Adolescent sexuality

Material Information

Adolescent sexuality the gap between knowledge and action
Dreisbach, Susan Lynn
Place of Publication:
Denver, Colo.
University of Colorado Denver
Publication Date:
Physical Description:
xvi, 375 leaves : illustrations ; 28 cm

Thesis/Dissertation Information

Doctor of Philosophy
Degree Grantor:
University of Colorado Denver
Degree Divisions:
School of Education and Human Development, CU Denver
Degree Disciplines:
Health and Behavioral Sciences
Committee Chair:
Koester, Stephen
Committee Members:
Brett, John
Bull, Sheana Saylers
Corbett, Kitty K.
Main, Deborah


Subjects / Keywords:
Teenagers -- Sexual behavior -- United States ( lcsh )
Sexual ethics for teenagers -- United States ( lcsh )
Sexually transmitted diseases -- United States ( lcsh )
Risk-taking (Psychology) in adolescence ( lcsh )
Risk-taking (Psychology) in adolescence ( fast )
Sexual ethics for teenagers ( fast )
Sexually transmitted diseases ( fast )
Teenagers -- Sexual behavior ( fast )
United States ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 348-375).
General Note:
Department of Health and Behavioral Sciences
Statement of Responsibility:
by Susan Lynn Dreisbach.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
47058418 ( OCLC )
LD1190.L566 2000d .D73 ( lcc )

Full Text
Susan Lynn Dreisbach
B.S.N., Ohio State University, 1973
M.S.S., University of Colorado at Denver, 1996
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Health and Behavioral Sciences
; a

2000 by Susan Lynn Dreisbach
All rights reserved.

This thesis for the Doctor of Philosophy
degree by
Susan Lynn Dreisbach
has been approved
Stephen Koester
)Z\ )3\oo
Date '

Dreisbach, Susan Lynn (Ph.D. Health and Behavioral Sciences)
Adolescent Sexuality: The Gap between Knowledge and Action
Thesis directed by Associate Professor Stephen Koester
This inquiry uses three complementary qualitative methods to reveal the process by
which a complex web of intrapersonal, interpersonal, and environmental factors
influence the ability and willingness of adolescents to act on what they know about
protecting themselves from sexually transmitted infections including human
immunodeficiency virus infection. The narratives of thirty middle and upper class,
male and female, urban and suburban high school participants provide the basis for a
model showing how they navigate a tortuous path through competing messages
about sexuality and gender roles and integrate that information with personal
experience and personal attributes. The integrated knowledge is central to their
decisions whether or not to have sexual intercourse under specific conditions, and if
so whether or not to use condoms. Their enactment of intended sexual behaviors
varies depending on the social situation of the moment. In that moment the
reciprocal and dynamic forces of sexual desire, self-efficacy, peers, parental
monitoring, social opportunities, and alcohol act in concert to either encourage
unprotected sexual intercourse or encourage safe sex or abstinence. The grounded
Model of Adolescent Sexual Behavior adds context to published quantitative
findings, informs prevention and intervention strategies, and provides a basis for
further qualitative and quantitative research in adolescent sexuality and risk
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.

I dedicate this thesis to my family for their unfaltering love, encouragement, insight,
humor, and patience during the process of creating this work and to those teens who
entrusted me with their thoughts and the stories of their lives.

I would like to acknowledge the Sexuality Fellowship Program of the Social Science
Research Council for their intellectual and financial support throughout this inquiry.
In addition, I would like to acknowledge the guidance, inspiration, and patience
provided by my dissertation committee. I am incredibly grateful for their enthusiasm
in mentoring me throughout this process.

Research Questions...............................................5
Scope of the Inquiry.............................................7
Situatedness of the Researcher...................................9
Situatedness of the Inquiry.....................................10
Structure of the Dissertation...................................11
Historical Context..............................................13
Adolescence in American Society in 1999.........................21
Social and Cultural Context of Adolescent Sexuality........22
Developmental Context of Adolescent Sexuality..............30
The Current State of Adolescent Sexual Behaviors................36
Adolescent Risk Taking.....................................36
Coincidence of Alcohol or Drug Use and Sexual Intercourse..44
Sexual Intercourse.........................................48
Non-coital Adolescent Sexual Activities....................54
Initiation of Sexual Intercourse...........................55
Condom Use and Perceived Susceptibility to HIV.............57

Summary of the Literature
Studying Adolescent Sexuality....................................63
Philosophical and Theoretical Foundations........................64
The Reciprocal Constuctivist Paradigm.......................64
Strategies of Inquiry.......................................66
Organizing Theory and Contributing Constructs.............. 68
Design Overview..................................................75
Systematic Data Collection.......................................81
SDC Participant Recruitment.................................82
SDC Consent and Acknowledgement of Receipt of Incentive....84
Demographics of SDC Participants............................84
SDC Setting.................................................86
SDC Data Collection Strategies..............................87
Handling and Analysis of SDC Data...........................89
Primary and Confirmatory Ethnographic Interviews.................90
Participants and Recruitment................................90
Ethnographic Interview Consent and Receipt of Incentive.....92
Demographics of Primary and Confirmatory Participants.......92
Ethnographic Interview Setting..............................95
Ethnographic Data Collection................................95
Ethnographic Data Handling and Analysis.....................97
Audio Journals...................................................99
Participant Recruitment....................................100
Audio Journal Consent......................................100
Audio Journal Training.....................................101

Audio Journal Data Collection, Handling, and Analysis....101
Data Confidentiality..........................................102
Overview of the Model.........................................103
Cognitive Phase...............................................106
Risk-Benefit Decisional Pathway...............................108
Action Phase..................................................110
Action as Feedback............................................Ill
Risk Means Taking Chances.....................................112
Insight on Perceived Susceptibility to STDs..............119
Rank Order....................................................120
Pile Sorts....................................................121
Summary of Systematic Findings Regarding Risk and Sexual Risk.125
Learning about Sexual Risks...................................127
Traditional Sources of Sexual Knowledge: School..........128
Traditional Sources of Sexual Knowledge: Parents and Siblings.. 134
Traditional Sources of Sexual Knowledge: Religion........142
Traditonal Sources of Sexual Knowledge: Gender Roles.....143
Popular Sources of Sexual Knowledge: Gender Roles........157
Popular Sources of Sexual Knowledge:
The Media, Music, TV, Movies, and Ads....................162
Popular Sources of Sexual Knowledge:
Learning about Sexual Risk from Peers....................164

Learning from Your Own Experience and Observations........169
The Kind of Person You Are.......................................175
The Kind of Person Yoti Are: Maturity.....................176
The Kind of Person You Are: Character.....................179
The Kind of Person You Are: Goals..........................184
Risks Assessment and Conditional Intentions.....................186
Things Happen: The Action Phase..................................201
Peer Influence..................................................203
Personal Social Networks...................................205
Good Friends Take Care of Each Other.......................207
Good Friends Provide Diversional Activities................228
Sexual Peer Pressure.......................................213
Friends Provide Opportunities for Unsafe Sexual Behavior...215
Special Opportunities for Unsafe Sex.......................219
Alcohol: Liquid Courage..........................................223
Sexual Desire....................................................226
Parental Monitoring..............................................230
Self-Efficacy: A Fragile Power...................................238
Implications and Applications...................................244
Sources of Knowledge about Sexual Risks...................246
The Kind of Person You Are as a Filter....................267
Common Threads............................................273
The Risk-Benefit Decisional Pathway.............................274
Defining Sexual Risk......................................275
Getting Through the Decisional Pathway....................277

Conditional Intentions.................................279
Translating Intentions into Action..........................280
Peer Influence.........................................303
Intrapersonal Influences...............................311
Future Research and Dissemination of the Findings...........314
Trustworthiness and Limitations.............................315
Lessons from Systematic Data Collection................302
Lessons from the Ethnographic and Confirmatory Interviews.305
Lessons from Audio Journals............................307
The Model...................................................310
Knowledge about Sexuality, Sexual Risks, and Risks in General..312
Information is Filtered by the Kind of Person You Are.......314
Defining Risk...............................................315
The Risk-Benefit Decisional Pathway.........................315
Translating Intentions into Action..........................316
Interventional Implications.................................318
A. RECRUITMENT..................................................320
B. SYSTEMATIC DATA COLLECTION...................................323
C. ETHNOGRAPHIC INTERVIEWS......................................327

D. CONFIRMATORY INTERVIEWS.....................333
F. AUDIO JOURNALS..............................343
G. THE PARTICIPANTS............................346
REFERENCE LIST......................................348

2.1 Percentage of High School Students Engaging in
Health Risk Behaviors in the Past 1-3 Months......................... 42
2.2 Comparison of Percentage of U.S. and Colorado High School Students
Engaging in Health Risk Behaviors in the Past 1-3 Months in 1995.....42
2.3 Proportion of High School Students Who Have Engaged
in Health Risk Behaviors at Least Once...............................43
2.4 Comparison of Percentage of U.S. and Colorado High School Students
Engaging in Health Risk Behaviors at Least Once in Their Lifetime....44
2.5 NSAM and NSFG Trends in the Proportion of Non-Married
15 to 19 Year-Olds Who Report Having Had Sexual Intercourse..........49
2.6 YRBSS Trends in the Proportion of High School Students
Who Report Having Had Sexual Intercourse.............................50
2.7 Proportion of Higher and Lower Income Females Aged 15-19
Who Report Having Had Sexual Intercourse.............................51
2.8 Trends in the Proportion of Male and Female High School
Seniors Who Report Having Had Sexual Intercourse..................... 53
3.1 Relationship Between Paradigm, Guiding Theories,
Contributing Constructs, Strategies, and Data........................67
3.2 Contribution of Multiple Data Sources to the Data....................78
3.3 SDC Recruitment Process.............................................. 84
3.4 Relationship between Ethnographic Participants and SDC Participants.. 91
3.5 Peer Group Affiliations of All Participants..........................93

4.1 Simplified Model of Adolescent Sexual Behavior.........................104
4.2 Model of Adolescent Sexual Behavior.................................... 105
4.3 Multiple Sources of Adolescent Knowledge of Sexual Risks............... 107
4.4 Personal Attributes that Filter Information and Mediate
Formation of Sexual Behavior Intentions............................... 108
4.5 Adolescent Sexual Behavior Decisional Pathway..........................109
4.6 Action Phase of the Adolescent Sexual Behavior Model....................Ill
5.1 Five Organizing Schemes that Emerged from the Pile Sorts............... 122
5.2 Johnsons Hierarchical Clustering of Risks.............................125
6.1 Multiple Sources of Adolescent Knowledge of Sexuality and Risk.........128
6.2 Family as a Source of Knowledge about Sexuality and Risks..............134
6.3 Contribution of Religion to Knowledge of Sexuality and Risks...........142
6.4 Contribution of Gender Roles to Knowledge of Sexuality and Risks....... 143
6.5 Contribution of Contemporary Gender Roles to Knowledge
of Sexuality and Risks.................................................157
6.6 Contribution of Commercial Media Messages about Sexuality...............162
6.7 Contribution of Peer Advice to Knowledge of Sexuality and Risks........ 165
6.8 Contribution of Personal Experience and Observation to Knowledge....... 169
6.9 Personal Attributes that Contribute to the Kind of Person You are......175
6.10 Attributes of Maturity that Contribute to the kind of Person You Are..176
6.11 Character Attributes that Contribute to the Kind of Person You Are....179

6.12 Goals Contribute to the Kind of Person You Are........................ 184
6.13 Cognitive Process of Risk Assessment and Behavior Planning.............187
6.14 Risk-Benefit Decisional Pathway........................................190
7.1 Translating Intentions into Action......................................202
7.2 Positive and Negative Peer Influence on Behavioral Intentions...........204
7.3 Composite Personal Network Map Showing Four Levels of Peers.............207
7.4 Peers Assert Positive Influence by Providing Diversional Activities....211
7.5 Peers Assert Negative Influence by Hosting Unsupervised Parties.........215
7.6 The Influence of Alcohol on Sexual Behavior.............................223
7.7 The Influence of Sexual Desire on Sexual Behavior.......................226
7.8 Positive and Potentially Negative Influences of Parents on the
Enactment of Adolescent Behavioral Intentions.........................230
7.9 Mediating Influence of Interpersonal Self-Effiacy on the
Enactment of Adolecsent Behavioral Intentions......................... 239
8.1 Simplified Model of Adolescent Sexual Behavior..........................245
8.2 Translating Intentions into Action......................................281

2.1 Proportion of High School Students Who Have
Had Sexual Intercourse by Alcohol and Drug Use.....................45
3.1 Demographics of SDC Participants.................................... 85
3.2 Master List of Risks Used for Pilesorts............................ 88
3.3 Demographics of Primary and Confirmatory Participants.............. 93
3.4 Activities of Primary and Confirmatory Ethnographic Participants... 92
3.5 A Sample of Topics Initiated by Investigator and/or Participants... 96
5.1 Frequency of Risks Listed by SDC Participants...................... 115
5.2 Summary of Rank Order of Risks.....................................121

The young... are full of passion, which excludes fear, and of hope which
inspires confidence. Aristotle, Rhetoric Book II (Quandrel)
Adolescence is a time of transition from childhood into adulthood that
requires young people to reinvent themselves by gaining new knowledge and skills,
trying new roles and opportunities, and creating a sense of self that will allow them
to become independent from their parents. One of the most challenging aspects of
reinventing oneself for teens is exploring and becoming comfortable with their
sexuality, those behaviors and characteristics that define a person as a sexual being.
Although sexuality is an intrinsic aspect of human development that is expressed
from birth through death, physical, psychological, cognitive, and social changes
related to sexuality are particularly pronounced during the teen years.
Consequently, as Anke Ehrhardt states, sexual feelings and behavior .. .may be
intensely experienced by many young people in their teens (Ehrhardt, 1996).
In addition to handling intense sexual feelings, todays teens must bridge a
maturation gap that stems from a disparity in timing between biological, cognitive,
and social maturation and the unique cultural and social demands of this historical
period. One critical factor contributing to this gap is the shift in the average age of
menarche and spermarche over the course of the twentieth century from the middle
teen years to around age twelve for girls and age fourteen for boys (Alan Guttmacher
Institute, 1994). At the same time, more young women as well as young men are
delaying marriage to pursue career interests. As a result, there is a gap of a decade or

more between the time when young people are biologically ready to have sex and the
time when they are married and socially approved to have sex (Alan Guttmacher
Institute, 1994).
Other fast-paced social and cultural changes in western culture during the
second half of the twentieth century have also added to the complexity of the
developmental tasks and the magnitude of the maturation gap confronting
adolescents today. Notably these changes include: the availability of effective
contraception; a trend toward more liberal sexual mores concerning sexual
experience prior to marriage; more freedom and unsupervised leisure time at a
younger age; greater expendable resources; and a pervasive consumerism that uses
sexual innuendo and explicit sexual images to promote products and entertainment.
Although these changes exemplify the economic success and technological advances
of the late 20th century, they may unintentionally expose teens to greater risks to their
health and well being. As a consequence, it appears that teens today must make
frequent, complex and serious decisions about their sexuality, sexual behavior, and
risk-taking behavior as a part of their everyday life.
This study focuses on the risks inherent in adolescent sexual activity,
particularly sexual intercourse. These risks take on added significance in an era when
a sexually transmitted infection such as human immunodeficiency virus (HIV) can
result in a chronically debilitating and often fatal disease, acquired
immunodeficiency syndrome (AIDS). At first glance, it seems that this risk might be
minimal in the teenage population since less than 5,000 American teens aged 13 to
20 have been diagnosed with AIDS (Centers for Disease Control and Prevention,
2000b). While adolescents account for less than 1% of the total AIDS cases
nationally, people between the ages of 20 and 29 account for 20% of new AIDS
cases each year (Centers for Disease Control and Prevention, 1997). Given the long
latency period between initial HIV infection and the development of AIDS, it is
reasonable to assume that large numbers of people diagnosed with AIDS in their 20s

probably were unknowingly infected with HIV as teenagers (Centers for Disease
Control and Prevention, 1997; Wortley & Fleming, 1997). Although the prevalence
of HIV infection among adolescents is largely unknown, recent data suggest that the
epidemic is gaining a foothold in the adolescent population with over 3,000 young
people ages 13-24 newly diagnosed with HIV in 1999 in the 34 areas that report HIV
cases to the Centers for Disease Control and Prevention (these areas do not include
the epicenters of AIDS such as New York, California, and Washington, D.C.)
(Centers for Disease Control and Prevention, 2000b).
The perception that adolescents are at low risk for HIV/AIDS changes
dramatically when you consider that an individuals risk of contracting a disease is
the product of the number of infected people (prevalence of disease) in the pool of
potential sexual partners times the frequency of high-risk behaviors such as having
unprotected sex and having multiple partners. Using this equation, it appears that
adolescents risk of HIV infection is being held precariously in check by the low
prevalence of HIV in the adolescent community, since 42% of sexually active high
school students report not using condoms at last intercourse, and 16% report having
more than four lifetime partners (Kann et al., 2000). This represents a large number
of teens taking sexual risks since according to the nationally-based 1999 Youth Risk
Behavior Surveillance Survey (YRBSS), half (49.9%) of all high school students and
65% of high school seniors report having engaged in sexual intercourse (Kann et al.,
2000). The consequences of unprotected adolescent sex are substantial as evidenced
by the 3 million teen cases of sexually transmitted diseases (STDs) reported annually
and nearly 1 million annual teenage pregnancies (Centers for Disease Control and
Prevention, 1993; Centers for Disease Control and Prevention, 2000a). In addition to
the behavioral risk factors, young women are biologically at higher risk for HIV
infection because just after puberty the cervical cells are more susceptible to
bacterial infections which in turn increases susceptibility to HIV infection (D'Angelo
& DiClemente, 1995). In essence, high rates of unprotected sex, combined with

multiple partners and increased biological susceptibility make adolescents
comparable to a pile of dry tinder, just waiting for a match to start a raging bonfire.
This means that there is a dwindling window of opportunity to develop protective
behaviors before there is a substantial increase in prevalence of the disease in the
adolescent population.
Since the advent of the AIDS epidemic in the United States in 1981, there
have been extensive public and school-based educational efforts aimed at increasing
the knowledge and modifying attitudes related to HIV and STDs as a strategy to
reduce risky teen sexual behaviors. Teens have gotten the message. They know a lot
offacts about HIV/AIDS. But knowledge alone has not been sufficient to change the
sexual behaviors that place teens at risk for contracting a sexually transmitted
infection such as HIV/AIDS ( Boyer et al., 1997; Hingson & Strunin, 1992; Keller et
al., 1991; Kirby et al., 1994; Morton et al., 1996; Rotheram-Borus & Koopman,
1991; Rotheram-Borus et al., 1995; Sikand et al., 1996).
Quantitative research has identified many factors highly associated with risky
sexual behaviors among adolescents. Studies have identified personal variables such
as psychological distress (Rotheram-Borus et al., 1989; Walter et al., 1992), alcohol
use ( Lowry et al., 1994; Stall et al., 1986; Temple et al., 1993), depression and
suicidality (Stiffman et al., 1992), and the way teens think about their vulnerability to
sexual risks (Gerrard et al., 1996). Additional studies have identified interpersonal
factors such as clustered problem behaviors (lessor & lessor, 1977), and the
perceived social norms and sexual behaviors of others in the social network
(DiClemente et al., 1996; Fisher et al., 1992; Walter et al., 1992). What is not
addressed in the literature is how personal, interpersonal, and environmental factors
interact to influence adolescent decisions about sexual behavior and the actual
behavioral performance. The literature also has not described from the adolescent
perspective what is happening in the teenage experience to account for this gap
between knowledge and action.

