2 ., . .
.35 -
p . unwind-mun
.. u. .

h 3.... I.
nunlx ”hair .
AT

"5 .
{M I. V .I
3mg... 3.3 ad.

.3; 2
mm - n! a ‘
~ r8 ruin. .

. t‘

31.1

bwi! , .
1.. n 129.! . ‘
z..;url..mfi. wwwf‘u

J. 5... . i

5a. tank:

 

 

 

. . ;
.9... 83!,

I) .

771

sfifinfigfi 3.

it:
4 «G2. T ._ ,
.9457.) s. 1‘

 

“ THESIS ’

2001;?

This is to certify that the
dissertation entitled

PERCEPTIONS OF INVULNERABILITY AND
ADOLESCENT SEXUAL ACTIVITY

presented by

Sherry M. Knoppers

has been accepted towards fulfillment
of the requirements for the

degree in Family and Child Ecology

 

 

Wings, Slim

Major Profv sor ’€819nature

’5 .2006

Date

 

MSU is an Affirmative Action/Equal Opportunity Institution

 

LIBRARY
Michigan State
University

 

 

 

 

 

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

 

DATE DUE I DATE DUE DATE DUE

 

065911506}!

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/ClRC/DaleDue.indd-p.1

PERCEPTIONS OF INVULNERABILITY AND
ADOLESCENT SEXUAL ACTIVITY

By

Sherry M. Knoppers

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Family and Child Ecology

2006

\K'

5L

it

 

 

 

ABSTRACT
PERCEPTIONS OF INVULNERABILITY AND
ADOLESCENT SEXUAL ACTIVITY
By
Sherry M. Knoppers

Every year three million teens contract a sexually transmitted disease (Alan
Guttmacher Institute, 1994) and close to one million teens become pregnant (The
National Campaign to Prevent Teen Pregnancy, 2001 ). Youth who choose to abstain
from sexual activity eliminate the risk of contracting a sexually transmitted disease
while avoiding pregnancy and possible emotional consequences. Yet, even with
successful intervention programs, not all youth enjoy the desired outcome. Human
ecology provides a framework to help better understand the multiple influences
impacting adolescents. The adolescent personal fable, first described by Elkind
(1979), is a concept that may also help explain why some youth choose to engage in
risky behavior even when aware of the potential consequences. Are adolescents with
higher levels of invulnerability more likely to engage in sexual activity? The design
of this study was a cross-sectional survey, the purpose of which was to examine the
relationship between perceived invulnerability to pregnancy and STD’s and sexual
activity. Using an adaptation of the Dane County Youth Assessment (University of
Wisconsin Board of Regents, 1999) measures, combined with vulnerability questions
from The New Personal Fable Scale (Lapsley, FitzGerald, Rice, & Jackson, 1989)
and questions developed by the researcher, 67 usable surveys were completed by

youth taking Life Skills classes in a mid sized Midwestern city and the surrounding

area. A Spearman Rho was used to look at the correlation between perceived
invulnerability and sexual activity and logistic regression was used to look at the
relationship between perceived invulnerability and other predictors of sexual activity.
While no relationship was found between feelings of invulnerability and sexual
activity even when controlling for other predictors of sexual activity, perceived
invulnerability did have some impact on the sexual behavior of youth with multiple
risk factors and fewer protective factors. The latter result offers some support for the
notion that perceived invulnerability may impact adolescent sexual activity,
especially in high-risk youth. Interventions to decrease feelings of invulnerability and

overall risk may be helpful in decreasing adolescent sexual activity.

Copyright by
SHERRY M. KNOPPERS
2006

ACKNOWLEDGEMENTS

I would like to extend my sincere appreciation to those individuals who have
contributed to the success of this research project. My deepest appreciation to
Lawrence Schiamberg, my chairperson, for his guidance, support, encouragement and
patience. I would also like to thank Thomas Luster, Robert Griffore, and John
Eulenberg, committee members, for their input, suggestions, and support. I am also
thankful for the time and energy the supervisor and group facilitators at Wedgwood
put into helping me obtain the needed surveys.

I want to extend a special note of appreciation to my husband John and sons,
Landon and Logan. Their patience and support through this long process was
commendable! Above all I hope that this work will bring honor and glory to God and

that its impact will be pleasing to Him.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................... ix
CHAPTER 1
INTRODUCTION AND STATEMENT OF THE PROBLEM ........................ 1
Introduction ........................................................................... 1
Purpose ................................................................................. 6
CHAPTER 2
REVIEW OF RESEARCH .................................................................. 8
An Ecological Perspective to Youth Sexual Activity ........................... 8
Individual Level ............................................................. 9
Contexts of Youth Sexual Behavior ...................................... 11
Positive Youth Development ...................................................... 17
Interventions ......................................................................... 22
Specific Processes and Contexts of Adolescent Sexual Activity ............ 29
Youth Cognitive Interpretation and Perception of Risk Factors. 29
The Adolescent Personal Fable .............. . ............................ 32
Risk and Protective Factors ............................................... 35
Conclusion ........................................................................... 41
CHAPTER 3
METHODS .................................................................................... 42
Purpose of the Study ................................................................ 42
Definitions and Measures ........................................................... 42
General Definitions and Measures ....................................... 42
Research Design ..................................................................... 47
The Sample ........................................................................... 48
Method of Data Collection ........ » .................................................. 49
Reliability and Validity of Measures ..................................... 49
Study Implementation ...................................................... 50
Hypotheses ............................................................................ 51
Hypotheses relating to sexual behavior .................................. 51
Hypotheses relating to individual risk factors ........................... 52
Hypotheses relating to microsystem factors ............................. 53
Hypotheses relating to macrosystem/community factors .............. 54
Data Analysis for Hypotheses ...................................................... 54
Protection of Human Subjects ...................................................... 55
Limitations ............................................................................ 56
CHAPTER 4
RESULTS ..................................................................................... 57
Surveys ............................................................................. 57
Descriptive Statistics ............................................................... 57

vi

Correlation between Feelings of Invulnerability and Sexual Activity ....... 70
Correlation of Factors Impacting Sexual Activity and Feelings of

Invulnerability ........................................................................ 84
CHAPTER 5
DISCUSSION OF RESULTS ............................................................. 94
Youth Demographics .............................................................. 94

Relationships between Feelings of Invulnerability and Sexual Activity. . .. 95
Youth with higher perceived invulnerability will be

more likely to be sexually active ........................................ 98
Youth with higher perceived invulnerability will

have greater frequency of sexual activity .............................. 99
Youth with higher perceived invulnerability will

have greater number of lifetime partners ............................... 99
Youth with higher perceived invulnerability will more

frequently have sex while drunk or high ............................... 99
Youth with higher perceived invulnerability will have

greater use of birth control ................................................ 100
Youth who have ever had an STD will have lower

levels of perceived invulnerability ...................................... 101
Youth who have ever been pregnant or gotten someone

pregnant will have lower levels of perceived invulnerability ........ 102

Youth with higher perceived invulnerability will be more likely

to be sexually active when other predictors of sexual activity are
controlled .................................................................... 102
Perceived invulnerability will have less impact on the sexual

behavior of youth with few risk factors and more protective

factors while having a greater impact on those with multiple

risk factors and fewer protective factors ................................ 104

Relationship between Factors Impacting Sexual Activity and Feelings

of Invulnerability .................................................................... 107
Younger youth will have higher levels of invulnerability ............. 107
Youth who use alcohol or drugs will have higher levels of
invulnerability ............................................................... 107
Youth with a history of sexual abuse will have lower levels of
Invulnerability ............................................................... 109
Youth with a history of physical abuse will have lower levels of
invulnerability ............................................................... 109
Youth who have witnessed physical abuse will have lower levels
of invulnerability ............................................................ 110
Males will have higher levels of invulnerability ........................ 110
Minority youth will have higher levels of invulnerability ............. 11]

Youth with higher GPAs will have lower levels of invulnerability...112
Youth with strong religious beliefs will have lower levels of
invulnerability ............................................................... l 12

vii

Youth who believe that teens should not have sexual intercourse

will have lower levels of invulnerability ................................ 113
Youth who live with both parents will have lower levels of
invulnerability ............................................................... 1 14
Youth whose parents have never been divorced or separated

will have lower levels of invulnerability ................................ 116
Youth with higher levels of parental monitoring will have lower

levels of invulnerability .................................................... 116
Youth with higher levels of parental support will have lower

levels of invulnerability .................................................... 117
Youth who have parents with higher levels of education will have
lower levels of invulnerability ............................................ 117
Youth with parents who are more approving of adolescent sexual
activity will have higher levels of invulnerability ...................... 118
Youth with parents who are more approving of adolescent alcohol
consumption will have higher levels of invulnerability ............... 119

Youth who report having friends who get into trouble, are sexually
active, drink or use drugs, or smoke will have higher levels of

invulnerability ............................................................... 120
Youth reporting supportive communities will have lower levels
of invulnerability ............................................................ 120
Conclusion ............................................................................ 12]
Limitations ................................................................... 121
Implications for future research, theory and social
policy/intervention .......................................................... 124
APPENDICES ................................................................................ 128
REFERENCES ................................................................................ 140

viii

Table 1
Table 2
Table 3
Table 4
Table 5
Table 6
Table 7
Table 8
Table 9
Table 10
Table 11
Table 12
Table 13
Table 14
Table 15

