by the Social Science/Behavioral/Edumtion Institutional Review Board (SIRB) at Michigan State
University. Approved 04/27/07 - valid through 04/26/08. This version supersedes all previous versions. IRB ll 07-320.

Page: 1 of2

Michigan State University
Facultad de Ecologia Familiar y del Nir‘io

Proyecto: Hacia un teoria ecolOgica del desarrollo del abuso de
drogas en los adolescentes venezolanos

Hoja de Consentimiento Informado

PROPOSITO: Este estudio tiene como objetivo aprender mas acerca de como los padres,
maestros y comunidades cooperan y colaboran para evitar que los niilos se involucren en las
drogas. Los resultados de este estudio seran empleados en el desarrollo de nuevos programas
para la prevencién y el tratamiento de adolescentes que tienen problemas de conducta y do
consumo de drogas en Venezuela. Estos programas dc tratamiento se han mostrado eﬁcaces en
otros paises, pero nunca han sido desarrollados para uso en Venezuela. Tu participacién
alurnbrara maneras en que podamos adaptar estos programas para que scan implementados en
Venezuela.

BENEFICIOS: Tu participacién en este estudio tiene e1 potencial de beneﬁciarte. Tal vez te
beneficies por sentirte valorizado(a) como una ﬁrente respetada dc informacién valiosa para un
proyecto que tiene potencial para ayudar a los adolescentes y sus familias, tanto como la misma
escuela. Esperamos que esto sea causa de mucha satisfaccién y autoestima para quienes
participan. Tal vez también seas beneﬁciario en el futuro de un descenso en actividades
relacionadas con la droga en tu trabajo o vecindario, y un aumento en la calidad de tus relaciones
personales. Ademas, al prestar tu apoyo al desarrollo de este estudio, entidades gubernamentales
y ou'os organisrnos que se interesan en ayudar a la familia veneeolana y a las escuelas puedan
llegar a estar mas atentos a las necesidades de los padres, adolescentes, y escuelas.

RIESGOS: En toda investigacién cientiﬁca existen riesgos para Ios que participan en dichas
investigaciones. Al participar en este estudio pueda que te lleve a pensar en temas relacionados
con tu familia que te incomoden o te hagan recordar dc momentos no placenteros en tu vida. Si
experimentas algunos efectos adversos por haber participado en este estudio, te animamos a
hablar con un psico’logo o consej ero que conozcas o contactar el psicélogo cuya informacién esta
al final de este documento.

PARTICIPACION: Participar en este estudio consiste en responder a un cuestionario que ha
sido implementado en escuelas en Panama, Costa Rica, Nicaragua, El Salvador, Honduras,
Guatemala, y Republica Dominicana. El cuestionario te pregunta sobre diferente aspectos del
arnbiente escolar y los estudiantes. Debido a los riesgos previamente mencionados, enfatizamos

Subject Initials

Date
Esta forma de consentimiento fue aprobada por el Consejo Institucional de Revision (SIRB) de Ciencias
Sociales/ConductuaVEducativo de la Universidad del Estado de Michigan. Aprobada a partir de 04/27/07
hasta 04/26/08. Esta version reemplaza a todas las versiones anteriores. lRB#07-320.

203

Page: 2 of2

que tu participacién en este estudio es voluntaria Tienes e1 derecho de decidir a no participar en
este estudio, de rehusar a contestar cualquier pregunta, o terminar tu participacién en cualquier
momento, sin ninguna penalidad. Se tomara aproximadamente 45 minutos para llenar cl
cuestionario, y tu participacién en este proyecto no te ocasionara ningi'm costo mas alla que el
tiempo que inviertes. En agradecimiento por su participacién en el estudio te estaremos
obsequiando un boligrafo y al colegio una impresora laser.

PRIVACIDAD Y CONFIDENCIALIDAD: La informacién recogida en este estudio se
mantendra en forma conﬁdencial. Esto signiﬁca que ni tu nombre, ni el nombre de la escuela
donde trabajas, ni cualquier otro dato que se podrla usar para identiﬁcarte, aparecera en los
documentos que se desarrollaran relacionados con este estudio. Se te esta dando una copia de
esta hoja como referencia en el caso que tienes alguna pregunta en el futuro.

PREGUNTAS Y CONTACTO: Si en cualquier momento tienes algunas preguntas o
preocupaciones acerca de este estudio, puedes contactar al Sr. Ronald Cox a: 001-517-282-7152
0 per correo electrénico a: coxronal@msu.edu. 0 per correo normal a: 107 Human Ecology,
Michigan State University East Lansing, MI 48824, USA También puedes contactar a Sr. Cox
en Venezuela a1 0212—915-0455. También puedes contactar a1 Dr. Adrian Blow at 001-517-432-
7092, por correo e1ectrénico a blow/@msuedg, o por correo normal a BB Human Ecology,
Michigan State University, East Lansing, MI 48824 USA. En el caso que tengas preguntas o
preocupaciones en cuanto a tus derechos en este estudio, o te sientes disgustado con cualquier
aspecto de este estudio, por favor comunicate con el Dr. Peter Vasilenko por teléfono a: 001-
517-355—2180, por correo electrOnico a: irb@msu.edu o por correo normal a: 202 Olds Hall East
Lansing, MI 48824, USA. Si gustas hablar con algi’m profesional acerca de cualquier
incomodidad que sientas por causa de tu participacién en este estudio, 1a psicOloga Lic. Mariela
Rodriguez esta al tanto de este estudio y esta dispuesta a ayudarte aqui en Caracas. Su m’rmero
es: 0414-257-9777

Si estas de acuerdo en participar en el presente estudio, por favor coloca tu nombre y ﬁrrna e1
documento. Tu ﬁrma indica tu participacién voluntaria en este estudio.

Nombre del participante:

 

Firma: Fecha:

 

Subject Initials

Date
Esta forrna de consentimiento fue aprobada por el Consejo Institucional de Revision (SIRB) de Ciencias
SociaIes/Conductual/Educativo de la Universidad del Estado de Michigan. Aprobada a partir de 04/27/07
hasta 04/26/08. Esta version reemplaza a todas las versiones anteriores. lRB#07-320.

204

Michigan State University
Facultad de Ecologia Familiar y del Niilo

Proyecto: Hacia una teoria ecolégica del desarrollo del abuso de
drogas en Ios adolescentes venezolanos

Hoja de Consentimiento Informado de los Representantes para un Participante Menor

Estimado Representante:

Un grupo de investigadores de la Universidad Estatal de Michigan quiere aprender mas acerca
de co'mo reducir e1 consumo de bebidas alcohélicas y otras drogas y planiﬁcar scrvicios para ayudar a
los jévenes que los requieran. Han sido desarrollados algunos programas de prevencién que funcionan
con jévenes, sus farnilias, y sus comunidades y estan mostrandosc eﬁcaces en varias partes del mundo,
y ahora hay interés en adaptarlos para su uso en Venezuela. Para tal ﬁn, se hace necesario realizar una
encuesta con jévenes escolares en Caracas mediante un cuestionario anOnimo que prueba diferentes
aspectos de la salud y el comportarniento de los jévenes, incluyendo el consumo dc alcohol y drogas.
El cuestionario ha sido utilizado con otros jévenes en Panama, Costa Rica, Nicaragua, El Salvador,
Honduras, Guatemala y la Republica Dominicana E1 cuestionario 1e hara preguntas a su hijo(a) acerca
de su relacién con su familia amistades, maestros, rcligién, vecindario, y las drogas y el alcohol.

El director del plantel donde asiste(n) su(s) hijo(s) ha revisado e1 cuestionario y los
procedimientos que seran implementados y ha dado su aprobacién. La encuesta ha sido diseilada para
ser anOnima, es decir, que nadie sabra que estudiante responde a las preguntas. Se instruira a los
estudiantes a no colocar sus nombres ni ninguna otra cosa que les podria identiﬁcar en ninguna parte ‘
del cuestionario. También se tendra cuidado para que nadie pueda observar las respuestas dc otra
persona. Se tomara aproximadarnente 45 minutos para llenar e1 cuestionario y la participacién de su
hijo(a) no le ocasionara ningt'rn otro gasto mas alla de su tiempo. En agradecimiento por su
participacién en el estudio 1e estaremos obsequiando a su hijo(a) un boligrafo.

La participacién en este estudio tiene e1 potencial de beneﬁciar a su hijo(a). Tal vez so
beneﬁcie por sentirte apreciado(a) como una fuente respetada de informacién valiosa para un proyecto
que tiene potencial para ayudar a los adolescentes y sus farnilias, tanto como la misma escuela
Esperamos que esto sea causa dc mucha satisfaccio’n y autoestima para quienes participan. Tal vez
tarnbién sea beneﬁciado(a) en el futuro por Im descenso en actividades relacionados con la droga en su
colegio, trabajo o vecindario, y por un aumento en la calidad de sus relaciones personales. Ademas, al
prestar su apoyo a1 desarrollo de este estudio, entidades gubemamentales y otros organismos que se
interesan en ayudar a la familia venezolana y a las escuelas puedan llegar a estar mas atentos a las
necesidades de los padres, adolescentes y escuelas.

En toda investigacién cientiﬁca existen riesgos para los que participan en dichas
investigaciones. A1 participar en este estudio pueda que a su hijo(a) sc lleve a pensar en temas
relacionados con su familia que le incomoden o Ie hagan recordar momentos no placenteros en su vida.
Si es el caso que su hijo(a) experimenta algunos efectos adversos por haber participado en este estudio,
1e animaremos a hablar con su maestra(o) 0 con director(a) del plantel para que orienten a su hijo(a).
Debido a los mencionados riesgos, queremos enfatizar que la participacién de su hijo(a) en este estudio
es de naturaleza voluntaria. E1(Ella) tiene cl derecho do no participar, dc rehusar responder a cualquier

Esta forma de consentimiento fue aprobada por el Consejo Institucional de Revision (SIRB) de Ciencias
SociaIes/Conductual/Educativo de la Universidad del Estado de Michigan. Aprobada a partir de 04/27/07 hasta
04/26/08. Esta version reemplaza a todas las versiones anteriores. lRB#07-320.

to
0
UI

pregunta, o descontinuar su participaciOn en cualquier momento sin ninguna repercusién. Si su hijo(a)
decide no participar, hay un aula de estudio supervisado donde puede hacer tareas mientras
terminamos con el proyecto.