Using in-depth interviewing and systematic data collection strategies, this
study provides insight from the perspective of 30 middle and upper class urban and
suburban teens into the complex world in which these young people struggle to
integrate facts, beliefs, and attitudes about sexuality and sexual behavior from
multiple traditional and popular sources in order to make decisions about their own
intended sexual behavior. It identifies forces in the lives of more affluent adolescents
that support or interfere with their acting on those intentions. And it introduces a
model that suggests how personal, interpersonal, institutional, community, and
public policy factors interact to influence teens decisions about sexual behaviors and
their subsequent sexual actions.
Research Questions
The idea for this study emerged from a professors experience talking to high
school students as part of an evaluation of an HIV/AIDS prevention program in a
nearby affluent resort community. The evaluation documented that the teen
participants had learned the facts about how HIV is transmitted and how teens can
protect themselves from infection. Yet, in talking to the teens after the program about
their beliefs and experiences surrounding safe sex practices, some young men
revealed that they probably would still have unprotected sex in some situations
(Brett, 1996). What was going on? Why would teens engage in unprotected sex when
they know there is a risk of contracting HIV or another sexually transmitted
infection? And short of following them into the bedroom, how can social scientists
gain an understanding of this apparent disparity between knowledge and behavior?
Three premises arose from this initial interest. The first is that there is a gap
between what teens know about risky sexual behavior and how they act sexually.
The second premise is that contextual factors contribute to the knowledge-action
gap. And the third premise is that established and innovative qualitative research
methods can provide the kind of first hand data needed to understand whats going

on from the teen perspective. These premises provided the foundation for the study
and for the two specific aims of the inquiry.
Specific Aim 1: To describe the process by which multiple contextual
factors in the lives of adolescents interact to influence their willingness
and ability (or unwillingness and inability) to act on what they know
about protecting themselves from HIV and other sexually transmitted
Specific Aim 2: To explore the feasibility of using innovative and
traditional qualitative methods of data collection to reveal male and
female adolescent perceptions of sexuality and the contextual factors that
influence teen sexual behavior.
As the investigation evolved, secondary research questions arose. It became
apparent from data collection and the literature that risky adolescent sexual behaviors
were part of adolescent risk taking in general. Knowing that, the question at hand
became how teens define what is risky in general, what they see as risky about sex
specifically, and how they relate the two.
Data from the first round of interviews soon challenged the existence of a
knowledge-action gap. It became conceivable that instead of there being a gap
between what teens know and how they act, that perhaps teens were acting on
knowledge and assumptions that differed from the facts learned in school, from
parents, and from public health messages. This possibility guided the inquiry to look
at the multiple and often inconsistent pieces of knowledge upon which teens base
their intentions about sexual behavior.
Exploration of the origins of teen knowledge about sexual behavior and
sexual norms led to another secondary question about the process by which
knowledge about sexuality and risks translates into behavior. Specifically, how does
knowledge about HIV/AIDS and adolescent sexuality translate into safe or unsafe
sexual behavior, and what contextual factors influence this process?
These secondary research questions, having emerged from the data as it was
collected, elaborated the specific aims and provided additional guidance for the

inquiry. Their influence is evident in the resulting model and in the interpretation of
the data.
Scone of the Inquiry
This study uses qualitative methods of data collection and analysis to focus
on adolescent sexuality, risky teen sexual behavior, risky behavior in general, as well
as those things in the lives of middle and upper class urban and suburban high school
students that influence their ability and willingness to translate knowledge of HIV
and STD prevention into intentions and subsequent actions. Although unprotected
sexual intercourse can lead to an unplanned pregnancy as well as to a sexually
transmitted infection, there are factors specific to unplanned teen pregnancy that are
not be examined here. Only factors that are common to both phenomena are
considered in this study. Because risky behaviors tend to occur in clusters (Jessor &
lessor, 1977) this study also touches on drugs, alcohol, and parties the risk
behaviors that the participants identified as most frequently occurring with risky teen
sexual behavior. This inquiry does not address youth violence, the risk behavior most
prevalent in the news currently, or other risky adolescent behaviors, such as driving
and suicide, that are not directly related to sexual behaviors. However, it is important
to recognize that factors that buffer sexual risks are likely to have a positive impact
on other youth risk behaviors as well (Brener & Collins, 1998; Dryfoos, 1990).
Public discourse about the private behavior of sex has been restricted by
social mores until quite recently (D'Emilio & Freedman, 1997). Even today, at the
beginning of the twenty-first century, the popular language used to describe sexual
behaviors is laden with slang, innuendo, and ambiguity. For example, many
participants in this study talked about somebody getting on another person. This
term could mean kissing, heavy petting, or having intercourse. In contrast, in the

study of sexual behavior, terms that describe human sexuality have more specific
For the purposes of this study, sex refers both to the biological designation of
male or female and to the activity of sexual intercourse. Sex and having sex refer to
vaginal sexual intercourse unless specified as anal or oral intercourse. However,
being sexually active has a broader, more vague definition, meaning that one is
engaged in sexual activities that serve as active outlets for sexual expression. These
activities include sexual intercourse as well as noncoital activities, such as individual
or mutual masturbation.
In contrast, the term gender or more accurately gender role, refers to a
socially and culturally constructed masculine or feminine role with culturally
prescribed behavior that indicates to others and the self the degree to which one is
masculine or feminine. It includes, but is not restricted to sexual arousal and
response. Gender role is the public expression of gender identity, the private self-
awareness of being male or female. In contrast, sexual desire is a personal emotional
and biological response to sexual arousal that is moderated by social norms in
addition to hormonal stimulation. The domain of sexual desire overlaps both gender
identity and gender role.
Sexuality is a comprehensive term that integrates biological, social, and
cultural components of being sexual. It is an inherent human attribute that exists
from birth to death with different forms of expression throughout the lifespan.
Sexuality involves thinking about sex, taking on gender roles, expressing
physiological sexual characteristics, and engaging in physical as well as emotional
sexual activities.
Although sexual terms may be contested within specific academic disciplines
(e.g., some transgender and queer theorists would argue that the terms sex and
gender offer false dichotomies (Herdt, 1993)) these are what I have found to be the

most common usage of the terms in sexuality research. As such, these terms are used
accordingly in this study.
Situatedness of the Researcher
It is my belief that all research, whether qualitative or quantitative, reflects
the background, personal attributes, desired outcomes, and biases of the researcher to
some degree, and that it is imperative that these sources of bias be explicitly
revealed. To that end, it is important to note that I was a fellow in the Sexuality
Research Fellowship Program of the Social Science Research Council during the
time of this inquiry. Not only did their funding enable me to conduct this study, but
interactions within the unique network of experts and fellow researchers also molded
my thinking. The fellowship encouraged me to seek out the complexities of
sexuality, to examine the data critically, and to philosophically appreciate sexuality
as a natural expression of humanity that occurs from birth to death.
As the youngest of three girls and the last try for a boy, I was introduced to
traditional and non-traditional gender roles as a child. The traditional values from my
two-parent, white, middle class Protestant family were challenged in my adolescent
years by the womens rights movement and sexual revolution, pushing me toward
more liberal and open attitudes toward sexual expression, sexual rights, and gender
equality. These attitudes have been reinforced by my twenty-some years of marriage
to a very special man who has made every effort to keep our relationship equitable
and intimate. My ability and willingness to pursue a graduate degree is indicative of
my being surrounded by people who value womens voices. My life experiences
have substantiated my belief that biological differences between the sexes give rise
to legitimate variations in gender roles, but that power differentials between men and
women are culturally and socially constructed and should be critiqued and equalized.
Being able to be a stay-at-home mom while raising our two daughters
allowed me to interact as a volunteer with the youth in our community and to

observe the developmental changes and struggles of many young people in different
circumstances. My unabashed enthusiasm and respect for adolescents most certainly
reflects the positive experiences I had shepherding our children, their friends, and
other youth through their teenage years.
My background in nursing and scientific (i.e. positivist) training shows
through in my need to make order of these disorderly phenomena. Perhaps one of the
greatest academic influences on this work is my recent completion of studies in a
transdisciplinary behavioral science graduate program. My exposure to and
appreciation of the multiple ways of examining any human phenomenon is evident
throughout this work. Looking at adolescent sexuality through many lenses has
inspired invaluable insights but simultaneously imposed an intellectual burden of
ever expanding possibilities of analysis.
Some composite of all these experiences, beliefs, and realities contributed to
the generation of the research questions and the sampling methods I selected. The
inescapable truth is that my being a white, affluent, middle-aged married woman
who was a novice to qualitative research colored my daily research activities, my
interactions with the participants, and inevitably their responses to my inquiry. I
hope that this acknowledgement of the role my background and current person have
played in this study will contribute to the understanding of adolescent sexuality
presented in this work.
Situatedness of the Inquiry
Any research project is also influenced by events that occur during the study.
Such influence is particularly pertinent for a study such as this that collects data in
the community over the course of many months. During the nine months of data
collection, three interesting events occurred that influenced participants responses
and provided natural openings for discussions about oral sex, high school sexual
norms, school response to diversity, and school violence. First, President Bill Clinton

revealed that he had received oral sex from a White House intern, and then denied
that oral sex was really sex. This event occupied the print media during the first
two or three months of data collection. Second, at the height of data collection, in
April 1999, 13 students were shot and killed by classmates at a high school in the
Denver metropolitan area. This event generated months of news articles about teens,
their unique culture, and their behavior. Third, was the release in the spring of 1999
of a movie, American Pie, that offered a satirical view of teen sexuality, teen social
norms, and sexual behaviors. While these events did not impact any of the teen
participants directly, they probably indirectly influenced the thinking of both the
investigator and the teen participants.
Structure of the Dissertation
This study draws on diverse perspectives and theories from multiple
disciplines to provide insight into how contextual factors in the lives of teens interact
to influence their ability to act on what they know about sexual health risks. In
general, the structure follows the model of adolescent sexual behavior that emerged
from the data. Chapter 1 introduces the research questions and situates both the
researcher and the study. Chapter 2 explores what is already known about adolescent
sexuality and sexual behavior by laying out the current status of adolescent sexual
behavior, a brief history of adolescence, and the recognized developmental, personal,
and environmental factors that influence both the decision whether or not to have
safe sex and the subsequent behavioral performance. Chapter 3 describes the design
of the study, addresses the philosophical and theoretical underpinnings, and details
the methods. Chapter 4 introduces the model of adolescent sexual behavior that
serves as the explanatory model for the study and as the organizing framework for
presenting the results and discussion. The model illustrates the complexity of the
pathway that adolescents must follow to integrate a barrage of information, weigh
that information, determine how they intend to act, and then test those intentions in

the real world of peers, parties, alcohol and drugs, and sexual desire. The stories of
how the teens in this inquiry actually navigated this pathway through the model are
presented in the voice of the adolescents in Chapters 5, 6, and 7. Specifically,
Chapter 5 focuses on all 30 participants perceptions of risk in general and their
perceived susceptibility to the risk of sexually transmitted infections including
infection from the human immunodeficiency virus (HIV). Chapter 6 focuses on the
cognitive half of the behavioral pathway illustrated in the model and reveals what the
teens see as the factors that contribute to the risk-benefit analysis that determines
what behavior is intended or planned mentally. Chapter 7 focuses on the factors that
either support or divert those intentions, thereby resulting in safe or unsafe sexual
behaviors. Chapter 8 addresses the meaning of the teens stories and of the model.
Comparing this qualitatively derived data with findings from the literature, this
chapter demonstrates that listening to the stories of a few teens can fill in some of the
contextual details missing from quantitative studies. Quantitative studies have
provided the what and this study offers insight into the how and why of teen
sexual behavior. From that understanding comes the implications for potential
interventions included in Chapter 8. This chapter also discusses the limitations of the
inquiry. Chapter 9 discusses the strengths and weaknesses of the methods used to
address the research questions. Finally, Chapter 10 provides a summary of the
conclusions discussed in the previous two chapters and highlights questions that
require additional study.


Looking at adolescent sexuality and sexual behavior out of context is like
walking into the theatre late and catching only the final scene of a play. It may not be
unintelligible, but it makes less sense than it would if you had witnessed the entire
performance. The more complex the story line, the less sense the piece makes by
itself. The story line behind the development of sexual roles, sexual decision-
making, and sexual behavior of teens is about as complex as it gets. In order to
understand the whole story, it is first necessary to set the stage to situate adolescent
sexuality at the end of the twentieth century within its complex web of historical,
social, cultural, interpersonal, and intrapersonal contexts. Like reading a playbill, this
provides the context that gives meaning to the action of the characters.
There is already a lot known about adolescents and the contextual pieces that
contribute to sexual development. There is a limitation to this knowledge, however.
For the most part, it does not show how these pieces fit together to construct the
sexual reality of individuals and subgroups of adolescents. Gaining an appreciation,
though, of what we already know does set the stage for a more meaningful viewing
of the whole picture, the story of sexuality that the teens themselves share as the
curtain rises in the subsequent chapters of this inquiry.
Historical Context
The social norms and expectations of sexual behavior for adolescents vary
among cultures and change throughout time according to broadly held social values
and the dominant political economy of the times. Standards of sexual behavior are
not only culturally and historically dependent but may be differentially applied to

various groups depending on their age, marital status, and social status. For example,
sexual behaviors that are considered socially acceptable for middle-aged white
married couples are not usually acceptable for single adolescents in American
In order to appreciate the relationship between American society and the
sexual behavior of middle and upper class youth at the end of the twentieth century,
it is helpful to understand how that relationship has evolved during recent history.
Since the adolescents participating in this inquiry are in the contemporary American
mainstream, the following history of adolescent sexuality focuses on youth who
represent the dominant class and culture of the times. As such, the stories of many
young people, the ethnic minorities, the poor, the educationally disadvantaged, the
runaways, the prostitutes, and incarcerated youth are not included.
In European cultures prior to industrialization, children contributed
substantially to the economics of the agrarian family. Children learned at a young
age by example what function they were to perform in life. They learned about sex
primarily from observing their parents in close living quarters. Fathers restricted
marriage options to control the inheritance of wealth and the general economic future
of their children. This gave economic value to the virginity of daughters and the
virility of sons and often prolonged the period between childhood dependence on the
family and adult independence (Modell & Goodman, 1990). During this time, the
term youth could apply to anyone aged seven to thirty and the term boy referred
simultaneously to male children, unmarried males, and male servants of almost any
age (Kett, 1977). Adolescent sexuality was generally ignored.
Colonial American social patterns initially duplicated the agrarian European
model. With time, parental authority diminished in colonial America as young
people encountered new choices and opportunities for economic advancement and
reliance on inheritance decreased. However, economic independence still determined
the point at which a young man was considered a freeman an independent, frill

member of the community (Luker, 1996). Accordingly, sexual independence was
reserved only for freemen. As youth took on the responsibility for making decisions
about where they would live, who they would marry, and what they would do, new
roles and expectations developed, placing a strain on colonial American social norms
and expectations concerning youth and sexual behavior (Modell & Goodman, 1990).
Through the eighteenth century, the church and courts remained the primary
sources of sexual standards for youth, encouraging suppression of lust and punishing
those who engaged in sex outside of marriage, especially young women. Literature
available in America at the time promoted the concept of sex solely for the purpose
of reproduction, highly discouraged masturbation, and ignored contraception.
By the early nineteenth century, the path to adulthood had become much
more uncertain as industrialization and urbanization broadened economic and social
opportunities for young people. Individual choice and initiative replaced inherited
status and wealth as the characteristics of both economic and social success, opening
a world of possibility for adolescents. New social norms inspired by the emergent
middle-class family shifted economic responsibility to fathers, moral responsibility
to mothers, and transformed children from a useful economic commodity to a
priceless sentimental commodity. Middle-class values encouraged self-control and
self-direction. Familial influence on adolescents decreased as educational institutions
and school-affiliated peer groups gained influence (DEmilio & Freedman, 1997).
Social and economic opportunities in the cities in the middle of the
nineteenth century supported the development of an age-based working-class youth
culture exemplified by the Bowery Bhoys and Ghals of New York City. Their
public expression of sexuality and independence through clothing, voice, and
behavior challenged dominant conservative values. At the same time, the burgeoning
world of commerce embraced sexuality as a desirable and saleable commodity
despite efforts by the middle class to retain sexual privacy and protect their youth
from more liberal urban behaviors. Literature detailing the dangers of masturbation