Table 16

Table 17
Table 18
Table 19
Table 20

Table 21

LIST OF TABLES

Youth Age ................................................................... 57
Youth Race/ethnicity ...................................................... 57
Average Grades ............................................................. 58
Where Youth Live .......................................................... 58
With Whom Youth Lives ................................................. 58
Mom’s Education .......................................................... 59
Dad’s Education ............................................................ 59
Parents Divorced or Separated ........................................... 59
Youth Smoking ............................................................. 60
Youth Drinking ............................................................. 60
More than Five Drinks in the Last Month .............................. 60
Youth Marijuana Use ....................................................... 61
Youth Use of Other Drugs ................................................. 61
Youth Don’t Fear Pregnancy with Unprotected Sex ................ 61
Youth Don’t Fear STDs with Unprotected Sex ...................... 62
Youth Think They Could Get Pregnant with

Unprotected Sex ............................................................. 62
Youth Think They Could Get an STD with Unprotected Sex ..... 62
Religion Important .......................................................... 63
Not Very Involved in Religious Activity ............................... 63
If Ever Had Sex Age 1St Voluntary Sex ................................. 64
How Often Youth Had Sex in Last Three Months ..................... 64

ix

Table 22

Table 23

Table 24

Table 25

Table 26

Table 27

Table 28

Table 29

Table 30

Table 31

Table 32

Table 33

Table 34

Table 35

Table 36

Table 37

Table 38

Table 39

Table 40

Table 41

Table 42

Table 43

Number of Lifetime Partners .............................................. 64
How Often Youth Have Sex While Drunk .............................. 65
How Often Youth Use Birth Control .................................... 65
Witnessed Abuse or Beating ............................................. 65
Neighbors Could Help with a Problem ................................. 66
Community Adult Would Tell Parent if Youth Did Something

Wrong ...................................................................... 66
People in the Community Know and Care about Each Other ...... 66
Youth Tell Adults Before Going Out ................................... 67
Youth Talk with Adults about Plans .................................... 67
Adults Ask When Youth Go Out ....................................... 67
Adult Knows What Youth Do after School ........................... 67
Parent There When Needed ............................................. 68
Parent Cares About Youth .............................................. 68
Parents Say Teens Should Not Have Sex .............................. 68
Parents Say Teens Should Not Drink ................................... 68
Youth Friends Help Them Stay Out of Trouble ..................... 69
Most Friends Do Not Have Sex ........................................ 69
Most Friends Do Not Drink or Do Drugs ............................ 69
Most Friends Do Not Smoke ............................................ 69
Youth Belief That Teens Should Not Have Sex ..................... 70

Logistic Regression of Sexual Vulnerability and Top 4 Correlates. 76

Logistic Regression of 1St Two Sexual Vulnerability Questions
and Top 4 Correlates ....................................................... 76

Table 44

Table 45

Table 46

Table 47

Table 48

Table 49

Table 50

Table 51

Table 52

Table 53

Table 54

Table 55

Table 56

Table 57

Logistic Regression of 2“d Two Sexual Vulnerability Questions

and Top 4 Correlates .....................................................

Logistic Regression of Personal Fable and Top 4 Correlates .......

Logistic Regression of 5 Items with Highest Loading on the

Personal Fable Scale and Top 4 Correlates ............................

Logistic Regression with Top 4 Dichotomous Variables and

Sexual Vulnerability ......................................................

Logistic Regression with Top 4 Dichotomous Variables

and 1St Two Sexual Vulnerability Questions ...........................

Logistic Regression with Top 4 Dichotomous Variables

and 2nd Two Sexual Vulnerability Questions ..........................

Logistic Regression with Top 4 Dichotomous Variables

and Personal Fable Scale .................................................

Logistic Regression with Top 4 Dichotomous Variables

and 5 Items with Highest Loading on the Personal Fable Scale

Logistic Regression with Total Risk and Sexual Vulnerability. . .

Logistic Regression with Total Risk and 1St Two Sexual

Vulnerability Questions ..................................................

Logistic Regression with Total Risk and 2"d Two Sexual

Vulnerability Questions ..................................................

Logistic Regression with Total Risk and Personal Fable Scale.

Logistic Regression with Total Risk and 5 Items with Highest

Loading on the Personal Fable Scale ..................................

Youth Age and Chronological Age of 1St Sex Cross Tabulation. .

xi

.. 76

..77

.77

.78

.78

.79

.79

80

83

83

84

84

84

97

Chapter 1

Introduction and Statement of the Problem

Introduction

Adolescent sexual activity is a topic of interest and concern to society.
Healthy People 2010 Ten Leading Health Indicators for the Nation and the Surgeon
General’s public health priorities both include promoting responsible sexual behavior
(Satcher, 2001). Every year three million teens contract a sexually transmitted disease
(Alan Guttmacher Institute, 1994). Nearly 1 in 4 sexually active young people will
get a sexually transmitted disease or infection (STD) (Kaiser Family Foundation,
2003). The Office of the Surgeon General (2001) reported that rates for gonorrhea are
greatest among 15-19 year olds and females in this age group have the highest rates
of Chlamydia. In addition, close to one million teens become pregnant each year
accounting for about 1/5th of all sexually active females (Kalmuss, Davidson, Cohall,
Laraque, & Cassell, 2003; The National Campaign to Prevent Teen Pregnancy, 2001).
Nearly half of all unintended pregnancies occur in adolescents (Office of the Surgeon
General, 2001). These adolescent pregnancies place a large burden on society both in
medical and social costs and as these children grow up, they are at a very high risk for
engaging in early sexual activity themselves (Levine, Pollack, & Comfort, 2001).

Rector, Johnson, and Noyes (2003) report that the majority of sexually active
teens wish they had waited longer before beginning sexual activity and that teens who
are sexually active are significantly less likely to be happy. They are also more likely
to feel depressed and attempt suicide than teens that are not sexually active (Rector et

al., 2003). Topolski et a1. (2001) looked at quality of life indicators and found that

youth who abstain from high risk sexual behaviors report a higher quality of life than
those who experiment or engage regularly in these behaviors.

Youth who choose to abstain from sexual activity can eliminate the risk of
contracting sexually transmitted diseases while avoiding pregnancy and possible
emotional consequences. The Physicians Consortium holds that it is best for
adolescents to wait until marriage for sexual involvement as an arbitrary delay to just
a “later age” still leaves youth at risk for STD infection (Diggs, Wallis, Mohn, &
Jones, 2001). Youth who become sexually active at an earlier age usually end up
having more sexual partners, which is linked to greater STD risk. Condom use
generally decreases over time, so youth who become sexually active at an earlier age
are likely to have a greater number of sexual acts unprotected against STDs and teen
pregnancy than teens that begin sex at a later age (Diggs et al., 2001). In a study of
rural youth, Yarber, Milhausen, Crosby, and DiClemente (2002) had similar findings
and surmised that rural youth who initiate sexual activity at an early age are at an
increased risk of engaging in subsequent sexual risk behaviors, such as having
multiple sex partners and not using a condom.

If adolescents wait until they are in stable marriage relationships before
becoming sexually active and remain faithful in these relationships, many societal
problems could be eliminated. Single parenting would be limited to those losing their
partner to divorce, still a societal concern, or death. Those who continue to remain
faithful in the marriage relationship would have no fear of STD’s. Even HIV infection
would decrease drastically, being spread only by those using contaminated drug

needles or by rare accidental exposure such as with medical personal.

With this in mind, few would disagree that sexual abstinence is best for
adolescents, yet the high percentage of sexually active teens indicates that many
youth are not abstaining from sex. In 1999 almost 50% of high school students had
engaged in sexual intercourse at least once with around 16% of these youth having
been with four or more sexual partners (Kann et al., 2000). By their senior year of
high school, nearly two-thirds of adolescents reported that they had engaged in sex
(Kaiser F amily Foundation, 2003).

While promoting abstinence is a laudable goal it is unclear how it can best be
accomplished. It is even more challenging when a clear definition of sexual activity
and sexual abstinence is lacking (Bailey, Young, Knickerbocker, & Doan, 2002).
Hawkins et a1. (2002) examined adolescent perceptions of the terms abstinence and
sexual activity. While perceptions of abstinence were related to age, other factors
such as gender, virginity status, attendance at religious services, and perceived
religiosity did not impact perception. Haglund (2003) found most sexually abstinent
African American females in this study referred to “having sex” as vaginal
intercourse and being abstinent as refraining from intercourse. Knoppers (2003b)
defines sexual abstinence as “not engaging in sexual activity that could put a youth at
risk of contracting a sexually transmitted disease or cause pregnancy.” This definition
focuses on the negative societal implication to sexual activity, framing it broadly
enough to cover potential consequences. It can also be viewed on a continuum
ranging from any type of physical contact at one end of the scale to doing any type of
sexual activity except actual vaginal intercourse at the other end (Knoppers, 2003b).

If youth view sexual activity at this later end of the continuum it could put them at

risk for the consequences of sexual activity if they do “everything but”, thinking this
is still abstinence (Haglund, 2003).

Many approaches to promoting sexual abstinence in teens have been tried
with varying results. The majority of programs provide youth with information about
the risks involved with sexual activity. If adolescents have this information why then
do so many still engage in sexual activity? What other factors impact their behavior?
Could feelings of invulnerable to the potential consequences play a part?

Anderson, Nyamathi, McAvoy, Conde, and Casey (2001) found that
adolescents in juvenile detention facilities did not see themselves as being at risk for
HIV infection even when sexually active. Because they associated HIV with drug use
and did not know many infected people, they were more worried about gangs, drugs,
and violence in their neighborhoods and did not see sexual activity as a concern. Even
with risk factors present, few girls in another study perceived themselves to be at risk
for HIV (Morrison-Beedy, Carey, & Aronowitz, 2003). Williams, Norris, and Bedor
(2003) had similar findings in a study of college students. Less than half of them used
condoms at last intercourse with no correlation to the type of partner they had, and
they did not express any fear of HIV or other STDs. Worse yet, DiClemente et al.
(2002) found that teens who had had sexually transmitted diseases were more
knowledgeable about STD prevention but were more likely to engage in unprotected
sex.