Si en cualquier momento tiene algunas preguntas o inquietudes acerca de este estudio, puede
contactar a1 Sr. Ronald Cox a: 001-517-282-7152 0 per correo electrénico a: coxronal@msu.edg, o por
correo normal a: 107 Human Ecology, Michigan State University East Lansing, MI 48824, USA.
También puede contactar al Sr. Cox en Venezuela al 0212-915-0455. También puedes contactar a1 Dr.
Adrian Blow at 001-517-432-7092, por correo electrOnico a blow@su.edu, o por correo normal a: 313
Human Ecology, Michigan State University, East Lansing, MI 48824 USA. En caso que tenga
preguntas o inquietudes en cuanto a los derechos de su hijo(a) en este estudio, 0 se siente disgustado
con cualquier aspecto de este estudio, por favor comuniquese con el Dr. Peter Vasilenko por teléfono
al: 001-517-355-2180, por correo electrOnico a: irb@msu.edu 0 per correo normal a: 202 Olds Hall
East Lansing, MI 48824, USA.

Por favor devuelva este documento a1 maestro(a) de su hijo(a) indicando si esta de acuerdo o no
esta de acuerdo con la participacién de su hijo(a) en este proyecto. Si no esta de acuerdo con la
participacién de su hijo(a), 61 o ella sera dirigido(a) a un lugar para hacer tareas mientras Ios demés
estudiantes estan llenando el cuestionario. Si no envia e1 documento permitiremos a su hijo(a)
participar y el director(a) actuara como intercesor a favor del estudiante para asegurar que sus derechos
son protegidos. Sin embargo, si no responde a esta notiﬁcacién, su hijo(a) mantiene el derecho de no
participar, de rehusar responder a cualquier pregunta o a terminar su participacién en cualquier
momento sin ninguna repercusién. Si hijo(a) decide no participar, sera dirigido a un aula para estudiar
mientras los demas estudiantes terminan e1 cuestionario.

Favor, indique su disposiciOn en cuanto a la participacién de su hijo(a) en el cuestionario a
través de su ﬁrma al lado de la declaracién que expresa su deseo. Por favor envie esta carta a la
maestra(o) de su hijo(a). Gracias.

 

Doy mi permiso para que mi hijo(a) participe en el estudio:

 

 

 

F irma

 

 

N0 doy mi permiso para que mi hijo(a) participe en el estudio:

 

Firma

 

 

 

Esta forma de consentimiento fue aprobada por el Consejo lnstitucional de Revision (SIRB) de Ciencias
Sociales/Conductual/Educativo de la Universidad del Estado de Michigan. Aprobada a partir de 04/27/07 hasta
04/26/08. Esta version reemplaza a todas las versiones anteriores. IRB#07-320.

206

Page: 1 of 2
Michigan State University
Department of Family and Child Ecology

Project Title: Toward an Eco-DeveIOpmental Theory of Adolescent Substance
Abuse.

Parental Informed Consent Form for a Minor Participant

Dear Parents/Guardians:

A group of researchers from Michigan State University is attempting to learn more about how
to prevent the spread of alcohol and drug use among the youth of Venezuela. Several prevention
programs that work with children, their families and their communities have been developed, and are
showing positive results in different parts of the world, and there is now interest in adapting these
programs for use in Venezuela. In order to help with this project, we are looking to survey students
from schools in Caracas using a questionnaire that touches on different aspects of health and behavior,
including the consumption of alcohol and drugs. The questionnaire has been used with other
adolescents in Panama, Costa Rica, Nicaragua, El Salvador, Honduras, Guatemala, and the Dominican
Republic. The questionnaire asks your child to respond to items concerning different aspects of his/her
relationship with family, peers, teachers, religion, neighborhood, and drugs and alcohol.

The principal of your child/children’s school has reviewed the questionnaire that will be used
and has approved of the procedures that the researchers have suggested. The survey has been designed
to be anonymous, which means that no one will know how a student responds to a question. The
students will be instructed not to place their name or anything else that could identify them on any part
of the questionnaire. Also, great care is being taken so that no one will be able to observe any other
person’s responses. The survey will take about 45 minutes to complete, and your child’s participation
in this research project will not involve any costs beyond his or her time. We will be giving your child
a ballpoint pen as a token of our appreciation for having participated in the study.

Your child/children’s participation in this study has some potential beneﬁts to him or her. Your
child may beneﬁt by being valued as a credible and valuable resource in a project that has the potenu'al
to help teens and families as well as the school environment. Hopeﬁrlly this will be a source of
satisfaction and self-esteem for those who participate. Your child may beneﬁt directly in the future by
seeing a decrease in drug activity in his/her school, work, or neighborhood, and an increase in the
quality of his/her relationships. Additionally, as a result of your child/children’s help in this study,
government agencies and other organizations, which support families and schools, may become more
aware of and responsive to the needs of parents, teens, and schools.

In any research study there are risks involved with participation. Participation in this study
may lead your child to think about issues related to his/her own family, or other relationships that make
him/her feel uncomfortable, or bring back unpleasant memories. Likewise, some questions may make
him/her think about past behavior that they are not particularly proud of. If they do experience some
adverse effects, we will encourage them to speak to their classroom teacher or to the school principal
in order to address your child’s concern. Because of the above-mentioned risks, we want to emphasize
that your child’s participation in this study is voluntary. He/she has the right to not participate in this
study at all, to refuse to answer any questions, or end his/her participation at any time, without penalty.
If he/she decides not to participate, there is a study hall set up for them.

This consent form was approved by the Social Science/Behavioral/Edumtion Institutional Review Board (SIRB) at Michigan State University.
Approved 04/27107 - valid through 04/26/08. This version supersedes all previous versions. IRB ii 07—0320.

207

Page: 2 of 2

If at any time you have questions or concerns about this study, you may contact Mr. Ronald
Cox at 001-517-432-3328 (US number), by email: coxron_al@msu.edu Lby mail: 107 Human Ecology,
Michigan State University East Lansing, MI 48824, USA, or at 0212-915-0455 in Venezuela. You
may also contact Dr. Adrian Blow at 001-517-432-7092 (US number), by e-mail at blowa@ﬂu.edg,
or by mail at 3B Human Ecology, Michigan State University, East Lansing, MI 48824. If you have
questions or concerns regarding your rights as a study participant, or are dissatisﬁed at any time with
any aspect of this study, please feel free to contact Dr. Peter Vasilenko by phone: 001-517-355-2180
(US number), e-mail: ucrihs@msu.edu or mail: 202 Olds Hall East Lansing, MI 48824, USA.

Please return this letter to your child’s school teacher indicating whether you agree or do not
agree with your child’s participation in this project. If you do not agree with your child’s participation
your child will be allowed to enter a study hall while the other students are responding to the
questionnaire. If you do not respond to this letter then we will allow your child to participate and the
school principal will act as an advocate to ensure that the rights of your child are protected. Even if
you do not respond to this letter, your child will have the right to refuse to participate, to refuse to
answer any questions, or to end participation at any time, without any penalty. Ifhe/she decides not to
participate, there will be a study hall set up for him/her to study while others are completing the
survey.

Please indicate your agreement or disagreement for your child’s participation in taking the
questionnaire by signing next to the statement that expresses your desire. Please return this letter to
your child’s school teacher. Thank you.

 

I give my permission for my child to participate in the study:

 

Signature

 

 

 

 

I DO NOT give my permission for my child to participate in the study:

 

 

Signature

 

 

This consent form was approved by the Social Science/Behaviorai/Education Institutional Review Board (SIRB) at Mehigan State University.
Approved 04/27107 - valid throagh 04126/08. This version supersedes all previous versions. iRB it 07-0320.

208

Page: 1 of2

Michigan State University
Department of Family and Child Ecology

Participant Informed Assent Form
Toward an Eco-Developmental Theory of Adolescent Substance Abuse.

PURPOSE: This research aims to learn more about how parents, teachers, and communities
work together to protect their children from getting involved with drugs. The results from this
research will be used to develop new prevention and treatment programs in Caracas for
adolescents who have behavior problems or who get involved with drugs. These programs has
received scientiﬁc support in other countries, but have never been deveIOped for use in
Venezuela. Your participation will shed light on how to best adapt these programs so that they
can be implemented in Venezuela.

BENEFITS: Your participation in this study has some potential beneﬁts to you. You may
beneﬁt by being valued as a credible and valuable resource in a project that has the potential to
help teens and families as well as the school environment. Hopefully this will be a source of
satisfaction and self-esteem for those who participate. You may beneﬁt directly in the future by
seeing a decrease in drug activity in your school, work, or neighborhood, and an increase in the
quality of your relationships. Additionally, by lending your voice to this study, government
agencies and other organizations, which support families and schools, may become more aware
of and responsive to the needs of parents, teens, and schools.

RISKS: In any research study there are risks involved with participation. Participation in this
study may lead you to think about issues related to your own family, or other relationships that
make you feel uncomfortable, or bring back unpleasant memories. Likewise, some questions
may make you think about past behavior that you are not particularly proud of. If you do
experience some adverse effects, we encourage you to speak to your classroom teacher or to the
school principal. They can help you, or they can set up a conﬁdential meeting with the school
psychologist for you if you feel that you need someone else to talk to. Or, you can contact the
school psychologist directly yourself.

PARTICIPATION: Participation in this study consists of responding to a questionnaire that
has been used with other adolescents in Panama, Costa Rica, Nicaragua, El Salvador, Honduras,
Guatemala, and The Dominican Republic. The questionnaire asks you to respond to items
concerning different aspects of your relationship with family, peers, teachers, religion, your
neighborhood, and drugs and alcohol. Because of the above-mentioned risks, we want to
emphasize that your participation in this study is voluntary. You have the right to not participate
in this study at all, to refuse to answer any questions, or end your participation at any time,
without penalty. If you do decide not to participate, there is a study hall set up for you.

PRIVACY AND CONFIDENTIALITY: Information collected will be anonymous. This

means that neither your name nor any other information that could identify you will appear on
any of the documents used in this study. You will select an envelope at random that has a

This consent form

the ' 6
University. Approved 04/27/07- El id throucm 0426/08. This version Supersedes all previggs versions. lRB it 97- 3:9,

 

209

Page: 2 of2

Scantron sheet along with a questionnaire. The envelope, and the Scantron sheet will have a
number on them that lets me know which school this is. However, since I don’t know which
number you will select, and nor do I know who you are, there is no way for anyone to link you
with your responses to the items on the questionnaire. I would like for you to keep the form we
are reading from for a reference in the event you have any questions later.

COSTS AND COMPENSATION: It should take about 45 minutes of your time to
complete the questionnaire. Your participation in this research project will not involve any
additional costs to you beyond your time. A $5.00 gift card will be given to you in appreciation
of your participation when you turn in the envelope with the Scantron.

QUESTIONS AND CONTACTS: If at any time you have questions or concerns about this
study, you may contact Mr. Ronald Cox at 517-432-3328, by email: coxronal@msu.edu , by
mail: 107 Human Ecology, Michigan State University East Lansing, MI 48824, USA, or in 0212-
915-0455. You may also contact Dr. Adrian Blow at 517-432-7092, by e-mail at
blow@msu.edu, or by mail at 3B Human Ecology, Michigan State University, East Lansing, MI
48824. If you have questions or concerns regarding your rights as a study participant, or are
dissatisﬁed at any time with any aspect of this study, please feel free to contact Dr. Peter
Vasilenko by phone: 517-355-2180, e-mail: ucrihs@msu.edu or mail: 202 Olds Hall East
Lansing, MI 48824, USA.