proliferated as did nutritional interventions to curb masturbation such as Grahams
crackers and Kelloggs cereals (Potts & Short, 1999). Young men were expected to
control their sexual urges while nurturing the aggressiveness necessary for social and
economic success. In contrast, young women were expected to deny the existence of
sexual desire and develop a role of self-denial as part of the new economic order.
The public became increasingly concerned about adolescent vices such as young
men and women showing affection in public. Religious conversion became the
agreed upon method for protecting youth from the temptations of sex. The Womens
Christian Temperance Union and Young Mens Christian Association assumed
responsibility for distributing pamphlets designed to terrorize adolescents into
avoiding the evils of sexual activity. And other new institutions such as the scouts,
settlement houses, and high schools arose to socialize youth as an effort to assist
families in their child rearing (D'Emilio & Freedman, 1997).
By the beginning of the twentieth century, a large number of white teens
from both working and middle classes attended high school at some point. The high
school created a socially sanctioned adolescent peer group and popularized
adolescence as a creative and progressive life stage (Ueda, 1987). At the same
time, G. Stanley Halls 1904 groundbreaking psychological study of young people in
their teen years introduced the term adolescence, established it as a critical
developmental phase with unique sexual characteristics, and brought the study of
adolescent sexuality into public discourse (Hall, 1904). Halls work attempted to
integrate psychological, anthropological, social, historical, and religious perspectives
of adolescent behavior and was the first publication to acknowledge the complex
nature of adolescence and the integral role of sexuality (Downs & Scarborough
Hillje, 1993). During this time, the teens years were seen as a time of upheaval,
storm and stress, a time when youths were vulnerable, needed support, and
particularly needed guidance to manage their sexuality. This novel idea of people
becoming more advanced over time in response to internal or environmental

challenges reflected a popular misinterpretation of Darwins relatively recent theory
of mankinds evolution and recent advances in theories of human development
(Modell & Goodman, 1990; Potts &Short, 1999).
The concept of adolescence as a middle-class creation persisted well into the
twentieth century. Adolescent sexuality continued to be seen as something to be
controlled and suppressed. At the turn of the century, more youth were in the work
force than in school, marriage did not occur until the mid-twenties, and peer groups
were based primarily in the neighborhood (Modell & Goodman, 1990). But within a
decade, teens increasingly were engaged in a school setting, spent their time with a
school-based heterogeneous peer group and delayed their entry into the workforce ..
until they had received more education. The rigid rules of courtship were slowly
replaced by a social system of dating that gave young people more control over their
selection of a marriage partner. At the same time, the age of marriage was steadily
declining from the mid-twenties to the early twenties.
By the 1920s, social mores, particularly adolescent sexual mores, had become
liberalized, promoted by the new freedom of adolescents, the increasingly
heterosocial teen world, the advent of available contraceptives, and a commercial
interest in the erotic. Sexual liberalism intellectually separated sexual activity from
reproduction. As a result, a broader range of sexual behavior became accepted as the
social norm. Sexuality education in schools was initiated (Peterson, 1999). And sex
was used to sell everything from cleaning products to cars.
However, the Great Depression of the 1930s curtailed much of this
liberalism, at least on the surface, as economic concerns took precedence and more
conservative values re-emerged. Nevertheless, young people continued to express
their newfound sexual freedom. While the economy was still depressed, Margaret
Sanger used the economic crisis to promote legislation for birth control. Her efforts
increased the availability and accessibility of contraceptives by the middle of the
decade, at least for adults (Gordon, 1990).

Following the economic hardship and personal losses of World War II, young
people were motivated to marry young and settle down with a family. Over two-
thirds of all 19-year-olds were married, many after conceiving a child (Furstenberg,
1990). A resurgence of family values and moral purity reigned in post-war America
as the baby boom surged and record numbers of babies were bom (Downs &
Scarborough Hillje, 1993; Modell & Goodman, 1990).
During this post-war Father Knows Best period, research into adolescent
development and human sexuality became popular. More rigorous, reliable, and
valid studies such as those emanating from the Kinsey Institute described sexual
behaviors of adolescents as well adults (Kinsey et al., 1948; Kinsey et al., 1953).
Another influential study was Colemans 1950s investigation of teen culture that
characterized youth as uniformly irresponsible, hedonistic, and opposed to adult
culture (Coleman, 1961). This image of adolescents as an antagonistic subculture
permeated mid-century American culture despite the failure of anyone to replicate
Colemans findings. Colemans research, however, stimulated many studies, some of
which revealed a broader range of adolescent social norms than detected by Coleman
and found consistently that youth actually reported a strong congruence with parental
political, religious, and moral values. Peer influence was acknowledged as playing
an important role in teen behavior, but not to the exclusion of parental influence or
personal decisions, and premarital sex was condoned for those who were committed
to an impending marriage (Modell & Goodman, 1990). Just over one quarter of
women under age 18 were sexually experienced in the late 50s (Alan Guttmacher
Institute, 1994) and many adolescent men were introduced to sex through prostitutes
(Kinsey et al., 1948).
In the late 1960s, behavioral expectations, social attitudes, and political
policies became more liberalized following the Civil Rights and Womens Rights
Movements. Out of this period of expanding rights and womens lib came the
sexual revolution, a period of greater sexual freedom in and outside of marriage.

The political shift toward greater individual autonomy and the introduction of oral
and intra-uterine contraceptives gave women the liberty to have sex outside of a
committed relationship without fear of pregnancy. Gender roles and sexual behavior
were re-examined, and women took responsibility for their own sexual decisions and
consequences (D'Emilio & Freedman, 1997). Simultaneously, the age of marriage
increased dramatically to the mid-twenties as growing numbers of young women
attended college and pursued careers prior to settling down (Cherlin, 1992; Schwartz
& Rutter, 1998). Issues of adolescent sexuality became more complex and more
visible as contraceptives became available to youth. In 1973, the Roe v. Wade
Supreme Court decision that liberalized abortion laws gave teens greater options for
dealing with unintended pregnancies (D'Emilio & Freedman, 1997). Simultaneously,
sex became an everyday commercial commodity not isolated to the thriving
pornography industry but incorporated into the everyday entertainment of television,
movies, novels, and popular music sexual reality (Vance, 1984). Social acceptance of
premarital sex spread beyond those engaged to be married and increased among high
school students (D'Emilio & Freedman, 1997). By the early 1970s, 55% of 18 year-
old males and 35% of 18 year-old females reported having experienced sexual
intercourse (Alan Guttmacher Institute, 1994).
By the 1980s, public sentiment was more open to sex outside of marriage
with 40% of the public agreeing that premarital sex was not wrong at all according
to a University of Chicago survey (Wood, 1990). More teens and younger teens were
having sex. The nation became concerned about teenage sexual activity in response
to high adolescent pregnancy and abortion rates. Increased numbers of teenagers
participated in risk behaviors such as drinking, smoking tobacco, substance use, and
violence (Centers for Disease Control and Prevention, 1993). And HIV, a potentially
lethal sexually transmitted virus, was identified as the pathogen responsible for the
AIDS epidemic that initially seemed confined to the homosexual community, but

invaded the adolescent heterosexual community by the 1990s (Institute of Medicine,
Even before the threat of HIV/AIDS, education about sexuality had shifted in
the second half of the twentieth century from the family and church to the schools. In
mid-century, the church and parents were assigned the role of moral development
while the schools became responsible for explaining the biological facts of life to
students facing puberty. As adolescent sexual activity became more common,
schools, churches, community organizations, and parents struggled to find new ways
to teach young people the facts about risky behaviors, the potential consequences,
and ways to protect themselves without inadvertently condoning such behavior.
Adolescent sexuality was framed in terms of a public health problem. National
surveys documented that 80% to 90% of adults supported sex education in middle
school and high schools that included instruction on contraception and disease .
prevention, as well as abstinence (Louis Harris and Associates 1988; Meckler,
1999; South Carolina Council on Adolescent Pregnancy Prevention, 1997; The Field
Institute, 1999). Nevertheless, controversy raged between proponents of
comprehensive sexuality education and conservative political and religious groups
advocating an abstinence-only curriculum that would encourage abstinence and
avoid discussion of contraception or the use of condoms to protect against sexually
transmitted infections.
At the end of the twentieth century, there is substantial debate at the
legislative and local school board levels over which of these two educational
interventions work best to prevent sexual risk-taking and reduce harm from risky
behaviors (Donovan, 1998). To date, there is no empirical evidence published that
demonstrates that abstinence-only education delays the onset of sexual intercourse or
reduces unprotected sex beyond a 3 month period (Kirby, 1997; Thomas, 2000). To
the contrary, the evidence shows that the programs most effective in delaying
initiation of sexual intercourse and promoting long-term safe sex are those that are

theory-based (Kim et al., 1997), address abstinence along with pregnancy and
disease prevention, and develop negotiation and other practical skills (Chief State
School Officers Council, 1999; Grunseit & et al., 1997; Kirby, 1999; Thomas, 2000).
Despite the scientific evidence, U.S. federal and state policymakers have
embraced a new effort, as part of the welfare reform legislation of 1996. This effort
created a five-year, $250 million entitlement to support state education programs that
promote abstinence outside of marriage as their exclusive purpose and are separate
from state programs that address contraception and safe sex behaviors.
Consequently, 70% of school districts nationwide currently require that abstinence
be taught either as the preferred option or only option for pregnancy and STD
prevention. In comparison, 14% of school districts have a comprehensive sex
education policy that not only promotes abstinence but includes discussion of
contraceptives as an alternative way to avoid pregnancy and STDs (Landry et al.,
It seems that adolescent sexuality is currently viewed as a problem that is
being politically legislated, manipulated by schools, promoted by commercial
interests, and judged by norms of sexual behavior that have changed nearly every
decade in the last century in America. This historical perspective suggests that the
most private behavior of sexuality is indeed connected to the public social world.
Adolescence in American Society in 1999
In the United States at the close of the twentieth century, adolescence is
accepted as a distinct time of life and given protected status by society (Elliot &
Feldman, 1990). However, these protections are ambiguous. Teens are protected
legally from responsibilities for which they are deemed too immature such as the
economic responsibilities of adulthood, yet high school students receive pre-
approved credit cards and are encouraged through advertising to be active
consumers. Laws have been designed to protect adolescents from the full

consequences of breaking the law, yet more and more frequently teens are being
tried and sentenced as adults in American courts (Males, 1996). Similarly, there are
laws in many states prohibiting sexual intercourse with girls who are in their young
teen years (ages and specifics of statutory rape laws vary among states), yet the
media promotes lipstick, make-up, and miniskirts for 5 year olds (Luker, 1996).
At the same time, American society places other ambiguous expectations on
youth, granting increasing self-determination and freedom at younger ages while
simultaneously legislating school attendance and restricting behaviors such as
driving, employment, voting, drinking alcohol, smoking cigarettes, and joining the
armed forces. Contemporary American values regarding teen sexuality are equally
ambiguous. Sexually provocative images are flaunted in advertising, music,
television, music videos, and movies while discussion of sexuality in schools is
repressed and access to confidential reproductive health services for adolescents is
restricted or denied (Annie E. Casey Foundation, 1998; Katchadourian, 1990). It is
not surprising that many teens are unsure of what is and is not socially acceptable
sexual behavior.
The resulting contusion is compounded by the wide array of role options
available to teens, and the lack of clear guidelines for selecting among them. This
combination of ambiguous expectations and values plus competing roles adds
additional stress to the process of adolescent development and increases the risk of
youth making costly errors in judgment (Elliot & Feldman, 1990).
Social and Cultural Context of Adolescent Sexuality
American teenagers have more freedom, leisure time, educational
advantages, and economic opportunities than ever before (Fine et al., 1990;
Steinberg, 1990). The majority of teens are enrolled in high school, which means
they spend from 25 to 60 hours a week mingling with other teens in academic and
extracurricular activities. Over half of all white adolescents work in addition to

attending school (Fine et al., 1990). Todays adolescents spend an average of over
six hours a day watching television, playing video games, listening to music, and
surfing the Internet. Although this average number of hours reflects about one in six
(16%) who spend more than ten hours a day using media, another 17% spend more
than five hours a day watching TV, and nearly two-thirds (64%) spend more than an
hour a day watching television. After TV, music is the medium of choice for most
adolescents, with most young people listening to an hour and a half of music per day.
This means that the majority of American youth are processing multiple
commercially produced messages about sexuality, sexual behavior, and sexual risk
every day, and basing daily and life decisions, at least in part, on this information
(Kaiser Family Foundation, 1999; Keating, 1990).
The impact of the media on adolescent sexuality, gender identity, and sexual
behavior remains controversial. Kilboume (1999) argues that despite the claim by
most people that they are not impacted by ads, the advertising industry would not be
so powerful and lucrative if advertising did not work. She suggests that the
advertising industry strives to form relationships between people and products,
focusing on the reliability of products and the fragility of human relationships. This
relational approach to consumerism makes females particularly vulnerable to the
influence of advertising since the feminine brain is organized in a relational
framework (Blum, 1997; Chodorow, 1978; Kilboume, 1999).
Studies that have tracked sexual content in television, movies, soap operas,
and music videos agree that the number and explicitness of sexual references,
images, and instances of sexual intercourse have increased dramatically in the past
20 years, and even from year to year. They also agree that the media most frequently
projects sex from a recreational rather than procreational orientation with an
emphasis on physical attractiveness (Greenberg et al., 1993; Huston et al., 1998;
Kilboume, 1999; Ward, 1995). The debate in the research community centers not on
the existence of sexual references in the media but on the impact of those sexual

references on the attitudes and behavior of American youth. Theoretically, sexual
images and messages in the media could have immediate and long-term effects on
adolescent sexual attitudes and behavior through the mechanisms of arousal
(Zillman, 1982), learning from observing role models (Bandura, 1977), and
emotional reactions to sexual content (Berkowitz & Donnerstein, 1986).
Experimental studies of the direct effects of the media have shown that exposure to
content portraying pre- or extra-marital acts of recreational sex increases adolescent
acceptance of sex outside of marriage as being normal and not so bad (Bryant &
Rockwell, 1994; Huston et al., 1998).
Other research suggests that teens are active participants in choosing the
media messages they want to hear, that media contributes to adolescents
development of identity, and that the media may be imitating the reality of teens
lives rather than driving that reality (Arnett, 1995). This is consistent with findings
that adolescents seek out content to address curiosity, emulate or reinforce their own
attitudes, and that they avoid content that repulses or frightens them (Greenberg et
al., 1993). At the very least, there seems to be an interaction between teens and the
media that requires greater understanding.
In addition to the media, teens have access to computers and information
technology that provide inexpensive, instantaneous global exchanges of ideas and
information, although only a small percentage (9%) spend more than an hour a day
using the computer for fun (Kaiser Family Foundation, 1999). However, access to
these high tech advantages is selective based on social and economic status, with
middle-class and more affluent whites having greatest access to educational,
economic, and information technology advances (White, 1999). Opportunities such
as these stimulate positive educational, career, and personal goals, while a lack of
access to similar opportunities may discourage high expectations and goals. It is
these goals and high expectations that appear to help protect teens from initiating

sexual intercourse, having unprotected sex, and getting pregnant (Hayes, 1987;
Hoffert, 1987; Luker, 1996; Musick, 1991; Resnick et al., 1997).
Developmentally teens are increasingly drawn to interacting with their peers
and breaking away from their parents and families. It is not surprising that peer
values are, in many cases, similar to those of the teens parents since parents usually
choose the neighborhood and school from which the teen selects peers. However,
there is concern that the naturally developing separation between teens and parents is
being accelerated by parents who are spending less time than ever with their
adolescent children as work demands compete with parenting demands (Furstenberg,
1990; Hersch, 1998). In addition, over 28% of teens live with only one parent (U.S.
Bureau of the Census, 1996), most often the mother, which can reduce a teens
experience with older males in loving and supportive roles (Pipher, 1994). This lack
of a meaningful father-teen relationship has been associated with early initiation of
sexual intercourse (Bearman & Bruckner, 1999b), poor body image, depression, poor
general adjustment (Musick, 1991), and teen pregnancy among adolescent girls
(Hayes, 1987). Adding to the rift between teens and parents, teens bedrooms have
become private havens with stereos, televisions, computers, and phones that displace
opportunities for parents and youth to interact on a regular basis (Hersch, 1998).
Teen activities, including sports and socializing, often edge into the dinner hour,
making family dinners a rarity for many teens. This leaves few times for teens and
parents to talk and share ideas, an activity shown to be helpful in preventing teens
from engaging in risky behaviors (Resnick et al., 1997).
Peer and leisure activities consume a major portion of teens time (Brown &
Theobald, 1999; Fine et al., 1990). Time spent with peers is instrumental in helping
teens develop a more independent sense of self (Steinberg, 1990). Peer relationships
become increasingly significant and can exert both positive and negative influence
on teen behavior, particularly on sexual and other risk-taking behaviors. Although
coercive peer pressure does exist, less coercive peer influence is more common