Chapin (2001) found that a group of high-risk African American adolescents
felt that they were less likely than their peers to become pregnant (or cause a

pregnancy). These youth displayed an optimistic bias, underestimating their personal

risk in relation to their peers. High school students with high feelings of perceived
invulnerability also perceived themselves as less susceptible to sexual risks. They
may well believe that if they had engaged in risky sexual behavior without any
negative consequences, they are immune (Chapin, 2001).

Another study found no relationship between frequency of participants’ risk-
reducing behaviors and their perceived probability of HIV infection (Cohen & Bruce,
1997) while Hutchinson (1999) did find a correlation in youth who were consistent in
using condoms, satisfied with their relationships, and perceived their partner as
uninfected. These young women had increased odds of believing they were at no risk
for sexually transmitted diseases or infections including HIV, likely related in part to
their perceived lack of risk and their risk reducing behavior. Apparently feeling good
about their relationships and believing that their partners did not have any STDs
along with taking the precaution of using condoms gave them a feeling of safety or
invulnerability.

Porter, Oakley, Guthrie, and Killion (1999) did not find initiation of sexual
intercourse correlated to any shifis in perceived costs or benefits but to increases in
the frequency of the intimate behaviors of going out together alone, kissing and
making out, and the opportunity for intercourse when home alone. So these youth
looked less at perceived costs and benefits and more at behaviors leading to sexual
intimacy and the opportunity to progress to that level.

Adolescents can perceive benefits and costs associated with unprotected sex,
though. Parsons, Perry, Bimbi, and Borkowski (2000) found that college students’

sexual risk behaviors were correlated to their perceptions of the benefits associated

with unprotected sex. They seemed to perceive the positive outcomes associated with
unprotected sex as greater than the costs or potential negative outcomes, in their
minds justifying their sexual risk-taking behaviors. In this study perceptions of the
positive outcomes of unprotected sex, along with the inability to resist temptation and
low self-efficacy for safer sex, were predictive of sexual risk-taking for these youth
(Parsons et al., 2000). In an earlier study Parsons, Siegel, and Cousins (1997) also
found that risk perception seemed less important than perceived benefits to the risks
involved with sexual risk taking. If youth’s perception of benefits better predicts risk-
taking behavior than perceived costs, addressing this area of adolescent prevention
may be important in decreasing sexual risk taking among adolescents.

Yet this aspect of adolescent sexuality needs further study. Numerous studies
have shown that many different variables correlated to adolescent sexual activity, but
what is the impact of feelings of invulnerability? Do adolescents’ perception of
invulnerability to the potential negative consequences of sexual activity impact their
intentions or behavior? While this piece alone warrants consideration, many other
factors also impact sexual behavior. What is the relationship between feelings of
invulnerability and other risk factors for adolescent sexual behavior?
film

The purpose of this study is to answer these questions: Is there a relationship
between youths’ perceived invulnerability to pregnancy and STDs and sexual
activity? And, what is the relationship between youths’ perceived invulnerability to
pregnancy and STDs and sexual activity when considering other predictors of sexual

activity (e. g. drug use, peer group, parental support)? This study also looked to

answer questions about how the individual risk and protective factors for adolescent
sexual activity as well as factors at the microsystem and macrosystem levels relate to

feelings of invulnerability.

Chapter 2
Review of Research

Human ecology provides a framework to help better understand the multiple
influences impacting adolescents. Viewed from a human ecological perspective, how
an adolescent develops is a function of the individual person and environment over
the period of time that person is developing (Bronfenbrenner, 1989). The ecological
model is a dynamic interactive model where the individual adapts to the environment,
and the environment positively or adversely impacts development (Blum, McNeely,
& Nonnemaker, 2001). Bronfenbrenner (1989) sees humans inextricably embedded in
their environment. Both individual factors and environmental factors may contribute
to an adolescent’s decision to be sexually active or to abstain from sexual activity.
The base of behavior and development lies in the dynamic relationships within the
individual-context system. The relative plasticity of developmental systems can help
in planning interventions to prevent negative developmental trajectories, and can also
provide ideas on how to promote system changes that can optimize positive healthy
functioning (Lerner, 2003).
An Ecological Perspective to Youth Sexual Activity

Perkins, Luster, Villarruel, and Small'(1998) found that factors at the
individual, family, and extra-familial level were all predictive of sexual activity in
adolescents from different ethnic groups, though there were a few gender differences.
Talashek, Norr, and Dancy (2003) also discuss individual, family, and environmental

factors impacting adolescent sexual risk taking.

Individual level.

On an individual level, youth with frequent attendance at religious services,
who are more actively involved in religious activities or attached to religious
institutions, have strong religious beliefs, or those who make virginity pledges are
much less likely to be sexually active (Brewster, Cooksey, Guilkey, & Rindfuss,
1998; Diggs et al., 2001; Kalmuss et al., 2003; Kirby, 2001b; Larnmers, Ireland,
Resnick, & Blum, 2000; Satcher, 2001; and Scales & Leffert, 1999). While religious
beliefs are very individual, the adolescent’s environment may have an influence as
well.

Also looking at the individual, Paradise, Cote, Minsky, Lourenco, and
Howland (2001) found that in their study of adolescent girls’ reasons for having or
not having sex, values and beliefs were cited by the majority of virgins. Most of these
girls felt that the time was not right for them, they wanted to wait until they were
older, or they wanted to wait until they were married. Self-efficacy can also impact an
adolescent’s decision to remain sexually abstinent (Taris, & Semin, 1999). Kirby
(2001b) found that adolescents’ emotional well being, their own sexual beliefs and
norms about sexual behaviors, their attitudes, their skills, and their motivation also
can affect sexual behaviors.

Halpern, Joyner, Udry, and Suchindran (2000) found that higher intelligence
was associated with postponement of the initiation of the full spectrum of sexual
activities. They concluded that higher intelligence in the individual is a protective
factor against early sexual activity during adolescence even when controlling for age,

physical maturity, and mother’s education.

The time between puberty and psychosocial maturity has grown larger with
puberty being reached at an earlier age today than it was in the past (Baumrind, 1987)
and many young adults getting married later. Youth who reach puberty earlier than
their peers are at increased risk (Bearman & Bruckner, 2001). Doswell and Braxter
(2002) also note this increased risk for early sexual behavior with early pubertal
development. Hormone levels can make a difference as well (Kirby, 2001b). Halpern,
Udry, and Suchindran (1997) found that testosterone and changes in testosterone were
significantly related to the timing of sexual initiation. While this physiological aspect
of the individual did impact timing of first sex, the frequency of attendance at
religious services acted as a social control variable for subjects. The effect of
testosterone was negated for those youth who attended religious services more
frequently (Halpem et al., 1997).

Green, Kromar, and Walters (2000) looked at sensation seeking as an
individual variable associated with high risk behaviors including risky sexual
behaviors. They also noted that this factor peaks during the adolescent years,
especially in males. This coincides with the rise in testosterone, which may contribute
to sensation seeking. They also acknowledged, though, that problem behaviors such
as sexual risk taking are related to family socialization and communication patterns as
well (Green, Kromar, & Walters, 2000). Owens and Shaw (2003) see individual
characteristics as among the most influential of the environmental factors.

Another relevant individual factor is the adolescent’s cognitive interpretation
or perception of risk related to sexual activity. Feelings of invulnerability, in

particular, may impact adolescent sexual behavior. This concept, under the heading of

10

adolescent personal fable, will be discussed further starting on page 32 while the
narrower aspect of feelings of invulnerability will be more carefully addressed after
the discussion of risk and protective factors beginning on page 35.

Contexts of youth sexual behavior.

Environmental impacts on individuals and their interactions with the
environment occur at several different levels. Bronfenbrenner (1989) categorized the
microsystem as a place where face-to-face interactions occur, the mesosystem as the
links between microsystems, exosystems as similar to mesosytems but with at least
one of the microsystems not containing the developing person, and the macrosystem
as consisting of the overarching belief system of a given culture or social context.

At the microsystem level of the family, parental involvement including close,
warm parent-child relationships and amount of supervision or monitoring, has been
shown to influence adolescent sexual activity (Blum & Rinehart, 1997; Satcher, 2001;
The National Campaign to Prevent Teen Pregnancy, 2001; Wu et al., 2003) especially
if the parents are good communicators (Karofsky, Zeng, & Kosorok, 2001; Miller,
Norton, Fan, & Christopherson, 1998). Blake, Simkin, Ledsky, Perkins, and
Calabrese (2001) even found that adding homework assignments that require parental
interaction to existing abstinence-only curriculum, in essence promoting parental
involvement, resulted in greater self-efficacy for refusing high-risk behaviors. The
effect of parental involvement may even go back to infancy. Werner and Smith
(2001) found that having had less anxious, insecure relationships with their caregivers
as infants and a stronger feeling of security as part of their families in adolescence led

to better outcomes for youth who had been teenage mothers but managed to do well

11

 

later in life. So while these factors did not prevent the adolescents from engaging in
sexual activity and the resulting pregnancy, they did correlate with better long term
outcomes for them. Owens and Shaw (2003) see these characteristics of the parent
and family environments as having a great impact.