If you agree to participate please stay seated and you will be given a questionnaire. Your
continued presence in the classroom indicates your voluntary agreement to participate in this
study. If you do not agree to be involved in the study then you may go with your teacher at this
time to a study hall.

This consent form was a roved b the Social Science/Benavioral/Education in trtutional Review Board SI 8 at Mi i an Stat
University. Approvg 04/27/07 - vpiig through 04126/38. This version supersedes ail previous vgrs, ims. lRB # 07-320.

 

 

210

Page: 1 of2

Michigan State University
Facultad de Ecologia Familiar y del Niﬁo

Hoja de Asentimiento Informado
Hacia una teoria ecoldgica del desarrollo del abuso de drogas en los adolescentes
venezolanos

PROPOSITO: Este estudio tiene como objetivo aprender mas acerca de como los padres,
maestros y comunidades cooperan y colaboran para evitar que los nifios se involucren en el
consumo de drogas. Los resultados de este estudio seran empleados en el desarrollo de nuevos
programas para la prevencion y el tratamiento de adolescentes que tienen problemas de conducta
y de consumo de drogas en Venezuela. Estos programas de tratamiento se hen mostrado eﬁcaces
en otros paises, pero nunca han sido desarrollados para uso en Venezuela. Tu participacion
alurnbrara maneras en que podamos adapter estos programas para que scan implementados en
este pais.

BENEFICIOS: Tu participacion en este estudio tiene e] potencial de beneﬁciarte, tal vez por
sentirte valorizado(a) como una ﬁrente importante de inforrnacion valiosa para un proyecto que
tiene potencial para ayudar tanto a los adolescentes y sus farnilias, como a la misma escuela.
Esperamos que esto sea causa de mucha satisfaccién y contribuya en la autoestima para quienes
participan. Tal vez también seas beneﬁciario en el futuro con un descenso en actividades
relacionadas con la droga en tu colegio, trabajo o vecindario, y un aumento en la calidad de tus
relaciones personales. Ademas, a1 prestar tu apoyo a1 desarrollo de este estudio, entidades
gubemamentales y otros organismos que se interesan en ayudar a la familia venezolana y a las
escuelas podrian llegar a estar mas atentos a las necesidades de los padres, adolescentes y
escuelas.

RIESGOS: En toda investigacion cientiﬁca existen riesgos para los que participan en dichas
investigaciones. Participar en este estudio podria hacerte pensar en temas relacionados con tu
familia que te incomoden o te hagan recordar momentos no placenteros en tu vida Si
experimentas algunos efectos adversos por haber participado en este estudio, te animamos a
hablar con tu maestra o director del colegio. Elios te pueden ayudar, o te pueden buscar una cita
conﬁdencial con el psicologo del colegio, si es que te hace falta alguien mas con quien hablar. 0
también puedes contactar a1 psicologo directamente.

PARTICIPACION: Participar en este estudio consiste en responder a un cuestionario que ha
sido implementado con jévenes en Panama, Costa Rica, Nicaragua, El Salvador, Honduras,
Guatemala y la Republica Dominicana El cuestionario te pregunta sobre diferentes aspectos de
tu relacion con tu familia, amistades, maestros, religion, vecindario, las drogas y el alcohol.

Esta forma de consentimiento fue aprobada por ei Consejo Institucional de Revision (SIRB) de Ciencias
Sociales/Conductual/Educativo de la Universidad del Estado de Michigan. Aprobada a partir de 04/27/07
hasta 04/26/08. Esta version reempiaza a todas las versiones anteriores. lRB#07-320.

211

Page: 2 of2

Debido a los riesgos previamente mencionados, enfatizamos que tu participacion en este estudio
es voluntaria. T ienes e1 derecho de decidir a no participar en este estudio, de rehusar a contestar
cualquier pregunta, o terminar tu participacion en cualquier momento, sin ninguna penalidad. Si
decides no participar hay un aula de estudio supervisado donde puedes hacer tareas mientras
terminarnos con el proyecto.

PRIVACIDAD Y CONFIDENCIALIDAD: Ia informacidn recogida en este estudio se
mantendra en forma anonima. Esto signiﬁca que ni tu nombre, ni cualquier otro dato que se
podria usar para identiﬁcarte, aparecera en el cuestionario o la hoj a de respuestas. Seleccionaras
un sobre a1 lazar que tiene una hoja de respuestas junto con un cuestionario. E1 sobre, y la hoja
de respuesta tiene un m’rmero que nos permite saber a cual colegio corresponde 1a data. Sin
embargo, debido a que no 56 cual numero vas a seleccionar, ni quien eres, es imposible que
alguien relacione una hoj a de respuestas con una persona. Quiero que preserves 1a hoja de la cual
estamos leyendo para una referencia en caso de que tengas alguna pregunta en el futuro.

COSTOS Y RECOMPENSAS: Responder al cuestionario se llevara aproximadamente 45
minutos. Tu participacion en este proyecto no debe costarte nada mas alla que el tiempo que
inviertes. Se te dara un obsequio en agradecimiento por tu participacion a1 terminar con el
cuestionario.

PREGUNTAS Y CONTACTO: Si en cualquier momenta tienes algunas preguntas o
inquietudes acerca de este estudio, puedes contactar al Sr. Ronald Cox a: (517) 432-3328 0 per
correo electronico a: coxronal@msu.edu, 0 pct correo normal a: 107 Human Ecology, Michigan
State University East Lansing, MI 48824, USA También puedes contactar a Sr. Cox en
Venezuela a1 0212-915-0455. También puedes contactar al Dr. Adrian Blow at (517) 432—7092,
por correo electronico a blow@msu.edu, o por correo normal a: 3B Human Ecology, Michigan
State University, East Lansing, MI 48824 USA. De tener preguntas o inquietudes en cuanto a tus
derechos en este estudio, o te sientes disgustado con cualquier aspecto de este estudio, por favor
comunicate con el Dr. Peter Vasilenko por teléfono a: (517) 355-2180, por correo electronico a:
irb@msu.edu o por correo normal 3: 202 Olds Hall East Lansing, MI 48824, USA.

Si estas de acuerdo en participar en el presente estudio, mantente sentado y se te dara un
cuestionario. Tu presencia en el salon de clase indica tu participacion voluntaria en este estudio.
Si no desees participar en el estudio, puedes acompanar a tu profesor a otra aula para hacer
tareas.

Esta forma de consentimiento fue aprobada por el Consejo lnstitucional de Revision (SIRB) de Ciencias
Sociales/ConductuaI/Educativo de la Universidad del Estado de Michigan. Aprobada a partir de 04l27/07
hasta 04/26/08. Esta version reemplaza a todas las versiones anteriores. lRB#07-320.

212

Page: 1 of 2

Michigan State University
Facultad de Ecologia Familiar y del Niﬁo

Proyecto: Hacia una teoria ecologica del desarrollo del abuso de
drogas en los adolescentes venezolanos

Hoja de Asentimiento Informado

PROPOSITO: Este estudio tiene como objetivo aprender mas acerca de como los padres,
maestros, y comunidades cooperan y colaboran para evitar que los nifios se involucren en las
drogas. Los resultados de este estudio seran empleados en el desarrollo de nuevos programas
para la prevencién y el tratamiento de adolescentes que tienen problemas de conducta y de
consumo de drogas en Venezuela. Estas programas de tratamiento se han mostrado eﬁcaces en
otros paises, pero nunca han sido desarrollados para uso en Venezuela Tu participacidn
alumbrara maneras en que podamos adapter estas programas para que scan immementadas en
Venezuela.

BENEFICIOS: Tu participacion en este estudio tiene el potencial de beneﬁciarte. Tal vez te
beneﬁcies por sentirte valorizado(a) como una fuente respetada de informacion valiosa para un
proyecto que tiene potencial para ayudar a los adolescentes y sus families, tanto como la misma
escuela. Esperamos que esto sea causa de mucha satisfaccién y autoestima para qui-es
participan. T a] vez también seas beneﬁciario en el futuro de un descenso en actividades
relacionados con la droga en tu colegio, habajo, o vecindario, y un aumento en la calidad de tus
relaciones personales. Ademas, a1 prestar tn apoyo al desarrollo de este estudio, entidades
gubemamentales y otros organismos que se interesan en ayudar a la familia venezolana y a las
escuelas puedan llegar a estar mas atentos a las necesidades de los padres, adolescentes, y
escuelas.

RIESGOS: En toda investigacion cientiﬁca existen riesgos para los que participan en dichas
investigaciones. Al participar en este estudio pueda que te lleve a pensar en temas relacionados
con tu familia que te incomoden o te hagan recordar de momentos no placenteros en tu vida. Si
experimentas algunos efectos adversos por haber participado en este estudio, te animamos a
hablar con tu maestra o director del colegio. Ellos te pueden ayudar, o te pueden buscar una cita
conﬁdencial con el psicélogo del colegio, si es que te hace falta alguien mes con quien hablar.
O, también puedes contactar al psicologo directamente.

Esta forma de consentimiento fue aprobada por el Consejo lnstitucional de Revision (SIRB) de Ciencias
Sociales/ConductuaI/Educativo de la Universidad del Estado de Michigan. Aprobada a partir de 04/27/07
hasta 04/26/08. Esta version reemplaza a todas las versiones anteriores. lRB#07-320.

Page: 2 of2

PARTICIPACION: La participacion en este estudio consiste en responder a un cuestionario
que ha sido implementado con jovenes en Panama, Costa Rica, Nicaragua, El Salvador,
Honduras, Guatemala, y Republica Dominicana. El cuestionario te pregunta sobre diferentes
aspectos de tu relacion con tu familia, amistades, maestros, religion, vecindario, y las drogas y el
alcohol. Debido a los riesgos previamente mencionados, enfatizamos que tu participacion en
este estudio es voluntaria. Tienes el derecho de decidir a no participar en este estudio, de rehusar
a contestar cualquier pregunta, o terminar tu participacion en cualquier memento, sin ninguna
penalidad Si decides no participar hay un aula de estudio supervisado donde puedes hacer tarees
mientras terminarnos con el proyecto. Se tomara aproximadamente 45 minutos para llenar el
cuestionario, y tu participacion no te ocasionara ningi’m otro gasto mas alla de tu tiempo. En
agradecimiento por tu participacién en el estudio te estaremos obsequiando un boligrafo.