and exerts its influence through role modeling of behavior, setting social norms, and
structuring opportunities (such as parties or going to movies) for their peers to pursue
or avoid risky activities (Bearman & Bruckner, 1999a).
Gender Roles. One of the most important and most conflicting tasks for
teens is the establishment of gender identity and gender role. This process is central
to adolescent development since gender identity and gender role go hand in hand to
form the foundation for intimate relationships throughout life (Erikson, 1968).
Gender role encompasses all the public and private behavioral expressions of ones
sexuality and gender identity. The term, gender role, introduced by John Money in
1955, includes, but is not restricted to sexuality in the sense of eroticism" (Money,
1955, p. 254).
Todays teens are exposed to a vast array of gender role models that are
portrayed, idealized, and critiqued in the media, at home, and throughout society. In
contrast to earlier times when male and female roles primarily replicated models
observed within the family, teens now have access through the media and through
increased social mobility to gender role models that were more or less invisible
even a few decades ago. At the end of the twentieth century, young women can
observe women as vixens or victims, stay-at-home or working mothers, intentionally
childless married or single professionals, women as single mothers (by design or
divorce), and heterosexuals as well as lesbians as singles, couples or mothers. In
addition, most young women are acquainted with female peers who have assumed a
broad array of sexual roles. Some dress and act provocatively but are sexually
abstinent. Others may be sexually aggressive, sexually abstinent, pregnant,
promiscuous, or trading sex for money or drugs.
Young men are exposed to an equally diverse potpourri of male role models.
No longer is the rugged Marlboro man or the committed father the norm. On any
given day, young men will see some men who are macho and others who are

sensitive; some who are promiscuous and others who committed to a caring
relationship. These men may be heterosexual, bisexual (having sex with men and
women), homosexual, or having sex with men but not identifying as homosexual.
These men may be married, single, or living with a lover. They may be stay-at-home
dads, committed dads, workaholic dads, or absent divorced dads. Like the young
women, their peers take on a variety of gender roles that range from sexually
promiscuous to abstinent, and from macho and dominant to caring and committed to
virginity until marriage. Although society, and families in particular, are eager for
their teens to experiment with career and other adult roles, they are far less tolerant
of any experimentation with gender roles that differ from those of the immediate
family or those that express non-traditional sexual behaviors (Schwartz & Rutter,
In contrast to the more androgynous gender roles that youth can observe in
contemporary American society, traditional dichotomous gender roles are still
pervasive throughout the middle and upper classes and are promoted as normative by
some religions, by some commercial media, and by many parents. These
traditional roles reflect the reproductive differences between males and females.
Accordingly, stereotypical females, as the ones primarily responsible for
childbearing and child rearing, are seen as the ones who should be nurturing and
maternal, seen and not heard, gentle and emotional, and sexually responsive to
male advances while simultaneously preserving their virginity until marriage.
Stereotypical males, on the other hand, as the primary providers and power brokers
are seen as the ones who should be aggressive, competitive, athletic, unemotional,
protective, and the initiators of sexual interactions (Moore & Rosenthal, 1993; Potts
& Short, 1999; Schwartz & Rutter, 1998).
Contemporary attitudes, beliefs, values, and knowledge from multiple
sources throughout society explicitly and implicitly reinforce these gender roles
(Huston et al., 1998). Religious fundamentalists actively promote traditional male

and female gender roles as being in concordance with those set forth in religious
doctrines (Allgeier & Allgeier, 2000; Potts & Short, 1999). Technological advances
in biology and genetics have enabled scientists to identify male and female cognitive
and behavioral differences that can be traced to levels of sex hormones. Evidence of
hormonally-driven behaviors that align with traditional male and female gender roles
is used to support the innate power differential between the sexes as well as the
complimentary nature of differences between the sexes (Blum, 1997; Moir & Jessel,
1991). Developmental theory has indirectly supported traditional gender roles by
claiming that male identity forms around career, mastery of the environment, and a
focus on separateness, all of which reflect the external and thrusting qualities of the
male genitalia. (Erikson, 1968). And even feminist theory has at times inadvertently
supported these roles by proposing that girls develop their gender identity in
relationship with their mother while boys separate from their mothers to establish
their masculine identity and role (Chodorow, 1978).
While many teens assume traditional gender roles congruent with their
religious, family, or community values, others are drawn to less traditional gender
roles that are supported by the feminist, social-constructionist, and postmodern
perspectives. Since the 1970s, feminists and other social scientists have challenged
the traditional dominant male and submissive female gender roles and called for
greater gender role equity in everyday life as well as in sexual interactions (Brown &
Gilligan, 1992; De Beauvoir, 1989; Foucault, 1978; Gagnon & Simon, 1973; Greer,
1971: Jackson, 1996; Millett, 1970; Vance, 1984). These critiques generally view
gender as a product of social interactions that occur in a patriarchal society that
values the experience and interests of men over those of women. In such a
hierarchical society, sexuality serves the interests of men and leads to coercive
heterosexual intercourse (Jackson, 1996). Social narratives and the biomedical model
of sexuality that support male dominance serve as instruments of oppression and
political power to maintain the status of the established elite (Foucault, 1978;

Foucault, 1986; Laumann & Gagnon, 1995; Vance, 1984). From the feminist and
postmodern perspectives, sexuality and desire are more political than behavioral.
Although the critique of biologically, and religious based traditional gender
roles has expanded the menu of potential gender roles, it has done little to clarify the
meaning of those roles for teens. Consequently, at the end of the twentieth century
the development of gender identity and the assumption of a gender role are not
straightforward tasks for teens. Most teens are assigned a gender at birth depending
on the male or female appearance of their genitals and are raised according to the
dominant gender roles of the culture in which they are reared. Although many
American families today encourage greater independence and competitive behavior
in their girls and greater sensitivity in their boys, traditional white middle-class
patterns of gendered behavior are still prevalent (Moir & Jessel, 1991). As children
age, hormonal changes and social experiences affirm or challenge their childhood
gender identity. It is not unusual for adolescents to be perplexed by their physical
attraction to members of the opposite sex or to the same sex. Acting on these
attractions involves experimenting with gender roles. When attractions and behavior
match the gender role that teens have learned to see as normal within their
community, internal and social conflict is minimal. However, when experiences and
experimental roles do not match the community norm, considerable conflict may
arise. Girls may find themselves much more or less sexually aroused than they
perceive they should be. Boys may find themselves confronted with unforeseen
moral dilemmas in deciding how and when to act on their sexual desires. Teens who
find themselves attracted to members of the same sex face even greater
psychological and social challenges as they attempt to define both the degree of
masculinity or femininity that they are comfortable assuming (gender identity) and
their sexual preference for same or opposite sex partners (Katchadourian, 1990;
Masters et al., 1992; Schwartz & Rutter, 1998).

Developmental Context of Adolescent Sexuality
Gaps and Conflicts. Most developmental authorities agree that the core tasks
of adolescence include: developing a positive body image; defining sex roles and
learning about relationships with the opposite sex; moving toward economic and
emotional independence from authority; preparing for future occupational and family
roles; and developing civic competence (Bandura, 1997; Elliot & Feldman, 1990;
Masters et al., 1992; Muus, 1988; Steinberg, 1990). With so many competing tasks it
is common for disparities in timing or objectives to create gaps and conflicts that add
stress to teens lives and complicate sexual behavior. Under the best of
circumstances, biological, social, and cognitive processes interact and put pressure
on each other to promote change. For example, hormonally driven changes in
physical appearance and sexual interest force teens to make substantial social
adjustments for which they may or not be ready, often tricking them into thinking
that looking physically mature makes them ready for sex (Musick, 1991).
The timing of biologically-driven sexual maturation exerts a strong influence
on social development. The average age of menarche and spermarche in the United
States has declined over the past century to around 12.5 and 14 years respectively,
presumably a result of improved nutrition for both sexes and possibly from exposure
to environmental estrogens for young women (Alan Guttmacher Institute, 1994;
Herman Giddens, 1997; Institute of Medicine, 1999; Nelson, 1995). Similarly, the
earliest signs of puberty in girls such as changes in breast tissue and growth of pubic
hair are no longer uncommon in girls as young as age 7 although the norm or median
age for the onset of puberty in girls remains closer to age 10 (Herman Giddens,
Maturation rates that differ from this norm create a developmental gap in
social and sexual interests, especially for early maturing girls and late maturing boys.
Early maturing girls often join an older peer group, and late maturing boys are often
ostracized for their childish stature and lagging athletic skills (Brooks-Gunn, 1990;

Brooks-Gunn & Furstenberg, 1990). Early maturation in either sex is associated with
earlier sexual debut for both biological and social reasons (Phinney et al., 1990;
Resnick et al., 1997).
Another biosocial gap is created by the current social trend to delay marriage
until young people are in their mid-twenties. This delay results in at least a ten-year
gap between the time of biological readiness for sexual activity and full social
acceptance of sexual activities. This creates many questions about the appropriate
timing for the initiation of intimate sexual behavior.
Cognitive gaps occur equally often as a result of teens limited life
experience and newly evolving critical thinking skills (see the following section
Adolescent Cognitive Development). Cognitive gaps create a situation in which
teens potential for competent decision-making may exceed their performance in real
situations (Keating, 1990).
The tension created by these developmental gaps may be magnified or
minimized by psychological and social responses, which vary according to cultural
context, social values, social networks, and individual personalities (Brooks-Gunn,
Development of Sexuality. As adolescents mature biologically, socially, and
cognitively, they move from autoerotic sexual arousal to sociosexual behaviors. This
usually, but not always, begins with petting (also called necking or making out)
which encompasses those sexual intimacies that stop short of sexual intercourse. In
the United States, petting usually follows a prescribed progression from simple
kissing and embracing, to deep kissing, fondling, and mutual masturbation
(Katchadourian, 1990; Miller et al., 1993). The normative pattern appears to be one
in which the level of intimacy and sexual activity increase as dating becomes more
serious and committed, and that males desire sexual intimacy earlier in the
relationship than females (McCabe & Collins, 1984). Interestingly, this pattern varies
somewhat between ethnic groups. There is some indication that white heterosexual

teens adhere to the progressive pattern and linger in the stages that involve fondling
breasts and genitals. In contrast, black heterosexual teens are more likely to engage
in vaginal intercourse than have direct contact with breast and genitals (Furstenberg
etal., 1987; Smith &Udry, 1985).
As adolescents age and gain experience in noncoital sexual behaviors, they
are increasingly likely to progress to having vaginal intercourse (Katchadourian,
1990), with a smaller number of adolescents progressing to heterosexual or
homosexual anal intercourse (Allgiers & Allgiers, 2000; Schuster, et al., 1996). It
appears that for the most part intercourse is mutually consensual, particularly among
older teens whose partners who are within two years of the same age (Darroch, et al.,
However, sexual behavior is not entirely voluntary for all teenagers. In a
national survey conducted by YM magazine and the Kaiser Family Foundation, 47%
of those respondents who had been in an intimate situation said they had experienced
pressure to do something sexual that they did not feel ready to do (Kaiser Family
Foundation, 1998). Younger adolescents are particularly vulnerable to sexual
coercion as evidenced by the 1987 National Survey of Children that shows that 74%
of young women who had intercourse before age 14 (16% of that age cohort) and
60% of those who had sex before age 5 (23% of that age cohort) did so involuntarily
(Moore et al., 1989). That means that nearly half a million young women aged 14
and younger (approximately 460,000) reported having been forced into sexual
intercourse. In comparison, for young women who initiated sex at ages 17 or 18,
both the percentage (3-5%) and actual numbers (approximately 100,000) of those
reporting involuntary intercourse were substantially reduced (Moore et al., 1989). It
is generally agreed that such examples of sexual coercion reflect .culturally dictated
gender roles that support male dominance in sexual relationships (Beal, 1994; Moore
et al., 1989; Potts & Short, 1999).

Cognitive Development. Puberty with its rush of hormones also signals a
time of cognitive growth and development. Cognitive development patterns are as
variable as physical development and may or may not be synchronous with physical
development. Since cognitive maturation is less discernible at a quick glance,
physical appearance may be a misleading marker of a teens ability to handle
sexually challenging social situations (Keating, 1990).
The developmental path of adolescent cognitive processing has been
extensively studied from a neuroscience perspective most often using middle and
upper class white youths or college students as subjects. Consequently, much of the
cognitive development literature reflects the thought processes of the dominant
American majority. Vygotskys (1978) hypothesis that social interaction and external
dialogue are essential to the development of cognition and the awareness of ones
thoughts establishes a theoretical basis for comparing cognitive processes among
adolescents who have been raised in various cultural and social environments. Yet,
the majority of studies investigating cultural influences on focus on outcomes such
as educational achievement and measures of cognition (i.e. test scores) between
ethnic and racial groups (Fraser, 1995) rather than focusing on culturally-based
differences in the development of thought processes. As such, it is not clear whether
there are cultural differences in the development of adolescent cognitive skills or
whether the differences are primarily in the expression and measurement of cognitive
processes (Jencks & Philliips, 1998).
Many assumptions about adolescent cognitive development stem from
Piagets cognitive development theory (1972) that agues that thinking shifts during
adolescence from concrete to abstract thought processes. While numerous studies
have challenged Piagets formal operations theory associated with the development
of abstract thinking (Carey, 1986; Gelman & Baillargeon, 1983), there is general
agreement among experts that cognition does develop in stages. There is also relative
consensus that by early adolescence, thinking is increasingly abstract rather than

concrete, multidimensional rather than focused on a single issue, relative rather than
absolute, and becoming self-reflective and self aware (Keating & Clark, 1980;
Keating, 1990; Peterson, 1988). Recent studies that map brain growth have found
changes consistent with the development of executive functions during adolescence
that are necessary for judgment and self control. These studies affirm that the final
stages of reasoning do develop during adolescence but suggest that they may not be
complete until early adulthood, especially in males (Thompson et al., 2000).
Adolescent Decision-Making. Developmental advances in cognitive skills
enhance teens abilities to effectively weigh risks against benefits and make
behavioral decisions. To better understand the overall decision making process,
Fischhoff (1992) identified five components of the decision-making process based
on his work with adults: 1) generating alternative options, 2) identifying potential
consequences, 3) evaluating the desirability of potential consequences, 4) assessing
the likelihood of those consequences, and 5) combining the information to make a
decision. Fischhoff s decision-making components are consistent with Mann,
Harmoni, and Powers (1989) conclusion that the ability to generate options, to look
at a situation from multiple perspectives, to anticipate consequences, and to evaluate
information sources increases throughout adolescence with particular gains in
cognitive skills around age 12 and again at age 15 or 16. This suggests that
adolescents are cognitively competent to make the daily and often complex decisions
required to protect their own health. Unfortunately, having the capability to make
decisions is not always sufficient to guarantee that those decisions will be enacted.
In opposition to this finding of adolescent cognitive competence, folk
wisdom and several authors have perpetuated the myth that adolescents possess a
unique pattern of thinking that they are invulnerable to harm, making them more
prone to taking risks (Burger & Bums, 1988; Rotherum-Borus & Koopman, 1990;
Whitely & Hem, 1991). However, there is no evidence that demonstrates that

adolescents differ from adults in their perception of vulnerability or susceptibility to
harm (Melton, 1988; Keating, 1990),;with both adolescents and adults tending to
underestimate their susceptibility to harm and disease (Kamler et al., 1987; Kuhlick
& Mahler, 1987; Quadrel et al., 1993).
Keating (1990) and Quadrel, Fischhoff, and Davis (1993) suggest that this
myth of invulnerability probably originated from Elkinds (1967) theoretical concept
of adolescent egocentrism that suggests that two phenomena may occur as
adolescents develop the ability to take on the perspective of another. First, teens may
fail to differentiate others thoughts from their own and imagine themselves as being
as central to others lives as to their own. This phenomenon, called the imaginary
audience, may account for teens tendencies toward self-consciousness and their
assumptions that they are in the desired and dreaded center of attention (Keating,
1990, p. 71). The second phenomenon, the personal fable, stems from adolescents
exaggerating their sense of uniqueness to the extent that it becomes a conviction
that he will not die, that death will happen to others, but not him (Elkind, p. 1013).
However, studies could not confirm any association between egocentrism and
perceptions of invulnerability (Enright et al., 1979; Dolcini et al., 1989). And
Quadrel, Fischhoff, and Davis (1993) concluded that there was no significant
difference between low-risk teens (primarily white middle-class), high-risk teens
(primarily minority, lower SES), or adults (primarily white middle-class) in their
perceptions of vulnerability to controllable or uncontrollable risks.
Although there is no evidence that teens deem themselves more invulnerable
than adults do, there is some evidence that adolescents give greater weight to
potentially less severe proximal consequences than potentially more severe distant
ones. For example, Kegeles, Adler, and Irwin (1988) found that teens intentions to
use condoms related more to condoms ease of use, acceptability by peers, and
facilitation of spontaneous sex than their potential to prevent STDs or pregnancy.
Again, this may or may not be unique to adolescent thinking.