In this microsystem youth also care about what their parents think, or at least,
what they think their parents think. One study found that the more disapproving
adolescents perceived their mothers to be toward their engaging in sexual intercourse
and the more satisfied they were with their relationship with their mothers, the less
likely they were to initiate sexual activity (Dittus & Jaccard, 2000). Kalmuss et al.
(2003) also found those who thought their mothers disapproved of their having sex
were less likely than others to engage in sexual activity. When parents communicate
stricter values to their teens about having sex or about premarital sex in general, the
youth are then more likely to wait longer before engaging in sex. So, family values
about sexual behavior can have an impact on the adolescents’ behavior (Kirby,
2001b). Miller (2002) agreed with this in finding that parental supervision along with
parents’ values against teen intercourse did impact adolescent risk taking.

Basic demographic aspects of the family microsystem can also make a
difference. For example, the presence of the father in the child’s household at ages
11-12 decreased the odds of initiating sex in early adolescence in a sample of
adolescents whose mothers had given birth to them as teens (Cooksey, Mott, &
Neubauer, 2002). Lammers, Ireland, Resnick, and Blum (2000) found youth in dual-
parent families showed lower levels of sexual activity and delayed onset of sexual

activity as well. Miller (2002) also found that family structure (mainly having two

12

 

parents in the home, not just one) had an impact, as did Kirby (2001b) who also found
levels of education and income correlated. Looking at factors associated with the
increase in the teen birthrate in the 1980’s, negative changes in family environments
(such as increases in family disruption) are believed to have contributed to this while
recent declines in teen birthrates may be attributed to positive changes in family
environments such as improvements in maternal education (Manlove, Terry, Gitelson,
Romano, & Russell, 2000).

Still looking at the parental aspect of the family microsystem, authoritative
parenting, which is not overly strict or authoritarian, helps guide children’s activities
with consistency and requires them to contribute to the family. Authoritative parents
are not afraid to confront their children so that they understand the family values and
respect these norms. Children with authoritative parents are loved, supported,
stimulated, and challenged. These parents balance the ratio of children’s autonomy to
parental control so that young children have stricter controls and adolescents are
allowed more autonomy (Baumrind, 1987). The family microsystem needs to balance
expressions of individuality with those of connectedness. If this microsystem leads to
emotional distance from the family it may result in youth who are more susceptible to
peer influences. Strong attachments between parents and youth that continue through
adolescence along with consistent management policies in traditional families help
shield youth from dysfunctional risk-taking behavior (Baumrind, 1987). Along these
lines, adolescents who are provided with rational explanations for parental decisions
are more likely to use their parents as role models and to choose friends that their

parents would consider appropriate (Baumrind, 1987). Hetherington and Elmore

l3

(2003) found that girls with divorced parents were more likely to live in a single-
parent home with non-authoritative parenting and often a sexually active mother. This
combination of risk factors was associated with earlier and more promiscuous sexual
activities and teenage pregnancy. Conversely having a close relationship with an
authoritative parent promoted well-being for children from divorced homes
(Hetherington & Elmore, 2003).

Kerr, Beck, Shattuck, Kattar, and Uriburu (2003) also found that higher levels
of parental monitoring and familial connectedness were consistently associated with
less problem-behavior in adolescents as did Kirby (2001b) and Martyn and Martin
(2003). With this in mind the value of statutes that allow adolescents to gain access to
reproductive health care without parental consent may be seen as undermining
parents’ ability to monitor their teens and may not be in the best interest of the
adolescent (Merrick, 1996).

Cauce, Stewart, Rodriguez, Cochran, and Ginzler (2003) consider that with an
ecological model of development more distal risks will be filtered through more
proximal environments allowing positive family and peer microsystems to decrease
youth vulnerability. Luthar and Zelazo (2003) also emphasize the importance of the
proximal environment of the family in resilience-based intervention for children
stressing the importance of the quality of parent-child relationships. Yet, the riskier
the setting, the less likely it is that protective factors will be present

Another aspect of many family microsystems is siblings. Having older
sexually active siblings or pregnant/parenting teenage sisters increases teen risk

(Miller, 2002). East and Kieman (2001) found an even stronger correlation if teens

14

had two or more parenting sisters living with them. The younger sisters in these
families had more permissive sexual beliefs and attitudes about having children and
the younger brothers were engaged in sexual activity at younger ages than those
without parenting teen sisters living with them.

Peer group interaction can also be viewed as a microsystem for influencing
sexual behavior (Kalmuss et al., 2003). Social influences in the peer group, including
having friends who are sexually active, significantly influences the adolescent’s
likelihood of becoming sexually active at an early age. For example, youth who
perceive that their friends were engaging in high-risk behavior were more likely to
engage in sexual risk taking (Boyer et al., 2000). Peers’ norms and behavior regarding
sex affect adolescents’ sexual behavior (Kirby, 2001b).

The school microsystem could impact adolescent sexual risk taking as well
(Kirby, 2002c). Involvement in school and attachment to school were all related to
less sexual risk-taking and lower pregnancy rates in adolescents. Mothers’
perceptions of a quality school environment were also correlated to adolescents being
less likely to get into trouble (Kowaleski, 2000), and the Search Institute lists a caring
school climate and clear rules and boundaries provided by schools as developmental
assets or factors that protect youth against risky behavior (Benson, 1997).

Considering cultural impacts at the macrosystems level, even being exposed
to the sexually explicit material that is so prevalent in our society can have a negative
impact on attitudes and expectations as well as increasing the likelihood of sexual
activity (Roberts, 1993; Strong, DeVault, & Sayad, 1996; Ward & Rivadeneyra,

1999). Ward and Rivadeneyra (1999) found not only viewing a considerable amount

15

but how involved the viewer was with the program impacted adolescents’ sexual
attitudes, expectations, and behavior. Those with greater exposure who also were
more involved with the sexual content on TV. were more likely to be accepting of
recreational sex, believe more of their peers were sexually activity, and to be more
sexually active themselves (Ward & Rivadeneyra, 1999).

At the macrosystem level, influential characteristics of communities may
include risk factors or protective factors. Communities with more protective factors
should be better able to deal with adversity. While positive relationships at the
microsystem level can help counter negative community effects, the opposite effect is
also possible where positive community environments can help counter negative
influences that may occur at the microsystem level. Denner, Kirby, Coyle and Brindis
(2001) found this to be the case where traditional values about family and
community, close ties to religious institutions, monitoring youth, and being protective
of girls in one Hispanic community resulted in lower teen birth rates than another
community with similar socioeconomic factors that did not have these characteristics.
Parental notification laws can be an asset to a community like this while statutes that
undermine parental monitoring may put youth at risk. With this in mind, it is
important to look at a community’s sexual attitudes and norms, especially in those
where a large percentage of teens are having sexual intercourse at young ages, to try
and deter sexual initiation among youth (O’Donnell, Myint, O’Donnell, & Stueve,
2003). O’Donnell, O’Donnell, and Stueve (2001) also point out that these
community factors must be taken into consideration when planning interventions, as

programs are usually more effective if they are able to reach youth before they

16

become sexually active. In communities with high rates of early sexual activity
among youth, interventions need to reach youth at younger ages since it can be very
difficult to promote abstinence to students when the majority of them may already be
sexually experienced.

Lammers, Ireland, Resnick, and Blum (2000) found higher socioeconomic
status associated with lower levels of sexual activity and delayed onset of sexual
activity. Seidman and Pedersen (2003) see the various effects of poverty cascading
into the neighborhoods, schools, peer groups, and families and impacting the daily
lives and experiences of adolescents. Poverty (also associated with low levels of
education, poor schools, high unemployment, poor housing, increased divorce rates
with more single mothers, higher levels of family dysfunction, and increased crime
levels) was also found to impact teen birthrates (Kirby, Coyle, & Gould, 2001; Miller,
2002) indicating increased sexual activity among youth. Kowaleski (2000) found that
residential stability decreased adolescent risk-taking attitudes, regardless of the level
of disadvantage present within the community, so certain community characteristics
can have a positive impact on behavior, even though others might have a negative
impact.

Positive Youth Development

While looking at multiple environmental influences helps put adolescent
sexual activity in context, much of the literature still emphasizes problems and
deficits, perhaps with some attention to prevention. Changing the focus to the
plasticity and strengths of adolescents and their families as well as their community

contexts, it may be possible to capitalize on the strengths of developmental systems to

17

better promote healthy functioning and develop more effective and informed
interventions (Lerner, 2003). Even if problems are prevented, adolescents may not, in
reality, receive the support or the assets they need for optimal development. With a
focus on resiliency, protective factors, and positive youth development communities
can strive for the goal of not only “problem free” but “asset rich” youth (Benson,
2003). Yates, Egeland, and Soufe (2003) see resilience as an ongoing process of
acquiring resources that can help adolescents adapt and provide them with a
foundation for handling later challenges. Resilience is not viewed as the reason youth
may do well when confronted with adversity but is more a developmental processes
that helps youth gain the ability to use internal and external assets to realize positive
adaptation even in the midst of adversity. When youth are in caring communities that
support positive development, their ability to develop morally and better contribute to
civil society is enhanced. Adolescents need to be provided with the individual and
ecological assets that can help provide them with the “five C’s” of positive youth
development: competence, confidence, connection (to family, peers, and community),
character, and caring/compassion (Lerner, 2003). Luthar and Zelazo (2003) see
resilient adaptation as resting on good relationships.