PRIVACIDAD Y CONFIDENCIALIDAD: La informacién recogida en este estudio se
mantendra en forma anonima. Esto signiﬁca que ni tu nombre, ni cualquier otro dato que se
podria usar para identiﬁcarte apareceré en el cuestionario 0 en la hoja de respuestas.
Seleccionaras un sobre a1 lazar que tiene una hoja de respuestas junto con un cuestionario. Bl
sobre y la hoja de respuesta tiene un numero que nos permite saber a1 cual colegio corresponde la
data. Sin embargo, debido a que no sé cual nt'unero vas a seleccionar, ni quien eres, es imposible
que a] guien relacione una hoj a de respuestas con una persona Quiero que preserves la hoja de la
cual estamos leyendo para una referencia en el caso que tienes alguna pregunta en el futuro.

PREGUNTAS Y CONTACTO: Si en cualquier momento tienes algunas preguntas o
preocupaciones acerca de este estudio, puedes contactar a1 Sr. Ronald Cox a: 001-517-432-3328
0 par correo electro'nico a: coxronal@msu.edu, o por correo normal a: 107 Human Ecology,
Michigan State University East Lansing, MI 48824, USA. Tambien puedes contactar a Sr. Cox
en Venezuela a: 0212-915-0455. También puedes contactar a1 Dr. Adrian Blow at 001-517-432-
7092, por correo electronico a blow@rpsu.edp, 0 par correo normal a 3B Human Ecology,
Michigan State University, East Lansing, MI 48824 USA. En el caso que tengas preguntas o
preocupaciones en cuanto a tus derechos en este estudio, o te sientes disgustado con cualquier
aspecto de este estudio, por favor comunicate con el Dr. Peter Vasilenko por teléfono a: 001-
517-355-2180, por correo electro'nico a: irb@msu.edu o por correo normal a: 202 Olds Hall East
Lansing, MI 48824, USA.

Si estas de acuerdo en participar en el presente estudio, mantente sentado y se te dara un
cuestionario. Tu presencia en el salon de clase indica tu participacion voluntaria en este estudio.
Si no desees participar en el estudio, puedes acompafiar a tu maestra a otra aula para hacer tareas.

Esta forma de consentimiento fue aprobada por el Consejo Institucional de Revision (SIRB) de Ciencias
Sociales/ConductuaI/Educativo de la Universidad del Estado de Michigan. Aprobada a partir de 04/27/07
hasta 04/26/08. Esta version reemplaza a todas las versiones anteriores. iRB#07-320.

214

Page: 1 of 2

Michigan State University
Facultad de Ecologia Familiar y del Niﬁo

Hoja de Consentimiento Informado
Hacia un teoria ecologim del desarrollo del abuso de drogas en los adolescentes
venezolanos

PROPOSITO: Este estudio tiene como objetivo aprender mas acerca de como los padres,
maestros y comunidades cooperan y colaboran para evitar que los niiios se involucren en las
drogas. Los resultados de este estudio seran empleados en el desarrollo de nuevos programas
para la prevencién y el tratamiento de adolescentes que tienen problemas dc conducta y de
consumo de drogas en Venezuela. Estos programas de tratamiento se han mostrado eﬁcaces en
otros paises, pero nunca han sido desarrollados para uso en Venezuela. Tu participacién
alumbrara maneras en que podamos adaptar estos programas para que scan implementadas en
Venezuela.

BENEFICIOS: Tu participacidn en este estudio tiene e1 potencial de beneﬁciarte. Tal vez te
beneﬁcies por sentirte valorizado(a) como una fuente respetada de informacion valiosa para un
proyecto que tiene potencial para ayudar tanto a los adolescentes y sus familias como a la misma
escuela. Esperamos que esto sea causa de mucha satisfaccion y contribuya en la autoestima para
quienes participan. Tal vez también seas beneﬁciario en el futuro de un descenso en actividades
relacionadas con la droga en tu trabajo o vecindario y un aumento en la calidad de tus relaciones
personales. Ademas, al prestar tu apoyo a1 desarrollo de este estudio, entidades gubemamentales
y otros organismos que se interesan en ayudar a la familia venezolana y a las escuelas pueden
llegar a estar mas atentos a las necesidades de los padres, adolescentes y escuelas.

RIESGOS: En toda investigacion cientiﬁca existen riesgos para los que participan en dichas
investigaciones. Participar en este estudio podria llevarte a pensar en temas relacionados con tu
familia que te incomoden o te hagan recordar momentos no placenteros en tu vida. Si
experimentas algunos efectos adversos por haber participado en este estudio, te animamos a
hablar con un psicélogo o consejero que conozcas o contactar e1 psicologo cuya informacién esta
al ﬁnal de este documento.

PARTICIPACION: Participar en este estudio consiste en responder a un cuestionario que ha
sido implementado en escuelas en Panama, Costa Rica, Nicaragua, El Salvador, Honduras,
Guatemala y Repﬁblica Dominicana. El cuestionario pregunta sobre diferente aspectos del
arnbiente escolar y los estudiantes. Debido a los riesgos previamente mencionados, enfatizamos
que tu participacion en este estudio es voluntaria. Tienes e1 derecho de decidir no participar en

Subject Initials

Date
Esta forma de consentimiento fue aprobada por el Consejo lnstitucional de Revision (SIRB) de Ciencias
SociaIes/Conductuai/Educativo de la Universidad del Estado de Michigan. Aprobada a partir de 04/27/07
hasta 04/26/08. Esta version reempiaza a todas las versiones anteriores. IRB#07-320.

215

Page: 2 of 2

este estudio, de rehusar a contestar cualquier pregunta, o terminar tu participacion en cualquier
momento, sin ninguna penalidad. '

PRIVACIDAD Y CONFIDENCIALIDAD: La inforrnacion recogida en este estudio se
mantendra en forma conﬁdencial. Esto signiﬁca que ni tu nombre, ni el nombre del plantel
donde trabajas, ni cualquier otro dato que se podn'a usar para identiﬁcarte, aparecera en los
documentos que se desarrollaran relacionados con este estudio. Se te esta dando una copia de
esta hoj a como referencia en el caso que tienes alguna pregunta en el futuro.

COSTOS Y RECOMPENSAS: Responder e1 cuestionario se llevara aproximadamente 45
minutos. Tu parn'cipacion en este proyecto no debe costarte nada mas alla que el tiempo que
inviertes. Se te dara un pequei'io obsequio en agradecimiento por tu participacion al terminar con
la recoleccion de datos.

PREGUNTAS Y CONTACTO: Si en cualquier memento tienes algunas preguntas o
inquietudes acerca de este estudio, puedes contactar a1 Sr. Ronald Cox a: (517) 432-3328 0 per
correo electrénico a: coxronal@msu.edu, o por correo normal a: 107 Human Ecology, Michigan
State University East Lansing, MI 48824, USA. También puedes contactar a Sr. Cox en
Venezuela a1 0212—915-0455. También puedes contactar al Dr. Adrian Blow at (517) 432-7092,
por correo electronico a blow@msu.edu. 0 per correo normal a: 3B Human Ecology, Michigan
State University, East Lansing, MI 48824 USA. En el caso que tengas preguntas o inquietudes
en cuanto a tus derechos en este estudio, o te sientes disgustado con cualquier aspecto de este
estudio, por favor comunicate con el Dr. Peter Vasilenko por teléfono a: (517) 355-2180, por
correo electronico a: irb@msu.edu o por correo normal a: 202 Olds Hall East Lansing, MI
48824, USA. Si gustas hablar con algt’m profesional acerca de cualquier incomodidad que
sientas por causa de tu participacion en este estudio, 1a psicologa Lic. Mariela Rodriguez esta al
tanto de este estudio y esta dispuesta a ayudarte aqui en Caracas. Su numero es: 0414-257-9777

Si estas de acuerdo en participar en el presente estudio, por favor coloca tu nombre y ﬁrma e1
documento. Tu ﬁrma indica tu participacion voluntaria en este estudio.

Nombre del participante:

 

F irma: Fecha:

 

 

Subject Initials

Date
Esta forma de consentimiento fue aprobada por el Consejo Institucional de Revision (SIRB) de Ciencias
Sociales/Conductual/Educativo de la Universidad del Estado de Michigan. Aprobada a partir de 04/27/07
hasta 04/26/08. Esta version reempiaza a todas las versiones anteriores. |RB#O7-320.

 

216

APPENDIX E

REPUBLICA Bouvxmxm DE VENEZUELA
MINISTERIO DE EDUCACION
Y naponras

 

ZONA EDUCA'I'IVA DEL DISTRITO CAPITAL
Cludad

N ° 1703-06

Caracas, 28 de julio de 2.006.
Ciudadano
SR. RONALD COX
Presente-

Me dirijo a usted, en atencién a su comunicacién de fecha 21-06-2.006 recibida en esta Zena
Educativa el 21-07-2.006 mediante 1a cual da a conocer Proyecto de Irrvestigacidn ereado por la
Universidad Estatal de Michigan, que tiene por objeto tratar en rasgos generales el uso de las drogas
por parte de los adolescentes, por tal motivo solicita autorizacion para ingresar a los‘ planteles a 10s
ﬁnes de apliear encuesta para recoger informacién respecto a los factores de riesgo que se relacionan
con el uso de dichas sustancias. Al respecto tengo a bien informarle que este Despacho ve con agrado'
este tipo de proyecto que favorece a nuestra poblacién estudiantil de jévenes y adolescentes, y como
Estado y responsable de los mismos tenemos 1a obligacién de asegm'ar que reciban la informacién
veraz, plural y adecuada a su desarrollo. En tal sentido, por la presente se le autoriza para desarrollar la
actividad propuesta una vez que hallan sido revisados y aprobados por la autoridad educative
competente de cada Distrito Escolar y bajo supervisidn de los mismos, los instrumentos que seran
utilizados para tal ﬁn. Dicha autorizacion se concede en virtud a lo sefialado en el Articulo 51 de la Ley
Organica para la Proteccién del Nifio y del Adolescente en relacién a la proteccién contra sustancias
alcohélieas, estupefacientes y sicotrépicas que establece:

217

“ElEstadoconlaactivapaﬂictpaddndelasocieMdebegm-andza'
politicos y programas deprevencién contra el uso ilt'cr'to dc srrstaudas alcohdlicas,
Marianas y sicoa'ripicas...”

Asi mismo 1e agradezco que a1 ﬁnalizar e1 proyecto informe a esta Dependencia sobre los
resultados del mismo.
Atentamente,

   
 

PROF. Ahmnmcmaz as

DIRECTOR DE LA ZONA EDU u.
DEL DISTRITO CAPITAL

ZEDC/AJ dé’mt.