These studies of adolescent perceptions of vulnerability, while instructive,
tell us little about how teens think about susceptibility to sexually transmitted
infections and how their perceptions are influenced by facts, experiences,
observations, and an adolescents character. It is these issues that are addressed by
this inquiry and reported in Chapter 5.
Another facet of the cognitive process involves evaluation of incoming
information. This requires not only the acknowledgement that all information does
not carry the same weight, but also that some information may be misleading or even
false. Besides factual information about sexuality, risks, and consequences acquired
from school or family, teens process about eight hours of commercially produced
messages in an average day (Fine et al., 1990; Keating, 1990). As Keating (1990)
notes, if we wonder why adolescents have difficulty reaching the cognitive levels
we deem important, we should consider what powerful social practices work against
that attainment (page 79).
The Current State of Adolescent Sexual Behaviors
Adolescent Risk Taking
Adolescent risk-taking behaviors have been defined as volitional acts whose
outcomes remain uncertain with the possibility of an identifiable negative health
outcome (Igra & Irwin, 1996). Risk behaviors are a concern because they are a major
cause of injury, morbidity, and mortality among teens and may persist into adulthood
as major contributors to health problems (Office of Disease Prevention and Health
Promotion, 1993; U.S. Preventive Services Task Force, 1989).
The definition, meaning, and importance of risk may vary depending on the
perspective of the person. As such, the literature identifies six popular ways that risk
is defined and categorized (Kendall, 1995). Baseline or involuntary risk is that which
is inherent in living. Baseline risk cannot be eliminated entirely, although society

sometimes acts and speaks as if in denial of this reality. Institutionally structured risk
is a component of daily life and involves risks such as driving, being exposed to
cigarette smoke, and having unprotected sex. Although they do involve individual
choice, these risks appear to be givens since they are so common. Additive risks are
those risks that cluster together in a hierarchical fashion. For example, sexual risk is
often embedded within the larger risk of being in a marginal social position.
Negotiated risks occur between two or more people. For example, a couple may not
consider unprotected sex risky based on their words that they are seronegative. The
risk from their perception is not that of contracting HIV but the risk that one or both
are lying and that their relationship is vulnerable. Time out voluntary risks are those
risks that are justified because they are temporary or just this once. And finally,
eroticized risk refers to taking risks to enhance sexual arousal. Adolescents face risks
from each of these categories and may engage in risks from more than one category
simultaneously, such as drinking and having sex while on spring break.
Explanations of adolescent engagement in sexual risks such as having sexual
intercourse prior to marriage and having unprotected sexual intercourse have a
theoretical basis in biological, social, and cognitive development and the tension that
accompanies disparities in the timing of the three components. Adolescent
development entails multiple core tasks that focus on: identity formation, sexual
awareness, increasing independence, gaining autonomy from the family, greater peer
affiliation, physiological and cognitive maturation, educational attainment, and
developing social and civil competence (Elliot & Feldman, 1990; Muus, 1988). Risk-
taking behaviors are part of this developmental process and serve different functions
at different developmental stages (Igra & Irwin, 1996). As described by lessors
problem behavior theory, smoking, drinking, drug use, reckless driving, and early
sexual activity can be considered meaningful and purposeful rather than perverse.
These risk behaviors serve to increase peer acceptance, establish autonomy from
parents, diminish stress and anxiety, and are seen as a rite of passage into adulthood

(lessor, 1991). Several other authors agree that risk taking is essential to healthy
adolescent development and provides opportunities to develop self-confidence,
autonomy, and initiative. However, they differentiate between developmentally
constructive risk taking behaviors that are experimental and those that are more
engrained and have the potential to jeopardize health, undermine prosocial
development, and disrupt the developmental process (Baumrind, 1987; Irwin &
Millstein, 1986; Turner et al., 1993).
In light of this developmental propensity for risk taking, it is not surprising
that many young people today engage in multiple risk-taking behaviors such as
smoking cigarettes, drinking alcohol, using drugs, and engaging in sexual intercourse
(see Figure 2.1). This clustering of risky behaviors is demonstrated by studies that
show that teens who engage in one risky behavior are more likely to engage in others
(lessor, 1991; lessor & Jessor, 1977; Lindberg et al., 2000; Millstein et al., 1992).
This is a particular concern for older adolescents since the likelihood of engaging in
multiple risk-taking behaviors increases with age, peaking in young adulthood
(Brener & Collins, 1998; Lindberg et al., 2000). This clustering of multiple risk
behaviors is related not only to developmental challenges but also to shared
biological, psychological, and/or environmental antecedents (Igra & Irwin, 1996;
Langer & Tubman, 1997). Consequently, when investigating the context surrounding
adolescent sexual behavior, it is essential to look at a broad spectrum of risk taking.
A great deal of current research on adolescent risk is focused on protective
factors that can prevent or reduce risk-taking behaviors. This approach to youth risk-
taking was initiated by Garmezy (1983) and Rutter (1979) in response to
observations that some youth are resilient and able to achieve their goals despite their
living in a high-risk environment. As it has progressed, this research has used an
ecological approach to explore family, peer, work, school, and community influences
on adolescent responses to risks such as sexual activity, substance use, dropping-out
of school, and juvenile crime. In the past decade, increasing numbers of studies have

Figure 4.2: Model of Adolescent Sexual Behavior
Cognitive Risk Assessment
and Behavior Planning Phase Action Phase
Religious Beliefs Family Values Self-Efficacy

identified patterns of co-occurrence between behaviors. However, as noted in
Chapter 3 in the discussion of ecological approaches to adolescent risk-taking
behaviors, there is little work that has attempted to integrate the complex interactions
between all these factors (DiClemente et al., 1996).
Studies of individual protective factors have looked at cognitive,
psychosocial, and interpersonal factors. Wemer and Smith (1992) found that having
well-developed problem-solving skills enabled youth to control their impulses
despite an unstable or chaotic home life. Other studies demonstrated that a belief in
ones ability to create change (self-efficacy) and having a sense of personal
responsibility for ones actions acted as safeguards for young people facing multiple
risks (Bandura, 1986; Rutter, 1987; Howard & McCabe, 1990). Going along with
this was the protective value of developing effective interpersonal skills (Wemer,
1990; Wemer & Smith, 1992) that contributed to delaying the initiation of sexual
intercourse when taught in a pregnancy prevention (Howard & McCabe, 1990) and
health-risk behavior prevention intervention (Hawkins et al., 1999). In a somewhat
different vein, religious commitment was found to be a strong predictor of adolescent
resilience and avoidance of risky behaviors (Benson, 1993; Hawkins & Fitzgibbon,
1993). Likewise, conservative values have been associated with a delay in sexual
activity and decrease in pregnancy (Blinn-Pike, 1999). Religiosity has been found to
protect teens from drug abuse, delinquency, and unintended pregnancy possibly by
providing young people and their caregivers a sense of coherence and stability
(Wemer & Smith, 1992). This may particularly valuable to young people living in
poverty who may be prone to a fatalistic attitude that there is nothing they can do to
escape the inevitable oppression of poverty (Fullilove et al., 1993; White, 1999).
In general, youth are more likely to engage in risk behaviors if they come
from a family in which there is minimal parental monitoring, that is, knowing where
your children are, who they are with, and what they are doing (Patterson &
Stouthamer-Loeber, 1984). A lack of family monitoring has been significantly

associated with sexual activity and alcohol consumption particularly among younger
teens (Beck et al., 1997; Steinberg, 1986). When inconsistent or overly permissive
parenting is added to a low level of parental monitoring, the effects are additive and
have associated with the highest levels of cocaine use among adolescents (Steinberg,
1991). Having a father present in a non-conflicted relationship decreases the
likelihood of early sexual debut and engaging in health-risk behaviors (Stem &
Smith, 1995). Interestingly, having a close relationship with at least one adult who
loves the young person regardless of their personal frailties can counteract the effects
of other adult relationships that may be dysfunctional or inconsistent (Benson, 1993;
Leffert et al., 1998; Werner & Smith, 1992). Having a stable affectionate adult
relationship in addition to one or more close friends is also associated with resilience
(Wemer, 1990; Wemer & Smith, 1992).
With the amount of time that teens spend in school, it is not surprising that
there is a reciprocal relationships between academic success or failure and risk
taking behaviors (Brooks-Gunn & Furstenburg, 1989; Entwisle, 1990). School
failure is highest among youth who are doing poorly academically in late elementary
school and who engage in increasingly risky behaviors in middle school and high
school (Hawkins & Fitzgibbon, 1993). Being highly connected to school, on the
other hand, is associated with decreased involvement in risk behaviors, and with
increased academic success (Benson, 1993; Resnick et al., 1997).
Part-time after-school jobs and household responsibilities have been
identified as protective factors especially when carried out in the context of helping
the family (Wemer, 1990). However, working more than 20 hours a week in addition
to attending school has been associated with drinking behaviors, possibly by
decreasing parental monitoring, increasing economic resources for purchasing
alcohol, and increasing contact with older youth who may introduce younger teens to
risky behaviors (Fine et al., 1990; Steinberg, 1991).

Research has uncovered some interesting associations between communities
and adolescent risk-taking behaviors. Numerous studies have shown an association
between social and economic deprivation and risky behaviors including violence and
other criminal activities (Fullilove et al., 1993; Fullilove et al., 1990; Masten et al.,
1990; Masten & Coatsworth, 1998; White, 1999). These neighborhoods often lack
cohesiveness which in turn limits communication, and they lack community norms
regarding curfews, drinking, and dating (Benson, 1993; Leffert et al., 1998). In
contrast, communities with greater economic resources are likely to have
infrastructures that can provide supportive and stimulating environments for both
youth and parents (Garmezy, 1991; Werner, 1990) that have been shown to increase
youth attachments to parents and school and generate low-risk community
behavioral norms (Hawkins et al., 1987).
The portrait of the risk-taking side of youth in Figure 2.1 comes from the
Youth Risk Behavior Surveillance Survey (YRBSS), a large national survey of
predominantly high school students. This survey is conducted every other year by the
Centers for Disease Control and Prevention to measure trends in teen risk behaviors
by having students at randomly selected high schools complete a written survey.
According to 1997 YRBSS data, more than one-third of high school students are
heavy drinkers and just over half drink at least once a month. Slightly more young
men than young women report drinking at least monthly (53% versus 48%
respectively). Rates of drinking alcohol have remained stable over the past decade,
but the percentage of young men and women who report currently smoking
cigarettes increased from 28% in 1991 to 36% in 1997. In the past decade, marijuana
use increased the most of any risk behavior with over one-fourth of high schoolers
(26%) reporting smoking marijuana during the past month in 1997 as compared to
15% in 1991. The frequent use of alcohol and rising use of marijuana are important
because as many as 35% of males reported in the 1995 National Survey of

Adolescent Males (NSAM) that either he or his partner was drunk or high at last
intercourse (Sonenstein et al., 1998).
Figure 2.1: Percentage of High School Students Currently Engaging in
Health Risk Behaviors
Source: Youth Risk Behavior Surveillance Survey. Centers for Disease Control and Prevention
Youth risk behaviors in Colorado have mirrored national trends for the most
part except for slightly higher rates of alcohol and marijuana use and lower rates of
sexual activity. For example, 31% of Colorado high school students in 1995 reported
having sexual intercourse within the past 3 months in contrast to 38% nationally
(Colorado Department of Education, 1998; Kann et al., 1999). Figure 2.2 shows a
comparison between Colorado and national risk behavior data for 1995.
Figure 2.2: Comparison of Percentage of li.S. and Colorado High School Students
Currently Engaging in Health Risk Behaviors in 1995
Alcohol Cigarettes Sex Marijuana
Source: 1995 Colorado Youth Risk Behavior Surveillance Survey
1995 U.S. Youth Risk Behavior Surveillance Survey

Compared to the proportion of youth who have engaged in these risky
behaviors in the past one to three months. Figure 2.3 shows that nationally, a
substantially higher percentage of youths have tried these activities at least once
during adolescence. The percentage of high school youths who have had a drink of
alcohol at least once (79%) has remained stable over the past decade. In contrast, the
percentage of high school students who have ever tried marijuana has increased
during that time from 31% in 1991 to 47% in 1997. The majority (70%) of teens has
tried cigarette smoking in their lifetime. This is a slight, but not statistically
significant, decline from the 72% who reported trying smoking in 1990. Similarly,
there is a decline in the percent of students reporting ever having sexual intercourse
from 54% in 1990 to 48% in 1997 and 50% in 1999.
Figure 2.3: Proportion of High School Students Who Have Engaged in
Health Risk Behaviors at Least Once
* Sex
* Cigarettes
' Marijuana
Source: Youth Risk Behavior Surveillance Survey. CDC
As in the case of current risk behaviors, lifetime Colorado youth risk
behaviors approximate national trends. As shown in Figure 2.4, in 1995, the
percentages of Colorado teens who had ever ingested alcohol (83%) or smoked
marijuana (48%) at least once were slightly higher than the national averages (80%
and 42% respectively). In contrast, the percentage of Colorado youth who reported
ever engaging in sexual intercourse was slightly lower (47%) than the national

average of 53% (Youth 97: Youth Risk Behavior Survey 1998; Colorado
Department of Education, 1998).
Figure 2.4: Comparison of Percentage of U.S. and Colorado High School Students
Engaging in Health Risk Behaviors at Least Once in Their Lifetime
Alcohol Sex Cigarettes Marijuana
Source: 1995 Colorado Youth Risk Behavior Surveillance Survey, CDC
1995 U.S.Youth Risk Behavior Surveillance Survey, CDC
Coincidence of Alcohol or Drug Use and Sexual Intercourse
For this inquiry, the coincident use of alcohol and/or drugs at the time of
sexual intercourse is a very pertinent risk-taking behavior associated with initiation
of sexual intercourse, having unprotected intercourse, having multiple sexual
partners, and getting a sexually transmitted infection. The 1999 YRBSS data
indicated that 25% of high school students reported combining these activities. This
would mean that approximately half of the sexually active teens had used drugs or
alcohol prior to or during intercourse at some point. In all racial/ethnic
subpopulations, male students (31%) were significantly more likely than female
students (19%) to have used drugs or alcohol at last sexual intercourse. In addition,
white female students (22%) were significantly more likely than African-American
female students (9%) to have used drugs or alcohol prior to or during sexual
intercourse (Kann et al., 2000). The proportion of students reporting drinking alcohol
or using drugs at last intercourse has remained substantially the same since 1995,

increasing by three percentage points from 1991 when 22% of high school students
reported combining these behaviors (Youth 97: Youth Risk Behavior Survey, 1998).
Studies specifically addressing the coincidence of alcohol or drug use and
sexual intercourse have produced mixed results, but the weight of the evidence
supports an association between increased sexual activity and being under the
influence of drugs or alcohol. A comprehensive review by the National Center on
Addiction and Substance Abuse at Columbia University using 1997 YRBS data
concluded after adjusting for the influence of age, gender, race, and parents
education level, that teens who drink alcohol are seven times more likely to have sex.
Similarly, teens who use drugs are five times more likely to have sex as shown in.
Table 2.1. They also determined that teens who ever drank prior to age 15 are two
times more likely to have initiated sexual intercourse by age 15. In addition, the 16%
of teens who report four or more sex partners are three times more likely to have
reported alcohol use as well (Foster et al., 1999; Kann et al., 1999). Similarly,
Santelli, Brener, Lowry, Bhatt, and Zabin (1998) found a strong association between
alcohol use and having multiple sexual partners that increased as the number of
alcohol-related behaviors such as binge drinking increased.
Table 2.1: Proportion of High School Students Who Have Had
Sexual Intercourse by Alcohol and Drug Use
Freauencv of Substance Use Percentage Ever Having Sex
Never Drank Alcohol 26%
Ever Drank Alcohol 63%
Frequent Drinkers (alcohol >10 days in life) 70%
Never Used Drugs 36%
Ever Used Any Drugs 72%
Heavy Drug User (any drug > 20 times) 81%
Source: National Center on Addiction and Substance Abuse at Columbia University analysis of 1997 YRBSS data

The relationship between the use of intoxicating substances and sex is not
clear and appears to be very complex. Not only are teens who use alcohol and/or
drugs more likely to have sex with more partners, but the more partners a teen has,
the more likely it is that he or she will use alcohol and/or drugs (Foster et al., 1999).
Some teens respond to sexual arousal with drug and alcohol use to reduce inhibitions
(Derman et al., 1998) or to justify having recreational or unprotected sex (Critchlow,
1983). In addition, drug and/or alcohol use (at least in small amounts) may stimulate
sexual arousal as a result of physical responses such as relaxation, sympathetic
nervous system arousal, or stimulation of dopamine release (Altheide & Johnson,
1994). Small amounts of drugs and alcohol may also stimulate sexual arousal
through psychological responses such as expectations that sex under the influence of
alcohol will be more pleasurable or easier to get (Derman et al., 1998). Studies of
date rape complicate the issue further by suggesting that intoxication from drugs or
alcohol may be a trigger for sexual aggression or for conduct that sends misleading
sexual cues (Ferris, 1997; Lang et al., 1975).
Similarly, there are mixed findings about the association between substance
use and condom use. Senf and Price (1994) reviewed three survey studies and
concluded that drinking alcohol did not influence rates of condom use in the small
adolescent populations studied. Doljanac and Zimmerman (1998) conducted
interviews with over 800 mostly African-American 9th graders and did not find an
association between substance use and condom use. Fortenberry and colleagues
(1997) examined coital logs from 82 young women who had attended an STD clinic
and did not find any patterns that associated alcohol or drug use with engaging in
sexual intercourse or using condoms during intercourse.
In contrast, a study based on YRBSS data from large representative samples
of high school students in the early 1990s did find strong associations between
sexual behaviors such as condom use and substance use. Lowry and colleagues
(1994) found that students who reported ever using marijuana, cocaine, or other

illicit drugs had the highest odds ratios for not having used a condom at most recent
intercourse. Students who reported only smoking cigarettes or drinking alcohol had
smaller, but still significant, increases in the likelihood of having had sex or having
multiple sex partners, but they were no more likely than non-substance users to not
have used a condom at last intercourse. However, it is important to note that this
study correlates drinking frequency and unprotected sexual behavior. It does not
provide any evidence that the two behaviors occurred on the same occasion and does
not offer evidence that explains how the two might influence each other.
Three innovative studies offer some additional insights into these
inconsistencies. One potential psychological explanation comes from Derman,
Cooper, and Agoch (1998) who looked at the relationship between alcohol and
condom use from a perspective of expectancies. They found that the association
between non-use of condoms and alcohol consumption held only for those teens who
held high expectancies that alcohol would facilitate sexual behavior by reducing
inhibitions. Another interesting study looked at public policy influences and
documented a consistently higher adolescent incidence of gonorrhea in states with a
younger legal drinking age and lower taxes on alcohol (Harrison & Kassler, 2000).
Studies exploring the association between drug use and STDs determined that selling
sex for drugs explained this association among crack cocaine users (Fullilove et al.,
1990; Fullilove et al., 1993).
In all, it appears that the relationship between risky sexual behaviors and
alcohol and drug use is complex, multi-faceted, and variable depending on the
individual and contextual situation. While many of these associations have been
revealed in the literature, the process by which the social, cultural, situational, and
intrapersonal contextual factors interact remains unclear.