The Search Institute’s 4O developmental assets provide a guide for enhancing
the strengths of individuals, families, and communities. When more assets are present
the chances of healthy development increase (Lerner, 2003). In the presence of risk
factors an additive or compensatory models suggests that more resources can
decrease the negatives resulting in better outcomes while a moderating model would

see these resources as decreasing the child’s susceptibility to the harmfulness of the

18

stresses or risk factors or somehow protecting them from the negative effects
expected from the threat (Masten & Powell, 2003). In one example the asset of
positive parenting had the effect of lowering mental health problems for children
exposed to the stress of parental divorce or death (Standler et al., 2003). Fergusson
and Horwood (2003) suggest that factors can act additively and may mitigate or
exacerbate the effects of exposure to childhood adversity. It is not usually a single
factor that causes difficulties in adolescents, but an accrual of difficulty that decreases
developmental capacity and it is not generally just one environmental factor that
makes a difference but rather a collection of risk in each family’s life (Samerotf,
Gutrnan, & Peck, 2003). In a longitudinal study of Hawaiians born in 1955 Werner
and Smith (2001) found that as the number of risk factors or stressful life events
increased, more protective factors or assets were needed to counter act the negatives
in the lives of these vulnerable children and to ensure positive deve10pmental
outcomes.

School success, leadership, valuing diversity, physical health, helping others,
delaying gratification, and overcoming adversity were found to be indicators that
youth were thriving (Lerner, 2003). Adolescents able to delay gratification and
concerned with physical health would be more likely to abstain from sexual activity
while the other factors also contribute to overall developmental health. Communities
that show cohesiveness, caring, and compassion are better able to support adolescents
and help them develop personal character and competence (Lerner, 2003).
Communities with these assets are also better able to instill moral values and support

spiritual and religious faith that also has a positive impact on adolescent development.

19

Positive youth development supports an ecological approach to human
development. The macroecological level is where policy and political structures
impact the community while communities strive to provide asset rich environments.
The role of the family in linking youth to the community remains important
especially when looking at parenting practices. Communities that focus on rebuilding
and strengthening their developmental infrastructure with a mobilization of public
will and capacity will likely fair better than those driven from the top down. Creating
a culture where all residents promote the positive development of children and
adolescents is a wonderful goal (Lerner, 2003). Asset building can occur from the
individual-level with residents in informal relationships to the macrosystem level with
community building helped by the local economic and governmental structures
(Benson, 2003).

Furthermore, even gang youth have developmental plasticity and the potential
for positive developmental change when linked in their ecologies to developmental
assets (Taylor et al., 2002). Enhancing community resources and assets could even
help youth with only a small number of assets to unlock their potential for positive
development (Taylor et al., 2002).

This work of community-based human development or asset-based
community and human development has a great impact on adolescent well-being.
Vulnerable populations, in particular, are served by an infrastructure that distributes
resources aimed at decreasing risks and promoting health. It is also important for
communities to support infrastructure that provides safe places and adult connections

while promoting competency building in youth. How attentive a community is to

20

these essential developmental needs determines the strength of the human
development infrastructure of a community (Benson, 2003).

Having long-term relationships with non-related adults, feeling a connection
to their neighborhood where other adults know and interact with them, and
participating in caring and supportive schools have all been related to positive
adolescent health outcomes (Benson, 2003). Transmitting values and standards,
providing support, controlling behavior, and promoting belonging are all impacted by
the consistent presence of adults in the community in a variety of contexts from the
home setting to the neighborhood and religious settings (Benson, 2003). Werner and
Smith’s (2001) findings support this. Looking at long range outcomes what mattered
most for the children followed in this longitudinal study was the emotional support of
members of the extended family, of peers, and of caring adults not in their household,
especially teachers and mentors who were able to serve as positive role models.

Sadly, assessment of developmental assets consistently finds that most
adolescents have only small number of such supports. In a study comparing male
gang members to community based organization members, Taylor et a1. (2002) found
that all youth had individual and contextual assets that could be used to promote
positive behavior and development, though not nearly as many as would truly
promote optimal development. While wealthier communities tend to have more
assets, the difference is negligible. Decreases in sexual activity are noted with
increases in assets (Benson, 2003). It seems that not only do “some kids need more”
but “all kids need more” (Benson, 2003). Concerns abound regarding the increased

chaos that seems to have infiltrated families, schools, and communities undermining

21

the stability that adolescents need for psychological growth (Benson, 2003). If youth
can be exposed to asset-building people and environments within multiple contexts
(developmental redundancy) and more assets can be provided to all youth
(developmental reach and breadth) gains in positive youth development would be
noted (Benson, 2003).

Individual adults willing to develop relationships with youth have the
potential to have a positive impact. Positive peer groups can also be influential.
Families, neighborhoods, schools and other organizations must work at positive
socialization of youth. Even community policy, social norms, and special programs
can provide needed assets to community adolescents (Benson, 2003). To have the
most impact a critical mass of individuals and institutions needs to be focused on
addressing the developmental needs of all community youth.

Thus, some of the factors that influence adolescent sexual activity may be
open to intervention, though many are not. With this in mind, many intervention
programs have sought to provide adolescents with the information they need to make
informed decisions about their sexual activity while others have targeted broader
areas of concern, still hoping for improved outcomes.

Interventions

Many approaches have been tried to minimize the consequences of adolescent
sexual activity and provide youth with assets that might help them avoid high risk
behaviors. An overview of these programs helps put this challenge in perspective.
Programs varied from those with the goal of decreasing risking sexual behavior, to

those aiming to reduce teen pregnancy, to others specifically promoting abstinence.

22

The CDC reviewed “Programs-That-Wor ”. The programs that were found to
reduce HIV, other STDs and unintended pregnancy included Reducing the Risk; Get
Real About AIDS; and Safer Choices for high school students and Be Proud, Be
Responsible; Becoming a Responsible Teen; Focus on Kids; Making a Difference;
and Making Proud Choices for youth in community settings. While these programs
met strict criteria for having a control group with at least a 4-week follow-up and
published data, they do not necessarily promote abstinence, just risk reductions
(Collins, Robin, & Wooley, 2002).

Most Americans support this comprehensive approach to sex education
believing schools should teach both abstinence and give teens enough information to
help them try to prevent unplanned pregnancies and STDs if they do decide to have
sex instead of teaching only abstinence until marriage (Kaiser Family Foundation,
2003). Landry, Kaeser, and Cory (1999) express concern that one in three school
districts surveyed did not allow any positive information about contraception to be
provided to students. But, the results of a randomized controlled trial by Coyle et al.
(2001) testing the effects of Safer Choices, one of the CDC’s “Programs-That-Work”,
found that sexually active subjects were more likely to use condoms and had fewer
partners but the subjects who had not previously been sexually active did not
significantly delay the onset of sexual intercourse compared to those in the control
group. While risk reduction is good, it is not 100% effective and still leaves teens at
risk. The Physicians Consortium (2002) also reviewed these curricula and expressed

concern that the claim stating they promote abstinence is deceptive and could

23

 

 

 

und

3C0.

fou
CV11

for

prog
addi
educ
patic
and l
prov
infor.
press
pallet
on-on

andc

 

 

undermine the public’s idea of what real abstinence education should be while
actually promoting “safe-sex”, which is not necessarily safe.

Kirby (2001a) also looked at programs to reduce teen pregnancy. This review
found that the Children’s Aid Society-Camera Program provided the strongest
evidence for impacting teen pregnancy while the Teen Outreach Program and Reach
for Health Community Youth Service Learning were also effective. Kirby (2001a)
suggests that professionals working with teens should continue to look at different
ways to prevent teen pregnancy by building on the successful elements of different
programs and exploring innovative approaches. Kirby (2002b) later found three
additional types of programs that were effective. They included certain sex and HIV
education curricula with specific characteristics, one-on-one prescribed clinician-
patient interactions in health settings, and service learning programs. Most of the sex
and HIV education programs emphasized abstinence as the safest choice but also
provided information on condoms and other contraceptives. They provided
information on the risks of teen sexual activity and activities addressing social
pressures that can influence sexual behavior as well (Kirby, 2002b). The clinician-
patient interventions were actually fairly brief but included the clinician having a one-
on-one interaction with the patient giving a clear message about appropriate sexual
and contraceptive behavior specific to the client (Kirby, 2002b). The success of the
service learning programs may be attributed to increased feelings of autonomy and
competence gained from the experience or the programs may have simply kept youth

busy allowing less time for high-risk behaviors (Kirby, 2002b).

24

Another evaluation of the Carrera program found that it was very effective for
female students who, after participation in the program, had significantly lower odds
of being sexually active than those not in the program (Philliber, Kaye, Herrling, &
West, 2002). It was not as effective with the male students, however. This may be due
to the fact that males often become sexually active at younger ages than females so
the male students may have already been sexually experienced and interventions have
been found to work better with students who were not yet sexually active at the
beginning of a program.

While these programs strive to decrease the consequences of adolescent
sexual activity such as pregnancy and STDs they do not have the goal of promoting
abstinence as their primary objective. Following are programs more specific to the
promotion of adolescent sexual abstinence.

Overall, adolescents who make virginity pledges are much less likely to have
intercourse than adolescents who do not pledge, but not necessarily in all
circumstances (Bearman & Bruckner, 2001). Even so, the delay effect was found to
be substantial. Pledging decreased the risk of intercourse substantially with pledgers
becoming sexually active on average 27 months later than non-pledgers (Diggs et al.,
2001). Diggs et a1. (2001) and The Physicians Consortium (2003) believe the success
of pledging is due to the way the pledge acknowledges that sexual activity can be
controlled, puts the locus of control on the individual, and makes the youth choose
purposefully to abstain, though there is really no way to control for self-selection bias
in youth choosing to pledge. And, while Koshar (2001) and Mohn, Tingle, and Finger

(2003) feel the impressive decline in teen pregnancy in recent years is due mainly to

25

teens choosing not to have sex, many youth are still engaging in sexual activity and
the negative consequences continue to be a huge societal problem.