“2.006 ANO BICEHENARIO DBL GENERAL FRANCISCO DE WANDA YDE LA PARTICH’ACION PROHGONICA Y DEL PODER POPUIAR”

218

 

REPUBLICA BQLNARIANA DEVENEZUELA
MINISTERIO DE EDUCACION Y DEPORTES

 

ZONA EDUCATIVA DEL DISTRI'I'O CAPITAL
DISTRITO ESCOLAR N" 2
Ciudad.-

Caracas, 14 de Julio del 2006
Ciudadano:
Mr. Ronald Cox
107 Human Ecology Building
Dept. 01' Family and child Ecology
Mlchiganm State University
East Lansing, MI

Estimado Sr. Cox,

Nos complace informarnos de en interés por el bienestar de la familia venezolana.

Estamos seguros de que el proyecto de investigacion sobre factores de riesgo y de
proteccién en el consumo de drogas hard una contribucién importante en la hrcha por
proheger a nuestrajuventud contra este problema.

Sirvase esta cata para informarle de nuestro deseo de colaborar con Ud. y la
Universidad de Michigan State on llevar a cabo el p‘oyecto planteado. Una vez revisados y
aprobados los instrumentos que seran utilizados en el proyecto, nos parece bien oﬁ‘ecerle
los planteles del Distrito Escolar N° 2 del Distrito Capital, para encuesta a los estudiantes
de basica, media, diversiﬁcaday profesional (de 12 a 1? adds), y entrevistar a maestros y
otro personal que desempeila labores en dichos planteles.

Ademas entendemos que sera necesario facilitarie el contacto con los padres y
representantes de los estudiantes con el ﬁn de logra- e1 consentimiento apropiado para que
sus representados participan en el estudio plmteado.

De nuevo 1e expresamos nuestra cornplacencia por su preocupacién por el
mejoramiento de nuesh'a sociedad a través del estudio del consumo de drogas entre los
jovenes Venezolanos y quedamos en espera de un informe detallado de los hallazgos al
ﬁnalizar el estudio. ‘

219

. ' -:P to”

2'...“ - "he. 0‘. 9
_ .ﬁoht n .0040
\‘ ' '

t

r ‘4;
. obi-e
if a
Y

    

     

I‘

 

in;

Sin otro particular a que hacer referencia, quedo de usted

)l 5 “A 3i-
J.» imitation £04122

  

. 90‘)
afifyg U ﬂ 96‘? J': \ cl
{Us ' .
.3: . ‘ .\
a 11;; ' w l
«,4, s \‘t'
\ q 96%
”Raw . .4 :irinuiutra

saw u./ l 184' z o ascomn N° 2

D'I'I‘O N°2IMIDIMSth

”20“ AND BICENTENARIO DEL JURAMENTO DEL GENERALISIMO FRANCISCO DE MIRANDA Y DE LA
PARTICIPACION PROTAGONICA Y DH. PODIR POPULAR”

220

Republics Bolivariana de Venezuela
Ministerio de Educaclon y Deportes

@ Caracas, 25 de julio de 2006
v

Zena Educative del Distrito Capital
Distrito Escolar N° 4
Caricuao - Caracas

Licenciado

Francisco Villamediamr

Coordinador Convem'o (MED-CE V-C PE)
Presente

Estimado Lic. Villamedr'ana:

Me dirijo a usted cordialmente en ocasion de saludarle y a la vez dar respuesta a su
comrrnicacio’n de fecha 12 de julio del aﬁo en curso, donde nos recomienda a1 cirrdadano
Msc., Ronald B. Cox, (investigador de la Universidad Estatal de Michigan), el cual
aplicard 1m Instrumento para rm proyecto, sobre losfactores que inciden en el consumo de
drogos en jo'venes de edades comprendia’as entre 12 y 1 7 aﬁos, en algunos planteles
adscrztos a este Distrito Escolar, Ios crrales se mencionan a continuacion:

 

Sin mos a que hacer referencia y agradeciendo el aporte por el mejoramiento de
nuestro sociedad, a través de dicho proyecto entre losjovenes venezolanos, se suscribe.

Atentamente,

We;

Prof Magaly asqu z
Jefe Distrito Escolar N" 4

 

DE 4/M l Tl'idc
25—0 7-2006

Direccién: Sector UD-3, Bloque 1, PB, Caricuao. Telefax: 431-49-89

O
“2006, ANO BICENTENARIO DEL JURAMENTO DEL GENERALLSIMO FRANCISCO DE
MIRANDA YDE LA PART ICIPACION PROTA GONICA Y DEL PODER POPULAR”

221

REPUBLICA BOLIVARIANA DE VENEZUELA
MINISTERIO DE EDUCACION Y DEPORTE
Unidad Educativa Nacional “PEDRO F ONTES”
Montalban — La Vega

Caracas, 14 de Julio de2006

Mr. Ronald Cox

107 Human Ecology Building
Dept of Family and Child Ecology
Michigan State University

East Lansing, MI

Estimado Sr. Cox,

Nos complace inforrnamos de su interes por el bienestar de la Farnilia
venezolana

Estamos seguros de que el proyectote investigacion sobre “Factores de riesgo y
de proteccién en el consumo de drogas” hara tma oontribucibn importante en la lucha
por proteger a nuestra juventud contra este ﬂagelo.

Le comunico nuestro deseo de oolaborar con usted y la Universidad de Michigan
State en llevar a cabo el proyecto planteado. Una vez revisados y aprobados los
instrumentos que seran utilizados en el proyecto, nos parece bien ofrecerle los planteles
del Distrito Escolar N° 3 del Distrito Capital, para encuestar a los estudiantes de Basica,
Media, Diversiﬁcada y Profmional (de 12 a 17 ar‘ios), y entrevistar a maestrOs y otro
personal que desempeﬂa labores en dichos Planteles.

Ademas 1e facilitamos e1 contacto con los padres y representantes de los
estudiantes con el ﬁn de lograr e1 consentimiento apropiado para que sus representados
participen en el estudio plmteedo.

De nuevo 1e expresamos nuestra oomplacencia por su preocupacion por mejorar
nuestra sociedad a través del estudio de consumo de drogas entre los jévenes
Venezolanos quedando en espera de un informe detallado de los resultados a1 ﬁnalizar
el estudio.

‘ Atentamente,

    
 

- a del Distrito Escolar N° 3
Directora (e)

222

REFERENCES

Anderman, E. M. (2002). School effects on psychological outcomes during adolescence.
Journal of Educational Psychology, 94, 795—809.

Anderson, A. R., & Henry, C. S. (1994). Family system characteristics and parental
behaviors as predictors of adolescent substance use. Adolescence, 29, 405—420.

Anthony, J. C. & the PACARDO Collaborators. (2001). The PACARDO: An
epidemiological study of adolescent health, behavior, and adaptation in seven
Latin American countries. Johns Hopkins University, Baltimore, MD, Electronic
Collaboratory for the Investigation about Drugs.

Apospori, E., Vega, W. A., Warheit, G. J ., & Gil, A. G. (1995). A longitudinal study of
the conditional effects of deviant behavior on drug use among three racial/ethnic
groups of adolescnets. In H. B. Kaplan (Ed). Drugs, crime and other deviant
adaptations: Longitudinal studies (pp. 211-230) New York: Plenum.

Baldwin, S. A., Murray, D. M., Shadish, W. R. (2005). Empirically supported treatments
or Type I errors? Problems with the analysis of data from group-administered
treatments. Journal of Consulting and Clinical Psychology, 73(5) 924-935.

Barnes, G. M., Reifman, A. S., Farrell, M. P., Dintcheff, B. A. (2000). The effects of

parenting on the development of adolescent alcohol misuse: A six-wave latent
growth model. Journal of Marriage and the Family, 62, 175—186

Bateson, G. (1972). Steps to an ecology of mind. Chicago: University of Chicago Press.

Bateson, G., Jackson, D. D., Haley, J ., & Weakland, J. H. (1968). A note on the double
bind (1962). In D. D. Jackson (Ed.), Communication, family and marriage (pp.
55-62). Palo Alto, CA: Science and Behavior Books.

Bauman, K. E., and Ennett, S. T. (1994). Peer inﬂuence on adolescent drug use.
American Psychologist, 49, 820—822.

Baumrind, D. (1991). The inﬂuence of parenting style on adolescent competence and
substance use. Journal of Early Adolescence, 11, 56-95.

Beauvais, F. & Oetting, E. R. (2002). Variances in the etiology of drug use among ethnic
groups of adolescents. Public Health Report, 117(Suppl 1), S8-Sl4.

Becvar, D. S., & Becvar, R. J. (2003). Family therapy: A systemic integration (2nd ed).
Needham Heights, MA: Allyn and Bacon.

Belcher, H. M., & Shinitzky, H. E. (1998). Substance abuse in children: Prediction,
protection, and prevention. Archives of Pediatrics &Adolescent Medicine, 152,
952-960.

Blum, R. W., Beuhring, T., Shew, M. L., Bearinger, L. H., Sieving, R. E., & Resnick, M.
D. (2000). The effects of race/ethnicity, income, and family structure on
adolescent risk behaviors. American Journal of Public Health, 90(12), 1879-1884.

Bollen, K. A. (1989). Structural equations with latent variables. New York: Wiley.
Bowen, M. (1974). A family systems approach to alcoholism. Addictions, 21(2), 3-11.

Bronfenbrenner, U. & Ceci, S. J ., (1994). Nature-nurture reconceptualized in

developmental perspective: A bioecological model. Psychological Review.
101(4), 568-586. '

Bronfenbrenner, U. (1979). The ecology of human development. Cambridge, MA:
Harvard University Press.

Bronfenbrenner, U. (1989). Ecological systems theory. Annals of Child Development, 6,
187-249.

Bronfenbrenner, U. (1995). Developmental ecology through space and time: A future
perspective. In P. Moen & G. H. Elder Jr. (Eds), Examining lives in context:
Perspectives on the ecology of human development (pp. 619—647). Washington,
DC: American Psychological Association.

Brook, J. S., Brook, D. W., & Pahl, K. (2006). The developmental context for adolescent
substance abuse intervention. In H. A. Liddle, & C. L. Rowe (Eds), Adolescent
substance abuse: Research and clinical advances (pp. 25-51). Cambridge
University Press

Brook, J. S., Brook, D. W., De La Rosa, M., Duque, L. F ., Rodriguez, E., Montoya, I. D.,
et a1. (1998). Pathways to marijuana use among adolescents: cultural/ecological,

family, peer, and personality inﬂuences. Journal of the American Academy of
Child and Adolescent Psychiatry, 37, 759-766

Brook, J. S., Brook, D. W., Gordon, A. S., Whiteman, M., & Cohen, P. (1990). The
psychosocial etiology of adolescent drug use: A family interactional approach.
Genetic, Social, and General Psychology Monographs, 116, 111—267.