Sexual Intercourse
Adolescent patterns of engaging in sexual intercourse are associated not only
with other risk behaviors, but also with historical trends, gender, race/ethnicity,
economic status, age, and biological maturation. Adolescent sexual behavior has
changed dramatically in the last decades of the twentieth century as demonstrated by
four large national quantitative surveys that monitor teen health-related behavior: the
National Survey of Family Growth (NSFG), the National Survey of Adolescent
Males (NSAM), the Youth Risk Behavior Surveillance Survey (YRBSS), and the
Longitudinal Study of Adolescent Health (Add Health). All four surveys ask teens
about engaging in sexual intercourse in general and none specify whether that is
vaginal, anal, or oral or with members of the same or opposite sex. As such, the
resulting statistics about adolescents engaging in sexual intercourse imply that all
respondents are talking about a single construct, heterosexual vaginal sex. By
collapsing all sexual behavior under one term, variations in teenage sexual behaviors,
such as homosexual behavior, are potentially underestimated.
Overall trends that show slowing or decreasing rates of sexual activity in the
1990s are consistent among the four surveys. However, the estimated rates of
adolescent sex resulting from the YRBSS are significantly and consistently higher
than the rates for a comparable group of youth broken out of the NSAM, NSFG, and
Add Health samples.. The other difference between surveys is an inconsistency in the
pattern of female sexual activity (Santelli et al., 2000).
During the 1980s, the NSAM and NSFG household surveys administered by
interviewers documented that an increasing proportion of all 15-19 year-old men and
women were engaging in sexual intercourse. Rates of sexual activity climbed for
both males and females from an overall rate below 50% in 1975 to 55% by 1988
(Forrest & Singh, 1990; Sonenstein et al., 1989). In the late 1980s and throughout the
1990s, these same surveys documented a statistically significant reversal in this trend

for 15-19 year-old males and a lesser decline for females in the same age group as
indicated in Figure 2.5.
Figure 2.5: NSAM and NSFG Trends in the Proportion of Non-Married
15 to 19 Year-Olds Who Report Having Had Sexual intewonrse
Source: National Survey of Family Growth (female data)
National Survey of Adolescent Males (male data)
With the introduction in the 1990s of the YRBSS, a written survey
administered almost exclusively to high school students, a slightly different picture
emerged as illustrated in Figure 2.6. The trend of declining rates of overall sexual
activity (males and females combined) and the significant decline in male sexual
intercourse through 1997 was consistent with the NSAM and NSFG. However, the
YRBSS did not show the same decline in female sexuality during the early part of
the decade, but showed a slight increase in female sexual activity. The decline in
female sexual activity did not appear in the YRBSS data until the second half of the
decade, and did not mirror the trend in males. Another unique difference in the
YRBSS data is the similarity of male and female rates of sexual activity in 1997
(Kann et al., 1999). Prior to this time, consistently more males than females had
reported having engaged in sexual intercourse (Alan Guttmacher Institute, 1994).
This shift meant that for the first time during the twentieth century, nearly equal

numbers of male and female high school students reported having had sex. However,
the 1997 rates may have been an anomaly since the rate of male sexual activity rose
again in 1999 (Kann et al., 1999).
Figure 2.6: YRBSS Trends in the Proportion of High School Students
Who Report Having Had Sexual Intercourse
50% +-
30% +-
10% +-
1991 1993 1995 1997 1999
Source: U.S.Youth Risk Behavior Surveillance Survey weighted data, CDC
In addition to looking at changes over time in sexual activity among high
school students by gender, additional insights into adolescent sexual behavior can be
gained by examining these trends from the perspective of other key variables such as
race/ethnicity, socioeconomic status, and age. Looking at ethnicity, it is interesting
that the rate of sexual activity declined significantly between 1991 and 1997 for all
ethnic groups of 15 through 17-year-old male YRBSS participants. In comparison,
the decline was significant only for white males in the NS AM, with a lesser non-
significant decline for Hispanic males and no change for black males. Recent
decreases in male sexual activity have been attributed to a national trend throughout
the 1990s toward more conservative sexual attitudes, increased exposure to AIDS
education, and to an increased awareness of male responsibilities associated with
teen pregnancy (Ku et al., 1998; Sonenstein et al., 1998). What these data do not
show is how these attitudes might have impacted adolescent thinking and action to

result in a behavioral change. In addition, it is unclear if this decline in sexual
intercourse will continue since the YRBSS rate of sexual intercourse for high school
males increased again in 1999 (Kann et al., 2000).
For female youth, the leveling off of rates of reported sexual intercourse for
15 through 17-year-old students reflects, in part, a decrease in reported sexual
intercourse among black adolescent females. This pattern of leveling off was
consistent among the studies although non-significant in the NSFG and significant in
the YRBSS (Kann et al., 1999; Santelli et al., 2000). The decrease in sexual
intercourse among black adolescent females was counterbalanced by a steady
increase in sexual activity throughout the 1980s and 1990s among higher income
adolescent non-Latino white females specifically and higher income females
generally (see Figure 2.7). Using NSFG data, the rate of intercourse for higher
income (at least 200% of poverty level) females aged 15 to 19 rose significantly
from 39% in 1982 to 48% in 1995 while the rate for lower income female
adolescents stayed stable at around 55% (Singh & Darroch, 1999). White non-Latino
females account for the majority of these higher income young women, since they
represent the largest population subgroup nationally and are predominantly above the
poverty level (Annie E. Casey Foundation 1998; U.S. Bureau of the Census, 1996).
Figure 2.7: Proportion of Higher and Lower Income 15 to 19 Year-Old
Females Who Report Having Had Sexual Intercourse
Higher Income
H Lower Income
Source: National Survey of Family Growth (adolescent sample)

Although sex is common among teenagers, it is not universal and depends
highly on age. Few (8%) adolescents have initiated sexual intercourse before age 13.
The majority of teens are virgins until nearly age 17, with up to 35% of young men
and young women maintaining their virginity through high school (Kann et al.,
2000). As a rule, sexual interest and involvement in sexual intercourse increases with
age. This coincides with rising levels of estrogen that stimulate breast development
and cyclical menstruation in females and increased levels of testosterone that are
responsible for beard growth, voice change and sperm production in males.
Likewise, a rise in testosterone in both sexes is highly related to initiation of sexual
activity and increased sexual interest (Halpem et al., 1997; Udry et al., 1985; Udry et
al., 1986). Social and cultural factors often act in complicated ways to reinforce
hormonal arousal in young men and restrain sexual desire in young women (Brooks-
Gunn & Furstenburg, 1989; Udry et al., 1985). For example, church attendance
strongly mediates testosterone effects on sexual debut in white females, but not in
black females or males (Halpem et al., 1994; Halpem et al., 1997). Athletics and
other competitive and aggressive activities may stimulate the release of additional
testosterone in young men, enhancing their athletic abilities as well as their libido
(Nelson, 1995). As research in this area has progressed, it is clear that a combination
of biological and socio-cultural factors best account for these age-related variations
in sexual behavior.
The increase in sexual activity that accompanies age is illustrated by the 1997
YRBSS data in Figure 2.8 that show 61% of high school seniors reporting being
sexually experienced as compared to 48% of all high school students (Kann et al.,
1999). Although the overall trend in sexual activity for this oldest group of high
school students mirrors the acceleration in sexual activity during the 1970s and
1980s seen for high school students in all grades, the rate shows more of a leveling
off in the 1990s rather than the decline noted for all high school students combined.
Most interesting, however, is the trend in sexual activity for female high school

seniors. It has continued to climb uninterrupted since the 1970s and by 1999 actually
surpassed the male rate of sexual activity (Alan Guttmacher Institute, 1994; Kann et
al., 2000; Santelli et al., 2000; Warren et al., 1998).
Figure 2.8: Trends in the Proportion of Male and Female High School
Seniors Who Report Having Had Sexual Intercourse
* Source: National Survey of Family Growth and National Survey of Adolescent Men
** Source: Youth Risk Behavior Surveillance Survey, CDC
Age and ethnicity are not only associated with teens having ever engaged in
sexual intercourse but also with the frequency of adolescent sexual intercourse.
Overall, there has been a trend toward less frequent sexual intercourse in the later
1990s as compared to the 1980s. Although teens in general report having sex less
frequently than young adults, 44% of sexually active 15 to 19 year-old women
reported having intercourse at least once a week. At the other end of the spectrum,
23% reported having intercourse once a month or less (Alan Guttmacher Institute,
1994). Patterns of sexual frequency vary across ethnic groups as well. During the
1990s, a decline in the frequency of intercourse has occurred among African
American young women, while the frequency of intercourse among white teens
increased. Despite increases in rate and frequency of sexual activity among higher
income white teens, there is still an overall higher proportion of African American

teens who report being sexually active (73%) than white (48%) or Latino (52%)
youth as of 1997 (Kann et al., 1999; Singh & Darroch, 1999; Warren et al., 1998).
It is unclear what, if any, role survey bias has played in the data and trends
discussed. It has been suggested that young men are more likely to over-report their
sexual activities and that young women are more likely to under-report their sexual
activities in order to comply with the double standard of social norms concerning
the acceptability of sex outside of marriage for the two sexes (Alexander et al., 1993;
Smith, 1988). However, there is no literature that shows that it is any more
acceptable for young women to acknowledge their sexual behavior in the late 1990s
than it was in the early 1990s. For young men, media messages and school-based
abstinence education may have made it more acceptable for males to acknowledge
their virginity, even though there is no evidence that these two factors have altered
adolescent male sexual behavior beyond a 3 month period (Kirby, 1997; Kirby et al.,
1994). Similarly, it is unclear how young people who engage in homosexual
behavior respond to these surveys and how their responses influence the results.
Despite the discrepancies and recent fluctuations in the rate and frequency of
sexual intercourse among high students, these national data show conclusively that -
large numbers of teens especially older teens are having sex before marriage, and
that it is not just one ethnic or socioeconomic group. However, the data provide little
insight into what this means for teens from their perspective and as they make
decisions about their own sexual behavior.
Non-Coital Adolescent Sexual Activities
In contrast to the amount of national data adolescent sexual intercourse, there
is much less research into the noncoital sexual experiences of teens. In this era of
AIDS, teens seem to be expanding their sexual repertoire. Schuster (1996)
documented that many adolescent virgins practice outercourse, a sexually
gratifying form of mutual masturbation that is an alternative to intercourse. In

addition, oral sex is increasingly common, with 50% of males and 41% of females
engaging in cunnilingus and 44% of males and 32% of females engaging in fellatio
(Newcomer &Udry, 1985).
Initiation of Sexual Intercourse
Beyond the statistical trends and demographics, biological, social, cultural,
and situational factors that encourage or discourage sexual activity in the teen years
have been identified as well. As noted, both young men and young women are more
likely to become sexually active when testosterone levels rise in puberty (Halpem et
al., 1997; Udry et al., 1985; Udry, 1988). Adolescents who live with both biological
parents are significantly less likely to be sexually active at a young age (Scott-Jones
& White, 1990; Upchurch et al., 1998). Adolescents are more likely to be sexually
active if they report looking older than their peers, work more than 20 hours a week,
and perceive that they have a high risk of untimely death (Resnick et al., 1997).
Peer influence on the debut of sexual intercourse is somewhat complex.
Some studies have documented that peers have strong influence on sexual and other
risk taking behavior (Costa et al., 1995; Romer et al., 1994). However, not all peers
have equal influence on sexual behavior. Best friends do not necessarily exhibit the
same sexual behavior. Instead, the next level of peers, the small group of close
friends (that is, those people an individual would identify as a friend, usually fewer
than 10 people) are the most influential (Bearman & Bruckner, 1999b). These teens
usually have remarkably similar interests, attitudes, and behaviors. This similarity is
most likely a result both of selecting friends who are similar and of influencing each
others thoughts over time (Brown & Theobald, 1999). The popular crowd in most
schools does not exert much influence on individual sexual behavior either.
However, the larger peer group or clique (those acquaintances with whom a person
socializes, usually a group of about 50 individuals) is influential, often by providing

opportunities to engage in unsupervised substance use and sexual activities
(Bearman & Bruckner, 1999b).
Having friends who do not engage in sexual intercourse, drink alcohol, or use
drugs reduces the risk of initiating sexual intercourse. Interestingly, having a few
high-risk friends (those not attached to school who regularly engage in risk-taking
behaviors such as drinking alcohol, using drugs, smoking cigarettes, and having sex)
does not seem to encourage more risk-taking among low-risk teens unless there is a
large age difference. In contrast, younger adolescents whose friends consist of
primarily older high-risk youth are particularly prone to early sexual debut. Not
surprisingly, being in a relationship that is perceived to be emotionally close is the
most important factor in the timing of first intercourse for a young woman,
suggesting that romantic partners may have greater influence on sexual behavior than
close friends (Bearman & Bruckner, 1999b).
In addition, there are a multitude of environmental (social, cultural, and
situational) factors that are associated with teens delaying the initiation of sexual
intercourse (Goodson et al., 1997). The national Adolescent Health Study (1997)
showed that teens who were highly connected to their school, who had taken a
pledge of abstinence, ascribed a higher level of importance to religion, participated
in activities with parents, and whose parents disapproved of adolescent sexual
activity and contraceptive use, were more likely to delay sexual intercourse (Resnick
et al., 1997). Other studies have determined that higher parental educational
attainment, higher family income (Furstenberg et al., 1987), greater parental
monitoring (Aseltine, 1995; Shaw & Zelnick, 1981), and higher educational and
vocational expectations promote a delay in sexual intercourse among teens (Hoffert,

Condom Use and Perceived Susceptibility to HIV
For those teens who become sexually active, protecting themselves from
pregnancy and STDs becomes a concern. In 1999, HIV/AIDS is increasingly
creeping into the heterosexual and homosexual adolescent population (Centers for
Disease Control and Prevention, 1999). Although adolescents account for only about
1% of the diagnosed AIDS cases in the United States each year, young adults
between the ages of 20 and 30 account for 20% of new AIDS cases annually
(Centers for Disease Control and Prevention, 1999; Keller et al., 1991). Given the
long latency period of 8 to 10 years between initial HIV infection and the
development of AIDS, it is reasonable to assume that large numbers of these young
adults were infected as teens and that there may be many teens who are unaware that
they are currently infected with HIV because they have no symptoms and have not
been tested. It is the latency of HIV combined with the high rates of STDs and
unprotected sex, and increased cervical susceptibility of adolescent women that make
the adolescent population particularly vulnerable to HIV/AIDS.
Despite increased awareness of the protective value of condoms over the
course of the HIV epidemic, prior to the late 1990s, over half of the teens who
engaged in sexual intercourse did not use condoms consistently. In contrast, in 1997,
over half of sexually active teens (57%) reported that they or their partner used a
condom during their last sexual intercourse. This increase in condom use has been
attributed to recent changes in attitudes such as a greater appreciation of partner
preference for condom use, a decreased belief that condom use reduces sexual
pleasure, and a decrease in embarrassment related to condom use (Murphy &
Boggess, 1998; Pleck et al., 1993; Shrier et al., 1999). Although this is a significant
increase from the 46% of teens using condoms at last intercourse in 1991, that still
means that over 6.5 million young people are having unprotected sex, potentially
exposing themselves to sexually transmitted infections including HIV/AIDS (Alan
Guttmacher Institute, 1994; Annie E. Casey Foundation, 1998; Kann et al., 1999).

Condom Use and Perceived Susceptibility to HIV
For those teens who become sexually active, protecting themselves from
pregnancy and STDs becomes a concern. In 1999, HIV/AIDS is increasingly
creeping into the heterosexual and homosexual adolescent population (Centers for
Disease Control and Prevention, 1999). Although adolescents account for only about
1% of the diagnosed AIDS cases in the United States each year, young adults
between the ages of 20 and 30 account for 20% of new AIDS cases annually
(Centers for Disease Control and Prevention, 1999; Keller et al., 1991). Given the
long latency period of 8 to 10 years between initial HIV infection and the
development of AIDS, it is reasonable to assume that large numbers of these young
adults were infected as teens and that there may be many teens who are unaware that
they are currently infected with HIV because they have no symptoms and have not
been tested. It is the latency of HIV combined with the high rates of STDs and
unprotected sex, and increased cervical susceptibility of adolescent women that make
the adolescent population particularly vulnerable to HIV/AIDS.
Despite increased awareness of the protective value of condoms over the
course of the HIV epidemic, prior to the late 1990s, over half of the teens who
engaged in sexual intercourse did not use condoms consistently. In contrast, in 1997,
over half of sexually active teens (57%) reported that they or their partner used a
condom during their last sexual intercourse. This increase in condom use has been
attributed to recent changes in attitudes such as a greater appreciation of partner
preference for condom use, a decreased belief that condom use reduces sexual
pleasure, and a decrease in embarrassment related to condom use (Murphy &
Boggess, 1998; Pleck et al., 1993; Shrier et al., 1999). Although this is a significant
increase from the 46% of teens using condoms at last intercourse in 1991, that still
means that over 6.5 million young people are having unprotected sex, potentially
exposing themselves to sexually transmitted infections including HIV/AIDS (Alan
Guttmacher Institute, 1994; Annie E. Casey Foundation, 1998; Kann et al., 1999).