Rector (2002) looked specifically at abstinence programs that do not provide
contraceptives or encourage their use and found them to be effective in reducing early
sexual activity. Virginity pledge programs were found to dramatically reduce sexual
activity in junior and senior high school students participating in abstinence programs
and pledging to remain abstinent. Rector (2002) also found Not Me, Not Now and
Operation Keepsake both dropped sexual activity rates among participants compared
to control school youth as did Abstinence by Choice and Teen Aid and Sex Respect.
Other successful programs included Family Accountability Communicating Teens
Sexuality, Postponing Sexual Involvement, Project Taking Charge, and Teen Aid
Family Life Education Project (Rector, 2002). Promoting true abstinence education
may be essential to reducing childbearing outside of marriage, preventing STDs, and
improving emotional and physical well being for youth. It can help them to develop
an understanding of commitment, fidelity, and intimacy that serves as the foundations
of healthy marital life in the future (Rector, 2002).

One mass media campaign to prevent teen pregnancy included a component
stressing abstinence. It found statistically significant changes toward abstinence with
the percentage of students who self-reported having sex by the time they reached 15
dropping from 46.6% to 31.6% (Doniger, Adams, Utter, & Riley, 2001). It also found
a decrease in the teen pregnancy rate from 63.4% to 49.5% showing that the change
in self-reported sexual activity was not likely due just to social pressure to deny being

sexually active (Doniger et al., 2001).

26

 

 

lm‘t
groi
b0)"
mor
cour

10“:

(km
facili
panii
inten

1110111.

used a
inlen‘

male-

dlSlIlt

 

 

A school-based intervention for urban youth used the Postponing Sexual
Involvement Curriculum (Aarons et al., 2000). This study found that intervention
group females were more likely to report virginity, self-efficacy to refuse sex with a
boyfriend, and the intention to avoid sexual involvement during the following 6
months. Intervention group males scored significantly higher than their control-group
counterparts in knowledge of birth control method efficacy but no change in attitudes
toward abstinence was observed.

Another program compared safer-sex education with abstinence education
(Jemmott, Jemmott, & Fong, 1998). This program used eight l-hr modules with adult
facilitators or peer co-facilitators and found that the abstinence intervention
participants were less likely to report having sexual intercourse in the 3 months after
intervention than the control group, but the difference did not continue to the 6 and 12
month follow up.

Following youth for 6-12 months, Aten, Siegle, Enaharo, and Auinger (2002)
used a control and 3 intervention groups of middle school students to compare no
intervention, ethnically diverse male-female pairs of adult professional educators,
male-female pairs of extensively trained high school peer educators, and school
district health teachers. The intervention was effective for regular teacher taught
students and peer-taught males and for those who were not already sexually
experienced at the beginning of the study.

Lieberman, Gray, Wier, F iorentino, and Maloney (2000) found a small-group

abstinence-based intervention somewhat beneficial. Focusing on mental health had

27

some impact on adolescents’ attitudes and relationships but was more helpful for
teens who were not already sexually active.

In addition to simply providing health instruction, adding a service learning
intervention with community involvement has been shown to have the long-term
benefit of reducing sexual risk taking among urban adolescents (O’Donnell et al.,
2002). Kirby (2002b) also found that service learning interventions decreased teen
pregnancy. This suggests that interventions focused beyond the basic microsystem
level can also be beneficial and that increasing developmental assets does have a
positive impact.

These programs provide a sample of the variety of interventions that have
been attempted, some working better than others. Many programs have focused on
sex-education, either abstinence only or abstinence plus information on safer sex.
While some studies have shown that this is indeed helpful and that adolescents can
make rational choices based on the potential consequences (Altman-Palm &
Tremblay, 1998; Blinn-Pike, 1999; Goldfarb, Duncan, Young, Eadie, & Castiglia,
1999) other studies indicate that just providing information may impact attitude but
this does not necessarily lead to a change in behavior (Arnold, Smith, Harrison, &
Springer, 1999; Kirby, Korpi, Barth, & Cagampang, 1997) or these studies did not
measure the impact on behavior (Agha, 2002; Spear, Young, & Denny, 1997). Many
different programs have shown encouraging outcomes, while some have had more
mixed results. Some of the successful programs have included a pledge to abstain
from sex and programs geared toward youth who were not already sexually active

have had better outcomes. Other programs, which may or may not have strong sex

28

 

 

beh
bah.-
undc

inter

decis
their .
bad a
at ext:
min.

This \

 

Choir

0051:,

31111:.

 

education components, but have stronger community involvement or service learning,
have demonstrated some positive outcomes. One common aspect of these programs is
their increased intensity with program leaders committed to a substantial investment
of time with youth. Yet is there more to promoting abstinence than just the
intervention?

Specific Processes and Contexts of Adolescent Sexual Activity

The developmental stages through which youth progress may also impact
behavior. While environmental influences may account for much of adolescent
behavior and interventions can certainly contribute to more positive outcomes,
understanding developmental processes and providing developmentally appropriate
interventions may be helpful.

Youth cognitive interpretation and perception of risk factors.

Ajzen and Fishbein (1980) see humans as capable of making rational
decisions based on the information available to them. People’s intentions are based on
their attitude toward the behavior or their judgment of whether they think it is good or
bad as well as their perceptions of social pressures related to the behavior. They look
at external variables or environmental impacts only in how they might affect people’s
attitudes about the behavior or their perceptions of social acceptance of the behavior.
This view can serve as a model of the cognitive processes underlying adolescent
choices about sexual activity (Gillmore et al., 2002). Adolescents should have the
cognitive ability to formulate rational behavioral intentions based on perceived
attitudes about the risks and benefits associated with engaging in sexual activity.

Hutchinson (1999) found that youth who were consistent in using condoms and who

29

0111
maj
Sun
ado]

beha

 

impa
em'ir
percc

becor

Ulcms

bfihm

 

this pj

 

attii J

 

did not think their partners were infected with any STD’s did not feel at risk of
contracting an STD. This is an example of how adolescents can perceive risk
logically (though it is unclear just how protective condoms really are against potential
STD’s and especially uncertain how accurate teens’ perceptions are that their partners
do not have an STD).

This approach argues that youth consider both the benefits and costs
associated with engaging in sexual activity, and are likely to avoid it if the costs far
outweigh the benefits (Mullan, Duncan, & Boisjoly, 2002). Measures of past behavior
may be used to predict intentions and can impact the affects of attitude. A study by
Sutton, McVey, and Glanz (1999) found that past behavior was predictive of
adolescent intentions and attenuated the affects of attitude and subjective norms on
behavior. Even so, they concluded that these beliefs about costs and benefits may be
impacted by information-based intervention programs. Hulton (2001) found that
environmental factors such as parental influences as well as social influences and
perception of benefits were factors related to the decision that adolescents make to
become sexually active or to abstain from sexual activity.

Donnelly, Ebume, and Eadie (1999) believe that youth have a tendency to see
themselves as invulnerable to the potentially negative consequences of high-risk
behavior. They propose educating students about risk management to help overcome
this perception. This is supported by Anderson et a1. (2001) who found that sexually
active youth in juvenile detention facilities did not think their sexual behavior could
result in their contracting a sexually transmitted disease. Another study of adolescent

females with risk factors showed that they did not perceive themselves to be at risk

30

 

 

 

IlOl

Chl

stud

inert
ofar.
hare
berrel
engaé
Show
Damn ‘
fOund

01‘th-

 

either (Morrison-Beedy et al., 2003) while Porter et al. (1999) found that increases in

the frequency of intimate behaviors and an increase in time alone and opportunity for
sexual activity had more of an impact on progression to sexual activity than any shifts
in perceived costs and benefits.

Awareness of the risks associated with high-risk sexual behavior alone does
not seem to be enough to change the behavior of many adolescents, though. Cohen
and Bruce (1997) found no correlation between perceived probability of HIV or
Chlamydia infection and frequency of risk-reducing behaviors in a group of college
students. Boyer et a1. (2000) found that youth were able to judge their increased level
of STD risk as their sexual risk taking increased. While they may have felt at
increased risk the lack of a negative outcome or their feelings that the consequences
of an STD or pregnancy would not have a long term negative effect on them may
have contributed to their continuing in the behavior. Perhaps they just do not see the
benefits of abstaining from sexual activity as outweighing the perceived benefits of
engaging in sexual activity. A study of sexually active female college students
showed that less than half used condoms for protection, regardless of the type of
partner and the level of risk of STDs or HIV (Williams et al., 2003). Another study
found that college students’ sexual risk behaviors were tied more to their perceptions
of the benefits of unprotected sex (Parsons et al., 2000). These adolescents saw the
benefits of unprotected sex as greater than the costs or possible negative results.
Based on another study Parsons et a1. (1997) also see the youth’s perception of the
benefits having a greater impact on risk-taking behavior than their view of the

potential costs. Nangle and Hansen (1998) believe cognitive behavior skills

31

interventions that directly teach adolescents new skills are an important component to
impacting behavior.

The adolescent personal fable.