Bubolz, M. M., & Sontag, M. S. (1993). Human ecology theory. In P. G. Boss, W. J.
Doherty, R. LaRossa, W. R. Schumm, & S. K. Steinmetz (Eds), Sourcebook of
family theories and methods: A contextual approach (pp. 419—448). New York:
Plenum.

224

Calvert, W. J. (1997). Protective factors within the family, and their role in fostering
resiliency in African American adolescents. Journal of Cultural Diversity, 4(4),
110-117.

Capaldi, D. M. & Dishion, T. J. (1988). Psychometric Properties of Fourteen Latent
Constructs from the Oregon Youth Study. New York, Berlin, Heidelberg, London,
Paris, Tokyo: Springer-Verlag.

Capaldi, D., DeGarmo, D., Patterson, G. R., & Forgatch, M. (2002). Contextual risk
across the early life span and association with antisocial behavior. In J. B. Reid,
G. R. Patterson, & J. Snyder (Eds), Antisocial behavior in children and
adolescents (pp. 123-145).

Carlo, G., Fabes, R. A., Laible, D., Kupanoff, K. (1999). Early adolescence and
prosocial/moral behavior 11: The role of social and contextual inﬂuences. Journal
ofEarly Adolescence, 19(2), 133-147.

Castro, F. G., Barrera, M., Martinez, C. R. (2004). The cultural adaptation of prevention

interventions: resolving tensions between fidelity and fit. Prevention Science, 5,
41—45.

Chassin, L., Presson, C. C., Sherman, S. J., & Edwards, D. A. (1992). Parent education
attainment and adolescent cigarette smoking. Journal of Substance Abuse, 4, 219—
234.

Chassin, L., Ritter, J ., Trim, R., & King, K. (2003). Adolescent substance use. In R.
Barkley & E. Mash (Eds), Handbook of child psychopathology (pp. 119—232).
New York: Plenum Press.

Chen, C., Dormitzer, C. M., Bejarano, J ., & Anthony, J. C. (2004). Religiosity and the
earliest stages of adolescent drug involvement in seven countries of Latin
America. American Journal of Epidemiology, 159(12),1 180-1 188.

Chilcoat, H. D., Breslau, N., Anthony, J. C., (1996). Potential barriers to parent
monitoring: Social disadvantage, marital status, and maternal psychiatric disorder.
Journal of the American Academy of Child & Adolescent Psychiatry, 35(12),
1673-1682.

Clark, D. B. & Winters, K. C. (2002) Measuring risks and outcomes in substance use
disorders prevention research. Journal of Consulting and Clinical Psychology, 70,

1207—1223.

Clark, D. B. (2004). The natural history of adolescent alcohol use disorders. Addiction,
99(suppl. 2), 5-22.

Cloninger, C. R. (1987). Neurogenetic adaptive mechanisms in alcoholism. Science,
236(4800), 410-416.

225

Colby, S. M., Lee, C. S., Lewis-Esquerre, J., Esposito-Smythers, C., & Monti, P. M.
(2004). Adolescent alcohol misuse: Methodological issues for enhancing
treatment research. Addiction, 99 (52), 47—62.

Comision Nacional Contra El Uso Ilicito de las Drogas. (2006). Estadistica consumo de
drogas en Venezuela. Republica Bolivariana De Venezuela Ministerio Del
Interior Y Justicia Oﬁcina Nacional Antidrogas, retrieved May, 2006 at:
http://www.ona.gob.ve/Pdf/Consumo_2006.pdf

Conchas, G. (2001). Structuring failure and success: Understanding the variability in
Latino school engagement. Harvard Educational Review, 71, 475-504.

Cook, T. D., Herman, M. R. Phillips, M., & Settersten, R. A. (2002). Some ways in
which neighborhoods, nuclear families, friendship groups, and schools affect
changes in early adolescent development. Child Development, 73:1283-1309.

Corroa, M. A., Guindon, G. E., & Sharma, N. (2000). Tobacco control country proﬁles.
Atlanta, GA: American Cancer Society.

Cox, R., & Ray, W. (1994). The role of theory in the treatment of substance abuse.
Contemporary Family Therapy, 16(2), 131-144.

Cummings, R. (1995). Adolescence: A developmental perspective. New York: Harcourt
Brace.

Curran, P. J., Stice, E., & Chassin, L. A. (1997). The relation between adolescent alcohol

use and peer alcohol use: A longitudinal random coefﬁcients model. Journal of
Consulting and Clinical Psychology, 65, 130—140.

Darling, N. and Steinberg, L. (1993) Parenting style as context: an integrative model.
Psychological Bulletin, 113, 487—496.

Davila, B., & Guarino, L. (2001). Fuentes de e'stres y estrategias de afrontamiento en
escolares Venezolanos. Revista Interamericana de Psicologia, 35(1), 97-112.

Delva, J ., Bobashev, G., Gonzalez, G., Cedeno, M., & Anthony, J. C. (2000). Clusters of
drug involvement in Panama: Results from Panama's 1996 National Youth
Survey. Drug and Alcohol Dependence, 60, 251-257.

Dishion, T. J ., Patterson, G. R., Stoolmiller, M., & Skinner, M. L. (1991). Family, school,
and behavioral antecedents to early adolescent involvement with antisocial peers.
Developmental Psychology, 27, 172—180.

Dormitzer, C. M. (2004). Family attention and youthful drug use: Protection against
involvement? Unpublished doctoral dissertation, Johns Hopkins University,
Baltimore, Maryland.

226

Dormitzer, C. M., Gonzalez, G. B., Penna, M., Bejarano, J ., Obando, P., Sanchez, M., et
al. (2004). The PACARDO research project: Youthful drug involvement in
Central America and the Dominican Republic. Revista Panamericana De Salud
Publica, 15(6), 400-416

Droomers, M., Schrijvers, C. T. M., Casswell, S., & Mackenbach, J. P. (2003).
Occupational level of the father and alcohol consumption during adolescence:
Patterns and predictors. Journal of Epidemiological Community Health, 5 7, 704—
710.

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

Duncan, S. C., Duncan, T. E., & Strycker, L. A. (2002). A multilevel analysis of
neighborhood context and youth alcohol and drug problems. Prevention Science,
3(2), 125-133.

Eccles, J. 8., Lord, S., Midgley, C. (1991). What are we doing to early adolescents? The
impact of educational contexts on early adolescents. American Journal of
Eduacation, 99(4), 521-542.

Eccles, J. S., Midgley, C., Wigﬁeld, A., Buchanan, C. M., Reuman, D., Flanagan, et al.
(1993). Development during adolescence: The impact of stage-environment ﬁt on
young adolescents’ experiences in schools and in families. American
Psychologist, 48, 90—1 01.

Eisenberg, N., Cumberland, A., Spinrad, T. L., F abes, R. A,. Shepard, S. A., Reiser, M.,
et al. (2001). The relations of regulation and emotionality to children’s
externalizing and internalizing problem behavior. Child Development, 72, 1112——
1134.

Ennett, S. T., Flewelling, R. L., Lindrooth, R. C., & Norton, E. C. (1997). School and
neighborhood characteristics associated with school rates of alcohol, cigarette,
and marij uana use. Journal of Health and Social Behavior, 38, 55—71.

Fabes, R. A., Carlo, G., Kupanoff, K., & Laible, D. (1999). Early adolescence and
prosocial/moral behavior I: The role of individual processes. Journal of Early
Adolescence, 19, 5-16.

Feldman, N. (1989). Desarrollo de la terapia familiar en Venezuela. Revista
Interamericana de Psicologia, 23(1), 5-20.

F eldt, L. S., & Charter, R. A. (2003). Estimating the reliability of a test split into two
parts of equal or unequal length. Psychological Methods, 8, 102-109.

F erketich, S. (1990). Focus on psychometrics: Internal consistency estimates of
reliability. Research in Nursing & Health, 13, 437-440.

227

Fletcher, A. C., Darling, N. E., Steinberg, L., & Dombusch, S. M. (1995). The company
they keep: Relation of adolescents’ adjustment and behavior to their friends’

perceptions of authoritative parenting in the social network. Developmental
Psychology, 31, 300-310.

Furstenberg, E (1994). History and current status of divorce in the US. Children and
Divorce, 4, 29-43.

Furstenberg, F. F. (2005). Banking on families: how families generate and distribute
social capital. Journal of Marriage and Family, 67, 809—821.

Fussell, E. & Palloni, A. (2004). Persistent marriage regimes in changing times. Journal
ofMarriage and Family 66 (5), 1201—1213.

Gfroerer, J ., Wright, D., & Kopstein, A. (1997). Prevalence of youth substance use: The

impact of methodological differences between two national surveys. Drug and
Alcohol Dependence, 4 7, 19-30.

Goldman, D., Oroszi, G., & Ducci, F. (2005). The genetics of addictions: Uncovering the
genes. Nature Reviews Genetics, 6(7), 521-532.

Golub, A., & Johnson, B. D., (2001). Variation in youthful risks of progression from
alcohol and tobacco to marijuana and to hard drugs across generations. American
Journal of Public Health, 91(2), 225—232.

Gonzalez, 0. (2005). Factores de riesgo en el consumo de drogas en los estudiantes de la
Escuela de Arquitectura. Multiciencias, 5(1), 1-21. Retrieved May 2006, at:
www.serbi.luz.edu.ve/scielo.php?script=sci_arttext&pid=S13 l7.

Gorman-Smith, D., Tolan, P. H., Zelli, A., & Huesmann, L. R. (1996). The relation of
family functioning to violence among inner-city minority youth. Journal of
Family Psychology, 10, 115—129.

Gorman-Smith, D., Tolan, P. H., Zelli, A., & Huesmann, L. R. (1996a). The relation of
family functioning to violence among inner-city minority youths. Journal of
Family Psychology, 10, 115-129.

Grifﬁn, K. W., Botvin, G. J ., Scheier, L. M., Diaz, T., & Miller, N. L. (2000). Parenting
practices as predictors of substance use, delinquency, and aggression among

urban minority youth: Moderating effects of family structure and gender.
Psychology of Addictive Behaviors, 14 (2), 174-184.

Grifﬁn, K. W., Botvin, G. J ., Scheier, L. M., Doyle, M. M., & Williams, C. (2003).
Common predictors of cigarette smoking, alcohol use, aggression, and

delinquency among inner-city minority youth. Addictive Behaviors, 28, 1141-
1148.

228

Hawkins, J. D., Catalano, R.F., & 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., Van Horn, M. L., Arthur, M. W. (2004). Community variation in risk and
protective factors and substance use outcomes. Prevention Science, 5(4), 213-220.

Hecht, M. L., Marsiglia, F. F., Elek, E., Wagstaff, D. A., Kulis, S., Dustman, P., &
Miller-Day, M. (2003). Culturally grounded substance use prevention: An

evaluation of the keepin’ it R.E.A.L. curricultun. Prevention Science, 4(4), 233-
248.