Many psychological, social, and cultural factors seem to influence condom
use among adolescents. Condom use may vary depending on the attitude of the
partner and the relationship between partners. There is some evidence that condoms
are used least often with steady partners, and are less likely to be used among
consistent oral contraceptive users (Maxwell et al., 1995; Plichta et al., 1992). On the
other hand, consistent condom use is more likely to occur in relationships lasting less
than 3 months and with partners who prefer condoms for contraception (Howard et
al., 1999; Plichta et al., 1992). This is consistent with the finding by Maxwell,
Bastani, and Yan (1995) that the decision to use a condom was most often unilateral,
but the decision not to use condoms was most often a joint decision in teens
attending an sexually transmitted infection clinic. Adding to this picture is the
finding by Manning, Longmore, and Giorrano (2000) that partners who consider
themselves going steady are more likely than those who have just met or are
just friends to use a contraceptive method the first time they meet.
Condom use is seen more frequently in younger teens, with 59% of sexually
active high school freshmen and sophomores reporting recent condom use as
compared to only 52% of seniors (Kann et al., 1999; Shrier et al., 1999). Pleck and
colleagues (1993) suggested that this trend is related to a decreasing perceived
susceptibility to HIV/AIDS that develops with age. Support for this explanation is
inconsistent. Some cross-sectional studies of a general population of adolescents and
young adults have shown a direct correlation between perceived susceptibility to
HIV and condom use (Basen-Enquist, 1992; Hingson et al., 1990). In other cross-
sectional studies of urban, multi-ethnic youth, the association between perceived
susceptibility and condom use was not demonstrated (DiClemente, 1992; Walter &
Vaughan, 1993). In addition to the possibility that different factors impact
populations situated in different cultural and socioeconomic contexts, Murphy and
Boggess (1998) found that adolescent men in general were less concerned about
getting HIV/AIDS in 1995 than in 1988. The researchers attributed this decline in

perceived susceptibility to longer exposure to the disease and the development of
better AIDS treatments. However, if perceived susceptibility to HIV/AIDS is indeed
associated with condom use, this historical decrease in perceived susceptibility
seems inconsistent with the trend toward increasing condom use. Other perplexing
findings about teen condom use include the findings that in some urban minority
populations, condom use is lower among those who have more sexual partners
(DiClemente et al., 1992; Sonenstein et al., 1989) and among those who engage in
more frequent intercourse (DiClemente et al., 1996). The authors of these studies
speculate that these phenomena may be associated with a tendency for these
subgroups of inner city teens to have a fatalistic attitude about their likelihood of
dying at a young age from AIDS or from violence. Such beliefs and attitudes may
also contribute to the differences found between the studies focusing on ethnic urban
youth and those studies focusing on a general adolescent population.
Studies measuring constructs such as social norms and self-efficacy from
Banduras social cognitive theory (Bandura, 1986) and intention from Ajzen and
Fishbeins Theory of Reasoned Action (Joffe, 1993), have added to our
understanding of condom use among teens. Several studies have demonstrated that
condom use is predicted by intentions to use condoms, perceptions about the sexual
risk and protective behaviors of friends, and expressed parental attitudes toward
adolescent sexual intercourse and contraception (Eisen et al., 1992; Jemmott et al.,
1992b; Kasen et al., 1992; Walter et al., 1994).
Other studies have focused on the ability of the individual to communicate
with his or her partner about condoms and are based on the assumption that in
managing sexuality, people have to exercise influence over themselves as well as
over others (Bandura, 1997, p. 180). These studies incorporated communication
skills training into AIDS education programs and showed an increase in intentions
and confidence in using condoms, although they did not measure actual condom-use
behavior over time (Jemmott et al., 1992a; Jemmott et al., 1992b). Two qualitative

studies focusing on power relationships between partners found that cultural norms
and gender roles may prevent some women from insisting on condom use and
prevent some men from seeing a need for condom use (Sobo, 1995; White, 1999).
Currently about 400 schools nationwide offer free condoms to students and
students do indeed take the condoms. However, it is not clear whether teens actually
use condoms more as a result. Studies by Schuster and colleagues (1998) and
Guttmacher and colleagues (1997) showed a statistically significant increase in
condom use resulting from school access, however, their studies lacked comparison
groups and had insufficient sample sizes. A similar evaluative study by Kirby and
colleagues (1999) did include a comparison group and had a larger sample size and
did not find a significant change in condom use associated with school access. A less
rigorous study by Furstenberg and colleagues (1997) also did not detect a significant
change. Although outcome results are mixed, all four studies agreed that having
condoms available in the school setting did not lead to an increase in sexual activity
(Furstenberg et al., 1997; Guttmacher et al., 1997; Kirby et al., 1999; Schuster et al.,
1998) which was a concern voiced by parents in other studies (Raffert & Radosh,
In addition to promoting condom use to prevent the transmission of STDs,
efforts have been made since the 1980s to encourage teens to restrict the number of
sexual partners to decrease the risk of contracting a disease like HIV/AIDS.
Accordingly, fewer teens today are having sex with multiple partners, with 16% of
teens in 1997 reporting four or more lifetime partners as compared to 19% in 1991.
Young men are slightly more likely than young women (18% versus 14%) to report
having had sexual intercourse with four or more partners during their life. Not
surprisingly, the number of sexual partners increases as youth age with 21% of high
school seniors reporting four or more lifetime sexual partners as compared to only
12% of high school freshmen (Kann et al., 1999).

Summary of Literature
In summary the literature shows that adolescents in the United States at the
end of the twentieth century are living in a unique period in which they are
simultaneously granted many freedoms, protected from select adult responsibilities,
and restricted from activities that adults consider a threat to their safety and well
being. They are exposed to commercial messages that suggest that sex before
marriage is socially desirable, messages from parents and church that denounce the
social acceptability of adolescent sexual activity, and mixed educational messages
about the skills required to protect themselves from sexual risks. Broadly held
American values that support dominant male gender roles undermine .equitable
relationships and promote a double standard for sexual conduct.
The literature indicates that adolescents have the cognitive capacity to make
informed choices about their sexual behaviors and that they are as capable as adults
of seeing themselves as vulnerable to sexual risks such as pregnancy and sexually
transmitted infections such as HIY/AIDS. Other studies show that personal
experience, age, and a general decrease in public concern about the risk of
HIV/AIDS may contribute to young men feeling less susceptible to HIV in the later
1990s than in the previous decade.
National studies indicate that large numbers (about half) of high school
students engage in sexual intercourse and that nearly half of those do not use
condoms consistently and that about one quarter have had multiple sexual partners.
These same studies demonstrate that many youth combine risky behaviors such as
drinking alcohol and having sex. Other data focus on the outcomes of unprotected
sex showing that nearly one million young women become pregnant each year and
that over three million teens report contracting a sexually transmitted infection each
The literature identifies multiple factors that contribute to our understanding
of adolescent sexuality, their sexual choices, and sexual behaviors. However, most

published studies are highly reductionists, investigate a limited number of factors at
one time, and employ a narrow theoretical perspective. In addition, statistical
correlations tell us little about the direction or nature of the relationship between
identified factors. Similarly, aggregated statistics provide little insight into individual
or community patterns of sexual decision making and performance.
Although the existing literature hints at the complexity of adolescent
sexuality and sexual behavior, there are three shortcomings that will be addressed by
this inquiry. First, there is an inadequate understanding of the context surrounding
the factors that influence adolescent sexual decision making and performance.
Second, there is little understanding of the process by which teens translate what
they know about protecting themselves from HIV and other STDs into action. And
third, there is no existing model of adolescent sexual decision making and
performance that integrates the adolescent voice and perspective with a
transdisciplinary theoretical framework.

Studying Adolescent Sexuality
From 1890 to 1970, the initial scientific studies of human sexuality focused
primarily on the individual since sexual behavior was believed to be motivated by
the sex drive embedded within the individual mind and body. Authorities viewed
societies and cultures as responding to sexual behavior rather than shaping it. It was
assumed that the essential or underlying nature of sexuality was biological and
remained fundamentally the same across time and cultures. Only in the past two
decades of the twentieth century, driven in part by the challenge to curtail the
transmission of HIV/AIDS, has research viewed sexuality as something constructed
by society and culture within a particular socio-historical context (Foucault, 1978;
Gagnon & Parker, 1995). Reflecting this new perspective, John Gagnon (1977), one
of the pioneers in social constructionist sex research, suggests, people become
sexual in the same way they become everything else. Without much reflection, they
pick up directions from their social environment. They acquire and assemble
meanings, skills and values from the people around them (p. 2).
As a result of the essentialist influence on sexuality research, the majority of
research into adolescent sexual behavior has a strong grounding in psychology and
human development and is focused on individual attributes that contribute to
sexuality and adolescent sexual behavior patterns (see for example Jessor, 1991).
More recent studies have investigated interpersonal factors such as the influence of
peers, parents, and other supportive adults on adolescent sexual behavior (see
Benson, 1993; or Bearman & Bruckner, 1999b). A few studies have linked teen risk
taking behaviors to institutional and community antecedents such as ones attitude

toward and connectedness with school (for example, Hawkins, et al., 1999). Few
studies have examined the contextual nature of adolescent sexuality (White, 1999).
And fewer still, if any, have explored the relationship between the intrapersonal
(including biological), interpersonal, institutional, cultural, social, and political
factors that influence adolescent sexuality and adolescent sexual behavior. It is the
intent of this inquiry to add to the understanding of the context surrounding
adolescent sexuality and to explore the process by which teens translate what they
know about sexual behavior and sexual risks into action by looking at the
relationship between the intrapersonal, interpersonal, institutional, cultural, social
and political factors and contingencies.
Philosophical and Theoretical Foundations
The Reciprocal Constuctivist Paradigm
A constructivist paradigm provides the philosophical umbrella for this
inquiry supplying a basic set of beliefs that guides action (Guba, 1990, p. 17). The
constructivist paradigm proposes that the purpose of research is to understand how
people construct social meaning. It assumes that multiple, apprehensible, flexible,
and potentially conflicting social realities are created through social interactions that
occur within a cultural context. These social realities are dynamic and may vary in
response to new information or changes in the social or cultural environment
(Neuman, 1990; Guba & Lincoln, 1994). The constructivist paradigm is an
interpretive rather than a positivist approach that assumes that people may or may
not experience social reality in the same way. Whereas positivist research values
statistically significant findings that contribute to the discovery of an absolute truth,
constructivist research values common sense and folk explanations that are used in
everyday experience to organize and give meaning to events in the world (Neuman,
1990). Unlike positivist research, this paradigm does not assume that research is

value-free, but explicitly acknowledges the influence of values on the investigators
and participants interpretations of reality (Denzin & Lincoln, 1998). Within the
constructivist paradigm, the investigator serves as a facilitator who manages the data
collection and analysis process and projects her own voice in addition to that of the
participants (Guba & Lincoln, 1994).
The constructivist approach has been used in sexuality research since the
1960s reflecting an underlying assumption that sexuality is constructed in response
to specific historical and social circumstances rather than being based solely on
internal drives. Its application has become particularly widespread since the
inception of the AIDS epidemic in the early 1980s when it became apparent that
preventing the spread of AIDS required a better understanding of the social and
cultural contexts that surround sexual behavior (Gagnon & Parker, 1995). The
paradigm has guided both quantitative and qualitative research, with the majority of
studies being quantitative and focused on isolated social variables. However, Diaz
(1997), Luker (1996), Sobo (1995), and White (1999) and have produced four
notable qualitative studies of sexuality guided by a constructivist paradigm.
Although each of the four studies focused on a different population (gay Latino men,
pregnant urban teens, disadvantaged urban women, and adolescent black women
respectively), each added to the understanding of human sexuality by revealing
complex webs of cultural beliefs and social structures that influence peoples
perceptions of their sexuality, potential sexual risks, and their capacity to protect
themselves from sexual risks.
For this inquiry, the constructivist paradigm has been modified to reflect the
observed reciprocal nature between sexual behavior and the social and cultural
environment. Accordingly, a unique term, reciprocal constructivism, is introduced in
Figure 3.1 to describe a modified paradigm that assumes that sexuality and sexual
behavior are not only socially and culturally constructed, but that they likewise
influence society and culture. The two beliefs, that social reality is created through

social interactions within a cultural context and that there is reciprocal relationship
between the person, his or her behavior, and the environment, provide the
overarching principles that guide this study.
In accordance with the constructivist paradigm, this inquiry is grounded in
the everyday practices of individuals. As such, it was conducted in the community in
collaboration with the teen participants, with every effort made to preserve the teens
voices. Methods of collecting and analyzing data were qualitative and inductive,
meaning that they focused on the meaning of sexuality and sexual behavior and that
the explanatory model emergedfrom the data.
Strategies of Inquiry
Strategies of inquiry were based on ethnography and grounded theory-
generating procedures. Ethnographic strategies used recurrent semi-structured and
open-ended interviews to capture the teen voice and produce a rich description of
the cultural and social contexts that influence adolescent sexuality and safe sexual
behavior (Atkinson & Hammersley, 1994; Hammersley & Atkinson, 1983). Semi-
structured interviewing also produced a picture or map of the social links and
interactions (social networks) of those participants in the ethnographic phase of the
study (Trotter, 1999). Efforts were made to collect data as well by having teens
produce tape recorded audio journals to enhance teen-research collaboration and
gather additional teen explanations of adolescent sexual behavior. In addition,
structured interviews based on Weller and Romneys (1988) systematic data
collection methods were to understand how teens define risk and how they think
about sexual and more general risk taking.
Data collection and analysis followed Strauss and Corbins (1990) grounded
theory-generating procedures that outline an iterative process during which the data
are coded and analyzed for emerging themes as they are collected and those themes
are pursued in subsequent interviews. The result is an empirically grounded model

derived from the everyday reality of the teen participants. The model provides an
understanding of the process by which multiple factors interface to influence
adolescent sexual behavior.
Figure 3.1: Relationship Between Paradigm, Guiding
Theories, Contributing Constructs, Strategies, and Data

Organizing Theory and Contributing Constructs
The model of adolescent sexual behavior that emerged from this inquiry is
grounded in the socially and culturally constructed everyday experience of the
adolescent participants rather than in any pre-existing theory. However, established
social, developmental, and behavioral theories provided an organizational structure
for the inquiry and contributed specific constructs throughout the research process.
The organizing theory, the ecological model (Bronfenbrenner, 1979) described
below, provided a theoretical structure for the initial organization of the complex
web of data. This theoretically-derived model, along with the constructivist
paradigm, influenced the thought process of the researcher during data collection,
analysis, and subsequent interviews. The organizational structure derived from the
ecological model helped integrate the multiple perspectives (biological,
psychological, developmental, cognitive, cultural, social, economic, and political)
encountered in this study in a manner consistent within the constructivist paradigm.
Contributing constructs extracted from established behavioral theory, on the
other hand, were incorporated only following the analysis of the empirical data.
When constructs emerged from the data that exemplified constructs defined in
established behavioral theory, the established label was applied to prevent confusion.
This was the case with the construct, self-efficacy. In instances in which the
empirically derived constructs were either novel or a variant of a previously
described construct, new terminology taken directly from the participants discourse
was applied. An example of this is conditional intentions, those behavioral intentions
that have conditions attached, like I will not have premarital sex unless I am too
drunk to be able to control myself.
The relationship between the paradigm, strategies of inquiry, foundational
theories, and contributing constructs is depicted in Figure 3.1. The figure illustrates
how the reciprocal constructivist paradigm guided the choice of research strategies.
In addition, the paradigm focused the inquiry on adolescent sexual and risk taking

behaviors and their intersection with social structures and culture. The diagram
shows that the paradigm, strategies of inquiry, and the emerging data guided the data
collection process. The resulting psychological, biological, cultural, and social data
were then organized using an ecological model that provided a framework for
incorporating constructs drawn from established behavioral theories. Principles of
developmental theory were used according to their applicability to each category of
the ecological model. Three constructs derived from various health behavior theories
- self-efficacy, behavioral intention, and perceived susceptibility were so prominent
in the data, that they were used as codes and incorporated into the construction of the
emergent model of adolescent sexual behavior. Similarly, the personal social
network as defined in social network theory, so closely matched the social
relationships described by the participants that it was incorporated as a contributing
construct as well.
In the context of this study, ecological model refers to a theoretical model
based on ecological principles that addresses health and behavioral issues from
multiple levels. Such studies focus on the interrelationships between personal,
interpersonal, and impersonal environmental determinants and antecedents of
behavior (Baranowski et al., 1997; Bronfenbrenner, 1979; McLeroy et al., 1988). As
used in the ecological model, the term environment refers to the cultural and social
space beyond the individual as perceived by the individual. Ecological models have
been applied to studies in anthropology, economics, psychology, sociology, and
public health to explore the reciprocal relationships between organisms and their
Interest in the relationship between the environment, behavior, and health is
not new. As early as the mid-1800s, Rudolf Virchow demonstrated that
environmental factors such as poverty and social class that resulted in crowded living
conditions could be related to a disease process such as a typhus epidemic (Moos,
1979). Although environmental factors have been identified as variables in many

health related studies since Virchows time (Amick et al., 1995), it has been only in
the latter half of the twentieth century that interpersonal factors, environmental
factors, and behavior have been integrated into a single theoretical model, the
ecological model (Baranowski et al., 1997).
The ecological model, which is basically a systems model, has been modified
and expanded as it has been applied to new areas of inquiry. Bronfenbrenner (1979)
was instrumental in forwarding the model by defining four levels of influence on
social behavior: 1) the microsystem that encompasses interpersonal roles, activities,
and relationships between individuals and their families, peer group, and work
group; 2) the mesosystem that refers to the interactions between the social settings in
which the individual is involved, such as family, school and church; 3) the
exosystem that includes forces within the individuals larger social system, such as
the unemploy-ment rate or commercial use of sexual innuendo to sell products; and
4) the macro-system that refers to the cultural beliefs and values that influence the
other three systems. Bronfenbrenner also believed that the relationship between these
four systems was reciprocal, meaning that one system could initiate change in the
Bronfenbrenner applied the model to social phenomena in educational and
other institutional settings. His work and carefully defined model stimulated an
interest in applying an ecological framework to public health concerns such as health
behaviors, diseases, and health promotion. Belsky (1980) combined the ecological
model with a theory of individual development to account for individual, family,
social, and cultural influences in child abuse. Winett applied the framework to
healthy life-style choices (Winett, 1985) and other health promotion strategies
(Winett et al., 1989). And McLeroy, Bibeau, Steckler and Glanz (1988) modified
Bronfenbrenners model to study health promotion by systematically assessing five
primary sources of influence on health behavior: intrapersonal factors, interpersonal
processes, institutional factors, community factors, and public policy.