Since looking exclusively at the cognitive processes underlying adolescent
choices about sexual activity helps little in understanding adolescent behavior, the
adolescent personal fable, first described by Elkind (1979), is a concept that could
help better explain why youth may choose to engage in risky behavior even when
aware of the potential consequences. Most adolescents have the knowledge to
perceive risk correctly; yet do not integrate these risks in their decision-making.
Personal fable is characterized by the incapacity of youth to see themselves as similar
to others, so youth feel unique and self-focused believing that the risks that apply to
others do not apply to them (Greene, Rubin, & Kromar, 2002). When youth believe
the natural laws that apply to others do not affect them they may have feelings of
immortality. Egocentric adolescents cannot believe that anything bad can happen to
them (Buis & Thompson, 1989). Frankenberger (2000) looked at this personal fable
and found that in addition to adolescents it also seemed to extend into early
adulthood. The belief that adolescents and even young adults may have that the bad
things that happen to others won’t happen them should also be considered when
looking at all the ecological factors that can impact behavior (Knoppers, 2003a).
While this concept is not new, longitudinal studies on the topic are lacking, making it
unclear if youth stating after the fact that “I didn’t think it could happen to me” held

this belief initially or only in a more retrospective manner (V artanian, 2000).

32

Traditionally, changes in cognitive development as youth move from
childhood into adolescence have been used to explain the personal fable. More
recently the personal fable has been viewed in relation to socio-cognitive
development and the work of separation (Goosens, Beyers, Emmen, & van Aken,
2002). The related concept of willingness to take risks has also been viewed as a
developmental phenomenon, possibility related to an inability to assess the extent of
risk in a given situation. This developmental immaturity or lack of experience may
cause an adolescent to misjudge risk (Green, Kromar, & Walters, 2000). If youth
don’t perceive a risk it is more a matter of lack of recognition than an error in
judgment. This may help explain how adolescents seem to ignore messages meant to
help them make healthier choices (Green, Kromar & Walters, 2000). The personal
fable fits in this perspective with its emphasis on youth’s uniqueness and
invulnerability, which is believed to peak in the late junior high and early senior high
years. Green, Kromar, and Walters (2000) see the personal fable and more
particularly feelings of invulnerability as negatively associated with adolescent’s
ability to perceive risks as well as their intentions to avoid the risky behaviors and
found the highest levels of sexual risk-taking were reported in those with high
feelings of invulnerability and sensation seeking.

It is unclear, however, what factors contribute to youth having or not having
strong feelings of invulnerability. If youth feel invulnerable to the negative
consequences of sexual activity they are more likely to feel the benefits of engaging
in sexual activity outweigh the potentially negative consequences that they do not

think will happen to them. While extensive research has been done on many aspects

33

of adolescent sexual activity, little has been done on this topic. A review of the
literature searching CINALH, Education Abs, MEDLINE, ERIC, Soc Abs, Social
Science Abs, and Wilson Select Plus showed very little in regards to adolescent
invulnerability related to sexual abstinence or sexual activity other than those
mentioned above.

Does this personal fable prevent adolescents from benefiting from the
education most receive and inhibit them fi'om making rational choices based on the
actual likelihood of negative consequences? Chapin’s (2001) findings that high-risk
African American youth perceived themselves to be less likely than their peers to
become pregnant or cause a pregnancy would support this. Klaczynski and Fauth
(1996) also found that adolescents believe that they are less likely to suffer unwanted
effects than their peers, and may also have strong feelings of being physically and
socially indestructible. Jack (1989) found that repeated comments from youth such as
“I didn’t think it would happen to me” support the idea that they really do feel
invulnerable. A better understanding of these feelings of invulnerability could help
guide those trying to decrease adolescents’ sexual risk-taking (Jack, 1989). While
risk-taking and experimentation during adolescence may help them achieve a sense of
independence and self-identity, finding a way to balance this against the potential
consequences is important.

Yet, what is the influence of perceived invulnerability in relation to other
known and measured risk or protective factors for adolescent sexual activity? If
feelings of invulnerability to the potential negative consequences of sexual activity do

influence adolescents’ intentions, are there interventions that could minimize this

34

effect, helping teens realize their vulnerability so they might make wiser decisions
about their sexual activity? Out and Lafi'eniere (2001) found that after 2-3 days of
caring for a Baby Think It Over teens were better able to accurately assess their
personal risk for an unplanned pregnancy than teens in the comparison groups, but the
study did not follow up to see if this actually resulted in fewer teen pregnancies.
Interactive computer games may be another way to help youth realize their
vulnerability. Virtual reality simulations could allow youth to safely explore the
possibilities and even experience the virtual consequences of their behavior
(Knoppers, 2003a). While early versions of computer programs showed positive
behavior changes and recent anecdotal evidence is encouraging, this type of
intervention warrants further investigation (Knoppers, 2003a).

Risk and protective factors.

Perkins et al. (1998) examined the following 12 risk factors of adolescents’
sexual activity: age, alcohol use, physical abuse, sexual abuse, GPA, suicidal
ideation, religiosity, parental monitoring, family support, time spent home alone,
membership in negative peer group, and perception of school climate. Those most
highly correlated with adolescent sexual activity were alcohol use, sexual abuse,
GPA, and membership in negative peer group. While Perkins et al. (1998) did not
find parental monitoring (in this study, measured only by a single item) or family
support predictive of adolescent sexual activity other studies have (Blum & Rinehard,
1997; Luster & Small, 1994; Wu et al., 2003). The results (Perkins et al., 1998) for
religiosity were mixed, varying by gender and ethnicity with low religiosity not

correlating for African American males nor Latina females. Perkins et al. (1998) also

35

had mixed results for perception of school climate, which correlated for males, but
not females. Blum and Rinehard (1997) found both perception of school climate and
religiosity to correlate to adolescent sexual activity, though this study did not separate
youth by ethnicity nor gender.

Mullan et al. (2002) as well as Talashek et a1. (2003) looked at additional
factors to be considered such as socio-economic status and family structure while
Kalil and Kunz (1999) also included family size, minority status, and maternal
education, as did Small and Luster (1994) for parental education. Perceived parent
disapproval of sexual activity and parent/adolescent activities were included by Blum
and Rinehard (1997) as were perceived risk of untimely death and school
connectedness.

Many of the factors impacting adolescent sexual activity can be viewed as risk
factors or conversely as protective factors. If low GPA is a risk factor, having a high
GPA is a protective factor. Fergusson and Horwood (2003) present the idea that to be
meaningful, protective factors should be something more than the opposite of risk
factors. Some posit that the effect of a protective factor would be negligible in lower
risk populations but would be intensified in the company of one or more risk
variables. Protective factors should have an impact only in the face of hardship. But
others have used protective factors to describe values associated with advantageous
outcomes free of the occurrence of social disadvantage or unfavorable conditions or
even the positive role of risk factors perhaps better called promotive factors

(Sameroff, Gutrnan, & Peck, 2003).

36

An interactive relationship of the protective factors, the risk exposure, and the
outcome may result in the protective factor having beneficial effects on those exposed
to the risk factors that might not benefit those not exposed to it (though this
framework does not always fit). Two types of processes may lead to resilience. These
are protective processes and compensatory processes (F ergusson & Horwood, 2003).
Intelligence and problem-solving abilities, external interests and affiliations, parental
attachment and bonding, and peer factors are examples of factors that may contribute
to resilience in children raised in high risk environments (Fergusson & Horwood,
2003). Resilience factors may produce an effect by compensating for childhood
adversity not necessarily by acting in a protective role (Fergusson & Horwood, 2003).
Seidman and Pedersen (2003) took a holistic and contextual approach in looking at
risk/protection and competence and found multiple forms of contextual competence
suggesting the need to engage youth positively with two or more settings.

While it is helpful to look at these factors it is challenging to try to pinpoint
which ones might have the most impact when there are so many environmental
elements. The Search Institute’s forty developmental assets (Benson, 1997) can help
in gaining an understanding of the bigger environmental picture. The external assets
grouped under support, empowerment, boundaries and expectations, and constructive
use of time focus on the environment while the internal assets grouped under
commitment to learning, positive values, social competencies, and positive identity
are based at the individual level.

In an ambitious review of over 250 articles Kirby (2002a) found more that 80

antecedents to initiation of sex. The majority of them were at the individual level yet

37

many were environmental or contextual factors. Trying to place these factors in a
developmental and ecological framework viewing adolescents within the complicated
ecology of their families, peers, schools, and neighborhoods remains a challenge
(Gorrnan-Smith, & Tolan, 2003). This is a critical step in trying to target interventions
instead of simply listing corollary factors. Lists of protective factors have limited
practical use because all itemized indicators can never be addressed in a given
intervention (Luthar & Zelazo 2003) but may still be an important step in promoting
positive youth development.

Individual factors include alcohol use, GPA, suicidal ideation, and religiosity,
though even these are also impacted by the environment. Additional factors identified
in Kirby’s (2002a) review include those under the groupings of biological;
race/ethnicity; healthy behaviors; other problem or risk-taking behaviors; attachment
to and success in school; working more then 20 hours a week; emotional well-being
and distress; and sexual beliefs, attitudes, skills, and behaviors. Masten and Powell
(2003) also saw individual differences including cognitive abilities, self-efficacy and
self-esteem, temperament, impulse control, and positive outlook on life as factors that
increase adolescents’ resiliency and ability to overcome adversity for better
outcomes.

Looking at the family microsystem parental monitoring, family support, time
spent home alone, family size and structure, parent’s education, and parental approval
or disapproval of adolescent sexual activity can all impact adolescent sexual behavior.
Kirby (2002a) also identified that having a working mother, being a younger sibling

and having older siblings who have been sexually active or gave birth as adolescents

38

were risk factors while higher income level, having health insurance, and the mother
being older at first sex and first birth were protective factors. Greater family
religiosity can also impact at this level in addition to the individual level (Kirby,
2002a). Factors such as physical and sexual abuse often occur in the family
microsystem as well, but may also occur in extra familial settings.