Heise, D. R. (1970). Causal inference from panel data. Sociological Methodology, 2, 3-
27.

Helms, J. E., Henze, K. T., Sass, T. L., & Mifsud, V. A. (2006). Treating Cronbach’s
Alpha reliability coefﬁcients as data in counseling research. The Counseling
Psychologist; 34(5), 630-660.

Henry, K. L., & Slater, M. D. (2007). The contextual effect of school attachment on
young adolescents' alcohol use. The Journal of School Health, 77(2), 67-74.

Hemandez-Ponce, L. E., & Reimel, S. (2004). Calidad de vida y participacion
comunitaria: Evaluacion psicosocial de proyectos urbanisticos en barrios pobres.
Revista Interamericana de Psicologia, 38(1), 73-86.

Hettema, J. M., Corey, L. A. & Kendler, K. S. (1999). A multivariate genetic analysis of
the use of tobacco, alcohol, and caffeine in a population based sample of male and
female twins. Drug and Alcohol Dependence, 57, 69—78.

Hill, L. G., & Werner, N. E. (2006). Afﬁliative motivation, school attachment, and
aggression in school. Psychology in the Schools, 43(2), 231-246.

Hirschi, T. (1969). Causes of delinquency. Berkley: University of California Press.

Hirschi, T., & Gottfredson, M. (1993). Commentary: Testing the general theory of crime.
Journal of Research in Crime and Delinquency, 30(1), 47-54.

Hofstede, G. (2001). Culture 's consequences: Comparing values, behaviors, institutions,
and organizations across nations (2nd ed.). Beverly Hills CA: Sage.

Hooker, J. (2005). Indigenous inclusion/Black exclusion: Race, ethnicity and
multicultural citizenship in Latin America. Journal of Latin American Studies, 37,
285-310.

Hussong, A. M., (2002). Differentiating peer contexts and risk for adolescent substance
use. Journal of Youth and Adolescence, 31(3), 207—220.

229

Hussong, A. M.,& Chassin, L. (1997). Substance use initiation among adolescent

children of alcoholics: Testing protective factors. Journal of Studies on Alcohol,
58, 272-279.

Ialongo, N., Poduska, J ., Werthamer, L., & Kellam, S. (2001).The distal impact of two
ﬁrst-grade preventive interventions on conduct problems and disorder in early
adolescence. Journal of Emotional and Behavioral Disorders, 9(3),]46-160.

Ianotti, R. J ., Bush, P. J., and Weinfurt, K. P. (1996). Perception of ﬁiends’ use of
alcohol, cigarettes, and marijuana among urban schoolchildren: A longitudinal
analysis. Addictive Behaviors, 21, 615-632.

Jellinek, E. M. (1946). Phases in the drinking history of alcoholics: Analysis of a survey
conducted by the Grapevine, ofﬁcial organ of Alcoholics Anonymous. Quarterly
Journal of Studies on Alcohol, 7, 1-88.

Jessor, R., & Jessor, S. (1977). Problem behavior and psychosocial development. New
York: Academic Press.

Johanson, C. E., Duffy, F. F., & Anthony, J. C. (1996). Associations between drug use
and behavioral repertoire in urban youths. Addiction, 9], 523-534.

Johnson, M. K., Crosnoe, R., & Elder, G. H., Jr. (2001). Student’s attachment and

academic engagement: The role of race and ethnicity. Sociology of Education, 74,
318-340.

Johnston, L. D., O’Malley, P. M. and Bachman, J. G. (1994) National survey results on
drug use from the monitoring the future study, 1975—1993. DHHS publ. no. 94-
3810. US Government Printing Ofﬁce, Washington DC.

Johnston, L. D., O’Malley, P. M., Bachman, G. J. (2001). Monitoring the future national
survey results on drug use 1975-2000, volume 1 secondary students. The
University of Michigan Institute for Social Research. Retrieved November, 2006
at:
http://eric.ed.gov/ERICDocs/data/ericdocs2sql/content_storage_01/0000019b/80/
19/47/bb.pdf

J unger, M., & Marshall, I. H. (1997). The interethnic generalizability of social control
theory: An empirical test. Journal of Research in Crime and Delinquency, 46, 79—
1 12.

Kaplow, J. B., Curran, P. J ., & Dodge, K. A. (2002). Child, parent, and peer predictors of
early-onset substance use: A multisite longitudinal study. Journal of Abnormal

Child Psychology, 30(3), 199-216.

Keeney, B. P. (1983). Aesthetics of Change. New York: Guilford.

Khoury, E. L., Warheit, G. J ., Zimmerman, R. S., Vega, W. A., & Gil, A. G. (1996).
Gender and ethnic differences in the prevalence of alcohol, cigarette, and illicit

drug use in a cohort of young Hispanic adolescents in South Florida. Women and
Helath, 24, 21-40.

King, K. M., Chassin, L. (2004). Mediating and moderated effects of adolescent
behavioral undercontrol and parenting in the prediction of drug use disorders in
emerging adulthood. Psychology of Addictive Behaviors, 18(3), 239-249.

Kirisci, L., Mezzich, A., Tarter, R. (1995). Norms and sensitivity of the adolescent
version of the drug use screening inventory. Addictive behaviors, 20(2), 149-57.

Kline, R. B. (2005). Principles and practice of structural equation modeling. Second
Edition. NY: Guilford.

Kraemer, H. C., Stice, E., Kazdin, A., Offord, D., & Kupfer, D. (2001). How do risk
factors work together? Mediators, moderators, and independent, overlapping, and
proxy risk factors. American Journal of Psychiatry, 158, 848 - 856.

Kumar, R., O’Malley, P. M., Johnston, L. D., Schulenberg, J. E., & Bachman, J. G.
(2002). Effects of school-level norms on student substance use. Prevention
Science, 3(2), 105-124.

Lagerwey, M. D., & Phillips, E. (2003). Voices from the pipeline: High school
completion among rural Latinos. Journal of Cultural Diversity, 10, 42-49.

Lerner, R. M., Petersen, A. C., Brooks-Gunn, J. (1991). Encyclopedia of Adolescence,
Vols. 1, 2. New York: Garland.

Liddle, H. A., & Dakof, G. A. (1995). Efﬁcacy of family therapy for drug abuse:
Promising but not deﬁnitive. Journal of Marital and Family Therapy, 21, 511—
543.

Lodo-Platone, M. (2004). Families and communities. Journal of Prevention and
Intervention in the Community, 27(1), 71-87.

Luscher, K. (1995). Homo Interpretans: On the relevance of perspectives, knowledge,
and beliefs in the ecology of human development. In P. Moen & G. H. Elder Jr.
(Eds), Examining lives in context: Perspectives on the ecology of human
development (pp. 563—597). Washington, DC: American Psychological
Association.

Luthar, S. S., & Becker, B. E. (2002). Privileged but pressured? A study of afﬂuent
youth. Child Development, 73, 1593—1610.

Luthar, S. S., & D’Avanzo, K. (1999). Contextual factors in substance use: A study of

suburban and inner-city adolescents. Development & Psychopathology, 11, 845—
867.

Marcus, R. F., & Sanders-Reio, J. (2001). The inﬂuence of attachment on social
completion. School Psychology Quarterly, 16, 427-444.

Martin, J. P., & Peruga, A. (2002). The global youth tobacco survey: Results in the
Americas. Epidemiological Bulletin, 23, 6—9.

Martinez, C. R., DeGarmo, D. S., & Eddy, J. M. (2004). Promoting academic success
among Latino youth. Hispanic Journal of Behavioral Sciences, 26, 128-151.

Martino, S. C., Collins, R. L., Ellickson, P. L., Schell, T. L., & McCaffrey, D. (2006)

Socio-environmental inﬂuences on adolescents’ alcohol outcome expectancies: a
prospective analysis. Addiction, 101(7), 971—983.

Morgan, S. B. (1988). The autistic child and family functioning: A developmental—family
systems perspective. Journal of A utism and Developmental Disorders, 18, 263-
280.

Moming, A., (in press). Ethnic classiﬁcation in global perspective: A cross-national
survey of the 2000 Census Round. Population Research and Policy Review.

Mundo, M. (2003). Discontinuidad de la institucion escolar y exclusion temprana: Temas
para una agenda de inclusién. C uadernos del Cendes, 20(52), 103-144. Retrieved
August 1, 2007 at:
http://www.revele.com.ve/programas/indice/ria.php?id=13620&rev=cendes

Murguia, E., Zeng-yin, C., & Kaplan, H. B. (1998). A comparison of causal factors in
drug use among Mexican Americans and non-Hispanic Whites. Social Science
Quarterly, 79, 341-360.

Murray, D. M., & Harman, P. J. (1990). Planning for the appropriate analysis in school-
based drug-use prevention studies. Journal of Consulting and Clinical
Psychology, 58(4), 458—468.

Muthen, L. K. & Muthen, B. O. (2006). MPlus User's Guide. Fourth Edition. Los
Angeles, CA: Muthén & Muthén.

Naj aka, S. B. (2001). A meta-analytic inquiry into the relationship between risk factors
and problem behavior. Dissertation Abstracts International, 62(1-A), 340.

Nash, J. K., Bowen, G. L. (1999). Perceived crime and informal social control in the
neighborhood as a context. Social Work Research, 23(3), 171-186.

Navarro, H. M., & Pontillo, C. H. (2002). Autoestima del adolescente y riesgo de
consumo de alcohol. Actualizaciones en Enfermeria, 5(1), 7-12.

Osorio, R., Ever, A., Ortega De Medina, N. M., & Pillon, S. C. (2004). Risk factors
associated with drugs abuse among adolescent students. Revista Latino-
Americano Enfermagem, 12, 369-375. Retrieved May 2006, at:

http://www.scielo.br/scielo.php?script=sci_arttext&pid=SO104-
1 169200400070001 1.

Ozechowski, T. J., & Liddle, H. A., (2000). Family-based therapy for adolescent drug
abuse: Knowns and unknowns. Clinical Child and Family Psychology Review,
3(4), 269-298.

Patterson, G. R., & Stouthamer-Loeber, M. (1984). The correlation of family
management practices and delinquency. Child Development, 55, 1299-1307.

Petronis, K., R., Anthony, J ., C. (2003). A different kind of context: The contagion effect
and cocaine incidence in the US. Journal of Epidemiology and Community
Health, 57(11), 893-900.

Pirie, P. L., Murria, D. M., & Luepker, R. V. (1988). Smoking prealence in a cohort of
adolescents, including absentees, dropouts, and transfers. American Journal of
Public Health, 78(2), 176-178.

Prescott, C. A., & Kendler, K. S. (1999). Genetic and environmental contributions to
alcohol abuse and dependence in a population-based sample of male twins.
American Journal of Psychiatry, 156, 34-40,

Raudenbush, Bryk, & Congdon, R. (2004). HLM (version 6.02a) [computer software].
Linconwood, IL: SSI.

Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical Linear Models: Applications and
Data Analysis Methods (2nd ed). Newbury Park, CA: Sage Publications.

Recagno-Puente, .F. (1998). Familia y exclusion social. A VEPSO, 9, 41-62.

Reckase, M. D. (1996). Test construction in the 19903: Recent approaches every
psychologist should know. Psychological Assessment, 8, 354-359.

Resnick, M. D., Bearman, P. S., Blum, R. W., Bauman, K. E., Harris, K. M., Jones, J ., et
al. (1997). Protecting adolescents from harm. Journal of the American Medical
Association, 278, 823-832.

Resnicow, K., Soler, R., Ahluwalia, Braithwaite, R., & Butler, J. (2000) Cultural
sensitivity in substance use prevention. Journal of Community Psychology, 28,
271—290.

Sampson, R. J. (1988). Local friendship ties and community attachment in mass society:
A multilevel systemic model. American Sociological Review, 53, 766—779.

Sampson, R. J ., & Groves, W. B. (1989). Community structure and crime: Testing social-
disorganization theory. American Journal of Sociology, 94, 774—802.

Sampson, R. J., Raudenbush, S., & Earls, F. (1997). Neighborhoods and violent crime: A
multilevel study of collective efﬁcacy. Science, 227, 918—924.

Samuolis, J., Hogue, A., Dauber, S., Liddle, H. A. (2005). Autonomy and relatedness in
inner-city families of substance abusing adolescents. Journal of Child &
Adolescent Substance Abuse, 15(2), 53-86.

Sartor, C. E., Lynskey, M. T., Heath, A. C., Jacob, T. True, W. (2007). The role of
childhood risk factors in initiation of alcohol use and progression to alcohol
dependence. Addiction, 102(2), 216-225.

Schmitt, N. (1996). Uses and abuses of coefﬁcient alpha. Psychological Assessment, 8,

Schonert-Reichl, K. A. (1999). Relations of peer acceptance, friendship adjustment, and
social behavior to moral reasoning during early adolescence. Journal of Early
Adolescence, 19, 249-279.

Schuckit, M. A. Smith, T. L. Barnow, S. Preuss, U. Luczak, S., & Radziminski, S.
(2003). Correlates of externalizing symptoms in children from families of
alcoholics and controls. Alcohol Alcoholism, 38, 559-567.

Schwartz, S. H. (1999). Cultural value differences: Some implications for work. Applied
Psychology: An International Review, 48, 23-47.

Schwartz, S. H., (2006). A theory of cultural value orientations: Explication and
applications. Comparative Sociology, 5, 136-182.

Simmons, R. G, Blyth, D. A. (1987). Moving into adolescence: The impact of pubertal
change and school context. New York: A. de Gruyter.

Simons-Morton, B. G., Crump, A. D., Saylor, K. E. and Yu, K. (1999) Student—school
bonding and adolescent problem behavior. Health Education Research, 14, 99-
107.

Solomon, D., Battistich, V., Watson, M., Schaps, E., & Lewis, C. (2000). A six-district
study of educational change: Direct and mediated effects of the child development
project. Social Psychology of Education, 4, 3—5 1.

Spoth, R. L., Guyll, M., & Day, S. X. (2002). Universal family-focused interventions in
alcohol-use disorder prevention: Cost-effectiveness and cost-beneﬁt analyses of
two interventions. Journal of Studies on Alcohol, 63(2), 219—28.

Stanton, M. D., & Todd, T. C., (1982). The family therapy of drug abuse and addiction.
Guilford Press New York.

Steinberg, L., Darling, N. E., & Fletcher, A. C. (1995). Authoritative parenting and
adolescent adjustment: An ecological journey. In P. Moen & G. H. Elder Jr.

(Eds), Examining lives in context: Perspectives on the ecology of human
development (pp. 423—466). Washington, DC: American Psychological
Association.

Steinberg, L., Fletcher, A., & Darling, N. (1994). Parental Monitoring and Peer
Inﬂuences on Adolescent Substance Use. Pediatrics, 93(6), 1060-1064.

Streiner, D. L. (2003). Being inconsistent about consistency: When coefﬁcient alpha does
and doesn’t matter. Journal of Personality Assessment, 80, 217-222.

Substance Abuse and Mental Health Services Administration (2005). Results from the
2004 National Survey on Drug Use and Health: National Findings. Retrieved
November 15, 2006 from Substance Abuse and Mental Health Services

Administration Website:
http://www.oas.sarnhsa.gov/NSDUH/2k4NSDUH/2k4results/2k4results.htm#7.1 .

Sullivan H. (1953). The interpersonal theory of psychiatry. NY: WW Norton & Co.

Swadi, H. (1999). Individual risk factors for adolescent substance use. Drug and Alcohol
Dependence, 55, 209-224.

Szapocznik, J ., & Coatsworth, J. D. (1999). An ecodevelopmental framework for
organizing the inﬂuences on drug abuse: A developmental model of risk and
protection. In M. Glantz & C. R. Hartel (Eds), Drug abuse: Origins and
interventions. (pp. 331-3 66). Washington, DC: American Psychological
Association.

Szapocznik, J ., & Williams, R. A. (2000). Brief strategic family therapy: Twenty-ﬁve
years of interplay among theory, research and practice in adolescent behavior
problems and drug abuse. Clinical Child and Family Psychology Review, 3, 117—
135.

Tapia, M. I., Schwartz, S. J., Prado, G., Lopez, B., & Pantin, H. (2006). Parent-centered
intervention: A practical approach for preventing drug abuse in Hispanic
adolescents. Research on Social Work Practice, 16(2), 146—165.

Tarter, R. and Hegedus, A., 1991. The Drug Use Screening Inventory: Its application in
the evaluation and treatment of alcohol and drug abuse. Alcohol Health and
Research World 15, pp. 65—75.

Tarter, R. E., Kirisci, L., Mezzich, A., Cornelius, J. R., Pajer, K., Vanyukov, M., et al.
(2003) Neurobehavioral disinhibition in childhood predicts early age at onset of
substance use disorder. American Journal of Psychiatry, 160, 1078—1085.

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

235

Turner, W. L., Wieling, E. and Allen, W. D. (2004) Developing culturally effective
family-based research programs: implications for family therapists. Journal of
Marital and Family Therapy, 30, 257—270.

United Nations General Assembly Special Session on the World Drug Problem (June,
1988). Fact Sheet No. 4. Retrieved May 2007 at:
http://www.un.org/ga/ZOspecial/presskit/themes/demand-4.htm

Vega, W. A., & Gil, A. G. (1998). Drug use and ethnicity in early adolescence. New
York: Plenum Press.

Vega, W. A., Zimmerman, R. S., Warheit, G. J ., Apospori, E., & Gil, A. G. (1993). Risk
factors for early adolescent drug use in four ethnic and racial groups. American
Journal of Public Health, 83, 185-189.

Vitaro, F., Tremblay, R. E., Kerr, M., Pagani, L., & Bukowski, W. M. (1997).
Disruptiveness, friends’characteristics, and delinquency in early adolescence: A
test of two competing models of development. Child Development, 68, 676-689.

Von Bertalanffy, L. (1968). General systems theory: Foundations, developments,
applications. New York: Braziller.

Waldman, I. D., & Slutske, W. S. (2000) Antisocial behavior and alcoholism: a
behavioral genetic perspective on comorbidity. Clinical Psychology Review, 20,
255—287.

Wallack, J. M., & Corbett, K. (1990). Illicit drug, tobacco, and alcohol use among youth:
Trends and promising approaches in prevention. In H. Resnik, S. E. Gardner, R.
P. Lorian, & C. E. Marcus (Eds), Youth and drugs: Society’s mixed messages (pp.
5-29) (OSAP Prevention Monograph-6). Rockville, MD: US. Department of
Health and Human Services.

Warheit, G. J., Vega, W. A., Khoury, E. L., Gil, A. G., & Elfenbein, P. R. (1996). A
comparative analysis of cigarette, alcohol, and illicit drug use among an ethnically
diverse sample of young adolescents. Journal of Drug Users, 26, 901-922.

Warren, C. W., Riley, L., Asma, S., Eriksen, M. R, Green, L., Blanton, C., et al. (2000).
Tobacco use by youth: A surveillance report from the Global Youth Tobacco
Survey project. Bulletin of the World Health Organization, 78(7), 868-876.
Retrieved December, 2006 at:
http://whqlibdoc.who.int/bulletin/2000/Number%207/78%287%29868-876.pdf

Watzlawick, P., Beavin-Bavelas, J., & Jackson, D. D. (1967). Pragmatics of human

communication. A study of interactional patterns, pathologies and paradoxes.
New York, London: W.W. Norton & Company.

Watzlawick, P., Weakland, J., & F isch, R. (1974). Change: Principles of problem
formation and problem resolution. New York: Norton.

Wills, T. A., & Cleary, S. D. (1999). Peer and adolescent substance use among 6th—9th
graders: Latent growth analyses of inﬂuence versus selection mechanisms. Health
Psychology, 18, 453—463.

Wills, T. A., Resko, J. A., Ainette, M. G., & Mendoza, D. (2004). Role of parental social
support and peer support in adolescent substance use: A test of mediated effects.
Psychology of Addictive Behaviors, 18, 122—134.

World Health Organization. (1997) Cannabis: A Health Perspective and Research
Agenda. Geneva, World Health Organization, Retrieved November 15, 2006 from
World Health Organization Website:
http://whq1ibdoc.who.int/hq/1997/WHO_MSA_PSA_97.4.pdf

World Health Organization. (2004). Global Status Report on Alcohol. Geneva, World
Health Organization, Retrieved November 15, 2006 from World Health
Organization Website:
http://www.who.int/substance_abuse/publications/alcohol/en/

Wright, D. & Davis, T. R. (2001). Youth Substance Use: state estimates from 1999
National Household Survey on Drug Abuse. Rockville, MD: Substance Abuse and
Mental Health Services Administration, Ofﬁce of Applied Statistics.

Zhou, Q., King, K. M., & Chassin, L. (2006). The roles of familial alcoholism and
adolescent family harmony in young adults’ substance dependence disorders:
Mediated and moderated relations. Journal of Abnormal Psychology, 115(2),
320—33 1.

Zucker R. A., Fitzgerald H. E., & Moses H. D. (1994). Emergence of alcohol problems
and the several alcoholisms: A developmental perspective on etiologic theory and
life course trajectory. In: D., Cicchetti, & D.J. Cohen (Eds). Developmental
psychopathology: Vol. 2: Risk, disorder, and adaptation (pp. 677—711). New
York: Wiley.

to
b)
\l

 

 

   

uljllnljjjijijiilintuit

6 0178