A similar ecological model has been used in recent work by Werner (1990)
and Benson (1993) to identify factors that protect youth from engaging in high risk
activities. As such individual, peer, family, school, and community factors that
reduce the likelihood that an adolescent will participate in a risk-taking behavior
have been identified through numerous studies. Factors from each level of inquiry
have been noted in Chapter 2. However, it is interesting to note that the interactions
between all the factors have not been well documented at this time.
In comparison to the integrative perspective of the ecological model,
constructs derived from behavioral theories focus more narrowly in the intrapersonal
and interpersonal spheres of influence The terms applied to specific constructs were
borrowed from Social Cognitive Theory (self-efficacy), the Theory of Planned
Behavior (behavioral intention), and the Health Belief Model (perceived
susceptibility) because they fit the emerging data so well and had previously been
validated with teen or young adult populations. While they primarily contributed to
the coding and theory/model-building phases of the study, underlying assumptions
associated with these theories and constructs undoubtedly had at least an indirect
influence on the data collection and analysis.
Banduras Social Cognitive Theory (SCT) (1986) emphasizes the reciprocal
relationship between the person, behavior, and the environment, called reciprocal
determinism by Bandura. This concept of reciprocity suggests that behavior is not
just the result of the interactions between the person and environment, but that
behavior also impacts the interaction between the person and environment. Teen
discussion of self-efficacy (a persons confidence in being able to perform a specific
behavior in any situation), a central construct of SCT, emerged from the earliest data
collection phase of this study and repeatedly throughout all phases. Consequently,
that exact term, self-efficacy, was used to code all examples of teen confidence (or
lack of) to perform a specific behavior in a variety of situations and was included in
the emergent behavioral model. Additional SCT constructs that emerged from the

analysis of the data included situation (a persons perception of the environment),
behavioral capacity (knowledge and skill to perform a behavior), and observational
learning (learning by watching others). However, these exact terms were not
incorporated because the examples in the data did not precisely match Banduras
definition of the term.
Social Cognitive Theory and particularly the construct of self-efficacy were
selected to contribute to the inquiry because they fit with what the participants
were saying about having the confidence to act on their intentions. The constructs of
self-efficacy and interactive self-efficacy were the most applicable SCT constructs
because they have been shown to mediate between what people know and how they
behave (Maibach & Flora, 1993). According to Bandura, the belief in personal
efficacy forms the basis for human agency and consequently for self-control
(Bandura, 1997). Perceived self-efficacy is different from self esteem in that efficacy
is concerned with a judgement of capability to perform a behavior while self esteem
is concerned with perceived self worth. As such, it is self-efficacy that is predictive
of behavioral performance, the focus of this inquiry.
Perceived self-efficacy and interactive self-efficacy have been shown to be
associated with behaviors that prevent and reduce risky adolescent sexual activities.
Perceptions of low self-efficacy have been shown to be predictive of adolescent
intentions to become sexually active within the next year (Walter, et al. 1992).
Enhanced confidence in being able to manage sexual situations has been associated
with effective contraceptive use in adolescent women (Kasen et al., 1992; Levinson,
1986). Personal efficacy beliefs have also been identified as the primary factor
determining whether intentions to use contraceptives are put into action (Basen-
Enquist & Parcel, 1992). SCT based interventions that promote self-efficacy,
mastery of skills, and self-regulation of behavior through social modeling, role-
playing, and corrective feedback have increased intentions to abstain from sex or use
condoms (Jemmott et al., 1992a) and decreased unprotected sexual intercourse and

sex with multiple partners (Gilchrist & Schinke, 1985; Kelly, 1989). Not all studies
have shown SCT based interventions to reduce risky sexual behaviors, with
responses varying depending on gender, pre-intervention virginity status, and
overwhelming social factors such as poverty and being involved in a cluster of high
risk activities (Eisen et al., 1992).
Intention is another theoretically based construct that was incorporated into
the data analysis and emergent model. Although Bandura discusses the importance
of intention and attitude toward a behavior in predicting actual behavior, these
constructs are more attributable to the Theory of Reasoned Action (TRA) developed
by Azjen and Fishbein (1980) and revised as the Theory of Planned Behavior (TPB)
(Ajzen, 1991). These two sister theories emphasize attitude toward a behavior,
intention to act, and the role of normative beliefs in mediating between the intention
and action. The theories hypothesize that individuals intend to perform a behavior
when they have a positive attitude toward performing it and when they believe that
important others think that they should perform it (Montano et al., 1997).
TRA and TPB constructs have been applied to studies interested in
identifying attitudes and perceptions of behavioral norms that might be modified
through interventional health messages (Gillmore et al., 1994; Jemmott et al.,
1992a). One such study measured attitudes about condom use with steady partners as
compared to casual partners. The findings revealed that attitudes toward using
condoms (e.g., they are less spontaneous, they make your partner more relaxed) are
more important in predicting condom use with a steady partner, while perceived
control over condom use (e.g., having condoms available, partner suggesting
condom use) is a more likely predictor of condom use with casual partners (Montano
et al., 1997). The constructs of behavioral intention, perceived norms, and decisional
balance are included in modified forms with modified labels in the model of
adolescent sexual behavior that emerged from this study.

One other behavioral construct that has contributed to this inquiry is that of
perceived susceptibility to risk. This construct is drawn from the Health Belief Model
(HBM) developed by Rosenstock (1974) and Becker (1974) which is based on
Skinners (1938) principles of operant learning. The HBM contends that an
individual weighs the risk of becoming infected with HIV or an STD (perceived
susceptibility) plus the perceived severity of the consequences against the perceived
costs and benefits of the preventive behavior. Like self-efficacy, the construct of
perceived susceptibility emerged from the earliest data and consequently was applied
as a code and construct in the emergent model. However, the construct of perceived
susceptibility to risk also contributed to the design of the study, particularly in the
systematic data collection phase at the outset during which the teens defined the
domain of risk.
While perceived susceptibility has been measured frequently in studies of
risky sexual behavior among adolescents, results have been mixed. This may be the
result of variation in the measurement and operationalization of constructs (Strecher
& Rosenstock, 1997) or due to unmeasured contextual factors that varied among the
different study populations. Several cross sectional studies with adolescents found a
significant association between perceived susceptibility and decreases in risky sexual
behavior such as having unprotected sex (Basen-Enquist, 1992; Hays et al., 1990;
Hingson et al., 1990). Other cross-sectional studies of adolescents by Brown,
DiClemente, & Park (1992), Catania and colleagues (1992), and Walter,
(1992), as well as longitudinal studies by McKusick, Coates, Morin, Pollack, and
Hoff (1990) and Montgomery, et al. (1990) did not find a significant association
between perceived susceptibility and sexual behavior.
In addition to these three behavioral theories, social network theory as
described by Attneave (1986) and Trotter (1999) was invoked to guide the inquiry
into the people in teens lives who were most influential in encouraging or
discouraging safe sexual behaviors. Personal network mapping was used to describe

graphically the people identified as associates by the central person (in this case the
participant). Network data usually include size and demographic characteristics of
the network. In this inquiry, it also included a description of the kinds of advice
about sex and risk behaviors that different associates provided. Mapping individual
social networks and compiling an aggregate map provided an opportunity to reveal
the cultural context of the process by which teens decide on specific sexual
behavioral interactions and act on those intentions. Personal network mapping has
been used to investigate peer influence on adolescent risk behaviors (Brown, 1982;
Brown, 1990), pregnancy (Brown & Theobald, 1999), and sexual debut (Bearman &
Bruckner, 1999b; Resnick et al., 1997).
Design Overview
Adolescent sexuality and sexual behavior are challenging to study because
they are among the most private domains in Western culture. Since direct
observation of teen sexual behavior would be neither ethical nor practical, this
inquiry relied on the observations and recollections of teens to produce a composite
representation of adolescent sexuality, sexual behavior, and the process by which
teens translate into action what they know about protecting themselves from HIV
and other sexually transmitted infections.
Qualitative strategies for data collection offer an opportunity to collect rich
descriptive data about adolescent sexuality and the contextual factors surrounding
teen sexual behavior (Agar, 1996; Atkinson & Hammersley, 1994; Patton, 1990).
This inquiry used four complimentary qualitative strategies to collect verbal data
from the teen participants, a form of triangulation or bringing together multiple
sources of data to bear on the research question. Triangulation has been widely
applied to social science research to enhance the soundness of qualitative research
and make it more applicable to other settings. Specifically, triangulation enables the
data collected from each source to be confirmed, challenged, clarified, and/or

elaborated by data from other sources (Altheide & Johnson, 1994; Marshall &
Rossman, 1995). In this study, using multiple strategies of data collection was
combined with an iterative grounded approach to data collection and analysis in
which successive phases of data collection build on the findings from earlier rounds
of data collection (Strauss & Corbin, 1990).
Although four separate data collection strategies were used, the four samples
of participants overlapped. This overlap of samples is described and illustrated in
detail later in this chapter. The four sets of data were analyzed both separately and in
the aggregate. The intent was for the combined sum of the data to provide a more
accurate and in-depth picture of the contextual factors surrounding teen sexuality,
risk taking, and sexual behavior than could be revealed by any one data collection
The first data collection strategy included three systematic data collection
(SDC) activities: ffeelisting, similarity pile sorts, and rank ordering (Weller &
Romney, 1988). These SDC activities were used to define the domain of risk in
general and the domain of sexual risk more specifically. In addition to defining what
teens do and do not include in the domain of risk, these three systematic data
collecting activities helped reveal how teens think about risk taking and sexual risks.
The second data collection strategy employed in-depth ethnographic
interviews with three young men and three young women over a 9-month period.
This extended period allowed time to establish rapport and trust, and provided a
means to compare data over time. Semi-structured and open-ended interviews
focused on capturing the teen voice in their description of the meaning and context
surrounding adolescent sexuality, sexual behavior, and risk behaviors, and in their
explanation of the process by which teens translate what they know about protecting
themselves from HIV and other STDs into action. Semi-structured interviews were
used to gather information about each participant, the people with whom that
participant associated, and to clarify information from previous interviews. The

semi-structured interview format was also used to focus on themes and interesting
questions that other participants brought up that were identified during the ongoing
data analysis. The regular inclusion of open-ended questions as well, encouraged the
teens to initiate topics for discussion as collaborative research partners.
In addition to participating in the face-to-face interviews, four of the six
primary participants agreed to produce audio journals to record their thoughts and
observations and report the thoughts of peers. The intent of piloting this third
strategy of data collection was for the participants to act as youth researchers
collecting data from youth by youth. Audio journals were also seen as a potential
method for eliciting more sensitive or insightful data from teens. However, only one
audio journal was included in the final data set due to lack of compliance with
confidentiality guidelines, journals being lost in the mail, and non-completion of
A fourth data collection strategy was added after the initial analysis of the
first five months of data detected a pattern of differences in responses between
genders and between urban versus suburban participants. The investigator recruited
eleven additional young men and women to participate in two to three hours of semi-
structured interviews. These participants were selected and the interviews were
designed to confirm, challenge, clarify, and/or elaborate on the themes revealed by
the six primary participants interviews, one audio journal, and the systematic data
collection process.
How did the four sources of data contribute to the inquiry? To more clearly
describe the contribution of the multiple sources of data and the iterative process of
data collection and analysis, the data have been somewhat artificially separated into
two categories primary and refined as illustrated in Figure 3.2. Data collected
from the systematic data collection activities, from semi-structured and open-ended
interviews with primary participants, and from the one audio journal provided what
are called primary data. These data provided the basis for identifying recurrent

semi-structured interview format was also used to focus on themes and interesting
questions that other participants brought up that were identified during the ongoing
data analysis. The regular inclusion of open-ended questions as well, encouraged the
teens to initiate topics for discussion as collaborative research partners.
In addition to participating in the face-to-face interviews, four of the six
primary participants agreed to produce audio journals to record their thoughts and
observations and report the thoughts of peers. The intent of piloting this third
strategy of data collection was for the participants to act as youth researchers
collecting data from youth by youth. Audio journals were also seen as a potential
method for eliciting more sensitive or insightful data from teens. However, only one
audio journal was included in the final data set due to lack of compliance with
confidentiality guidelines, journals being lost in the mail, and non-completion of
A fourth data collection strategy was added after the initial analysis of the
first five months of data detected a pattern of differences in responses between
genders and between urban versus suburban participants. The investigator recruited
eleven additional young men and women to participate in two to three hours of semi-
structured interviews. These participants were selected and the interviews were
designed to confirm, challenge, clarify, and/or elaborate on the themes revealed by
the six primary participants interviews, one audio journal, and the systematic data
collection process.
How did the four sources of data contribute to the inquiry? To more clearly
describe the contribution of the multiple sources of data and the iterative process of
data collection and analysis, the data have been somewhat artificially separated into
two categories primary and refined as illustrated in Figure 3.2. Data collected
from the systematic data collection activities, from semi-structured and open-ended
interviews with primary participants, and from the one audio journal provided what
are called primary data. These data provided the basis for identifying recurrent

themes and interesting questions that would be pursued in subsequent interviews. In
contrast, data collected from subsequent clarifying interviews with primary
participants, the one audio journal, and all interviews with the confirmatory
participants provided the basis for refining the data from which the results and
grounded model are drawn. The process as shown in Figure 3.2 progresses from
primary to refined data, not through a direct linear pathway, but by going through an
iterative loop of clarification and confirmation.
Figure 3.2: Contribution of Multiple Data Sources to the Data

The teen participants in the inquiry were primarily white and relatively
affluent. The decision to study middle and upper class teenage male and female high
school juniors and seniors reflects four assumptions derived from the literature. The
first assumption is that studying more affluent teens would eliminate confounds
associated with poverty that have been found to contribute significantly to teens
having unprotected sex. Such confounds include not having high educational or
career goals and having parents with less than a high school education (Igra & Irwin,
1996; Luker, 1996; Musick, 1991; White, 1999). The second assumption is that
researchers are more likely to gain the cooperation and confidence of participants
with whom they share cultural common ground. This assumption comes from my
personal experience as a nurse and from Erlandson, Harris, Skipper, and Allen,
(1993) who state that recruiting participants from a culture congruent with the
investigators improves the trustworthiness of a qualitative study. The third
assumption is that high school juniors and seniors, the adolescent students most
likely to have experienced sexual intercourse or have friends who have (Kann et al.,
1999; Kann et al., 2000), would be better informed about and able to discuss risky
adolescent sexual behaviors than younger high school or middle school students.
And the final assumption is that studying male and female teens from the same
schools and geographic areas would provide an opportunity to compare and contrast
responses between genders to produce a balanced picture of adolescent sexuality in
this socioeconomic group. Since few qualitative studies have included both males
and females, it seemed that including both would fill an important gap as well.
Consequently, criteria for inclusion in any phase of the study included being
of middle or upper socioeconomic status as indicated by three proxy measures:
1) living in a neighborhood with mean housing costs of $200,000 or greater
(DRCOG Housing Value and Household Estimates 1999,1999; The Piton
Foundation, 1998); 2) attending a public or private high school with college entrance
scores that are consistently above the state and national average (Bingham, 1999;

Camara & Schmidt, 1999; Colorado Department of Education, 2000), and 3) having
a parent or parents with greater than a high school education. In addition, participants
needed to be at least 16 years old and not older than 19 years old. Race and ethnicity
were not specified nor was sexual orientation or virginity status.
Secondary inclusion criteria were based more on purposive(Patton, 1990)
or judgement sampling (Bernard, 1988) in which participants are selected for a
specific reason or purpose. The specific reasons varied with each strategy of data
collection. For the systematic data collection, the intent was to gather data from a
broad spectrum of upper and middle class male and female high school juniors and
seniors, allowing for comparisons between demographically different groups
(Strauss & Corbin, 1990). Accordingly, the sample was designed to contain
approximately even numbers of males and females with approximately even
numbers from urban and suburban neighborhoods, and private and public schools. In
addition, there was an effort to recruit teens who seemed to represent different social
groups based on their interests and extracurricular activities. These groups included
jocks (athletes), band kids (musicians), school leaders(involved in school
organizations like student council), skaters (skateboarders), hip hop (dress in hip
hop style, often with pierced body parts and tattoos) and druggies (teens who
indulge in drugs, especially marijuana, on a regular, often daily basis). Teens who
tended not to socialize with peers were not actively recruited (but were not excluded)
since they might not have had as much personal experience with the contextual
factors that influenced sexual risk behaviors.
Since participants for the primary ethnographic interviews were selected from
the systematic data collection sample, they already met the preceding criteria prior to
being selected to participate in the primary ethnographic interviews. Teens from the
SDC phase were selected for the in-depth ethnographic interviews based on their
apparent personal and observed knowledge of adolescent risk taking, sexual
behavior, and sexual decision-making. Selection criteria also included their being

articulate, being observant, seeming trustworthy, and showing an interest in the topic
of inquiry.
Participants for the confirmatory interviews were selected based on purposive
sampling and convenience. These participants were the most specifically selected
group in that their purpose was to confirm, deny, elaborate, or clarify data previously
collected. Accordingly, since the young women who participated in the primary
interviews were suburban, it was necessary to hear from young women who were
urban. Likewise, the young men who participated in the primary interviews were
urban, so suburban young men were recruited for the confirmatory interviews.
Study protocols and consent forms were approved by the University of
Colorado at Denver Human Subjects Committee (see Appendix D).
Systematic Data Collection
Systematic Data Collection (SDC) strategies and activities were used to
define the cultural domain of risk within which sexual risk is situated. The
structured interviewing process was used to identify the boundaries of the domain of
risk, identify how sexual risk and other risks relate, and capture adolescent thought
processes related to risk and sexuality (Weller & Romney, 1988). The premise
underlying SDC is that individuals presented with the same list of items (in this case,
teenage risks) will organize and relate the items according to unique mental
constructs that vary depending on the individuals social arid cultural experiences.
Using the methods described by Weller and Romney (1988), participants completed
three activities: 1) freelisting or generating a spontaneous list of adolescent risks;
2) unconstrained pilesorting or grouping of risks; and 3) rank ordering or
prioritizing risks. The methods used for each activity are specified below and reflect
some modifications of the Weller and Romney method that were made to
accommodate the adolescent population.