The peer microsystem also plays a part. Beal, Ausiello, and Perrin (2001)
looked at parent disapproval of health-risk behaviors, parent modeling of health-risk
behaviors, parent monitoring of health-risks, peer disproval of health risks, and peer
modeling of health-risk behaviors. They found that peer social influences were more
evident than parental influences in impacting sexual activity in 7th graders attending
an urban magnet middle school. Perception of peers’ sexual behavior can also
influence adolescent behavior, perhaps even more than the peers actual behavior
(Nahom, Wells, Gillmore, 2001). Also at the peer group level, having older friends,
having peers with lower grades who drink or have permissive attitudes toward
premarital sex, having sexually active peers, dating alone, going steady and having
older romantic partners were risk factors (Kirby, 2002a). A protective factor at this
level is having close friends who are close to their parents (Kirby, 2002a).

Another important microsystem is the school. Attending a parochial school
was a protective factor for youth while being popular with peers and engaging in
physical fights were risk factors (Kirby, 2002a). Being connected to school, though
more at the individual level, was a protective factor (Kirby, 2002a).

Larger societal issues at the macrosystem level like socioeconomic status and

being a minority are also important. Other community antecedents include higher

39

divorce rates, higher rates of residential turnover, and higher unemployment rates
(Kirby, 2002a). Kirby (2002a) also identified living in a community with a higher
percentage of college educated people, higher family income and greater
neighborhood monitoring by the adults as protective factors.

Yet, with this large amount of research, few studies look at the individual
factor of perceived invulnerability to the potential consequences of sexual activity
and none were found that used a longitudinal approach to assess its relationship to
initiation of sexual activity. Chapin (2001) found a correlation between high feelings
of perceived invulnerability and feelings of being less susceptible to the consequences
of sexual risk taking. Youth who consistently used condoms and believed that their
sexual partner was not infected with an STD perceived themselves to be at no risk of
contracting an STD (Hutchinson, 1999). Pete and DeSantis (1990) found that a belief
in their lack of vulnerability to become pregnant correlated with sexual activity in
young black adolescent females while Quandrel, Fishchhoff, and Davis (1993) found
that most adolescents perceived themselves to be at less risk than others but this study
did not look at how this impacted the adolescents’ intentions regarding sexual activity
or their behavior.

Looking at how these factors correlate to adolescent behavior regarding sexual
activity may give suggestions on helpful interventions to promote sexual abstinence.
Data on individual factors obtained using multiple questions should make it easier to
ascertain subtle differences in environmental influences but lengthy measures that
could increase the risk of subjects not completing the questionnaire need to be

avoided. It is also important that the questions use the highest level of measurement

4o

feasible to provide more useful data using ordinal or interval level data over nominal
level data whenever possible.
Conclusion

Numerous studies have shown that multiple factors impact adolescent
transition to sexual activity. These factors occur at the individual, family, and extra-
familial levels. In addition, a variety of programs to delay onset of adolescent sexual
activity have been tried. Programs ranging from abstinence only, to abstinence plus
information on safer sex, to programs promoting community involvement and service
learning have all shown some potential. Yet, with this great body of research, the
concept of feelings of invulnerability has rarely appeared. If feelings of
invulnerability impact adolescent sexual behavior, gaining an understanding of this
relationship could help in planning interventions. It is apparent no one factor can
account for adolescent sexual activity, but a combination of factors may give a clearer
picture. This study takes an ecological perspective to understanding adolescent sexual
activity looking specifically at an individual factor, the little explored concept of
perceived invulnerability, in relation to risk factors associated with significant

contexts of sexual activity and other individual factors.

41

Chapter 3
Methods

Pimpse of the Study

The purpose of this study was to answer these questions: 1) What is the
relationship between youths’ perceived invulnerability to pregnancy and STDs and
behavior regarding sexual activity? 2) What is the relationship between youths’
perceived invulnerability to pregnancy and STDs and behavior regarding sexual
activity when other predictors of sexual activity are considered (i.e. controlled)? 3)
How does perceived invulnerability impact adolescents with few risk factors and
more protective factors compared to those with multiple risk factors and few
protective factors? This study also adressed questions about how the individual risk
and protective factors for adolescent sexual activity as well as factors at the
mircrosystem and macrosystem levels relate to feelings of invulnerability.
Definition_s_ and Measures

This study used a tool with questions adapted fiom the Dane County Youth
Assessment 2000 (University of Wisconsin Board of Regents, 1999), the
invulnerability scale from Lapsley’s New Personal Fable Scale (Lapsley, FitzGerald,
Rice & Jackson, 1989) and additional questions developed by the researcher
providing basic demographic information as well as data on risk and protective
factors and feelings of invulnerability. (See Appendix A for specific questions from
the Youth Assessment Tool.)

General definitions and measures.

Youth: Youth is defined as students ages 12 to 19 years of age (question 3).

42

Perceived inwlnerabililtogreamcy and STDs: This is the youth’s belief
that he/she will not get pregnant (or impregnate partner) and the belief that he/she will
not contract one of many sexually transmitted diseases if he/she engages in
unprotected sexual activity as measured by self report on four items on the Youth
Assessment Tool (questions 15, 16, 25, and 26). Low scores on 15 and 16 (reverse
coded) and high scores on 25 and 26 indicate higher levels of perceived
invulnerability with total scores ranging from 0 for the lowest level of invulnerability
to 16 for the highest level of invulnerability. Cronbach’s alpha for this grouping of
questions was .75. With the low reliability of this four item scale, possibly related to
the reverse coding of two of the questions, the two sets of questions were analyzed
separately. Using a Spearman’s rho the first two questions had a .84 correlation
coefficient with an alpha of < .01 while the Spearrnan’s rho for the second two
measures had an r of .76 also with an alpha of < .01. The scores for the personal
fable vulnerability scale (47-60), some which were reverse coded, range from 0-56.
The Cronbach’s alpha for these questions (47-60) was .56. Afier performing a factor
analysis showing the five items with the highest loading on the first factor,
Cronbach’s alpha was run on only these factors (questions 51, 53, 57, 59, and 60 in
Appendix A) resulting in an alpha of .65.

Sexuafitctivity: Sexual activity is defined as sexual intercourse measured by
self-report (question 18) as well as frequency of sexual activity in past 3 months
(question 19), number of lifetime partners (question 20), and if drunk or high with

sexual activity (question 21). For question 18 scores range from 0 to 7 with higher

43

scores indicating earlier age of sexual initiation. For questions 19-21 individual scores
range from 0-6 with higher scores indicating higher risk behavior.

_S_T_D_s_: This is a self report of sexually transmitted disease history as measured
by the Youth Assessment Tool (question 22).

Birth control use: This is self report of use of birth control as measured by the
Youth Assessment Tool (question 23) with scores ranging from 0 for those who have
not been sexually active to 6 for those who are sexually active and have never used
birth control.

Pregnancy: Pregnancy is defined as self report of having been pregnant or
causing a pregnancy as measured by the Youth Assessment Tool (question 24).

Risk factors: Risk factors are characteristics of the youth that may have a
negative impact on behavior as measured by the Youth Assessment Tool. These
include low GPA, single parent home, low parental education, history of sexual or
physical abuse or witnessing physical abuse, parental approval for sexual activity and
alcohol consumptions, and negative peer group.

Protective factors: Protective factors are characteristics of the youth that may
have a positive impact on behavior as measured by the Youth Assessment Tool.
These include high GPA, living with both parents, high parental education, increased
religiosity, caring community, increased parental monitoring and support, and belief
that youth should not have sexual intercourse.

Specific risk and protective factors: definitions and measures.

Qflfi: GPA is the average grade the youth usually gets in school courses as

measured by self report (question 4). Scores on this question range from 0-7 with

44

higher scores indicating lower grades and higher risk while lower scores indicate
higher grades and lower risk.

f_amilv structure: Family structure is defined as who the youth lives with most
of the time and if parents are divorced or separated as measured by 2 questions on the
Youth Assessment Tool (questions 6 and 9).

P_arental education: This is how much education mother/stepmother and
father/stepfather completed as measured by 2 questions on the Youth Assessment
Tool (questions 7 and 8). Scores range from 0 to 6 and -7 with lower scores from O to
6 indicating more education and less risk combining for a maximum of 12 and a score
of -7 indicating the youth does not know the amount of education that parent had
causing exclusion from analysis for this item. Using a Spearman’s rho these two
measures had a .56 correlation coefficient (p < .01).

Substance use: Substance use is defined as use of tobacco, alcohol, marijuana,
and other drugs as measured by 5 questions on the Youth Assessment Tool (questions
10-14). Total scores range from 0 to 25 with higher scores indicating increased
substance use. This scale had a Cronbach’s alpha of .83.

Religious beliefs/‘mtivities: Religious beliefs/activities is defined as how
important religion/religious beliefs and religious activities are to the youth as
measured by 2 questions on the Youth Assessment Tool (questions 17 and 27 with
question 27 reverse coded). Scores range from 0 to 8 with higher scores indicating
decreased religiosity and higher risk. The Spearman’s rho was .68 for these two items

mmw<nu

45

Sexual abuse: This is report of sexual abuse as measured by the Youth
Assessment Tool (question 28).

Physical abuse: Physical abuse is defined as report of physical abuse as
measured by the Youth Assessment Tool (question 29).

Witnessing physical abuse: This is report of witnessing physical abuse as

 

measured by the Youth Assessment Tool (question 30).

Qaang Community: Caring community is defined as report of helping
neighbors, watchful adults and caring community as measured by 3 questions on the
Youth Assessment Tool (questions 31-33). These questions had total scores ranging
from O to 12 with lower scores showing stronger community and lower risk. The
Cronbach’s alpha for this scale was .71.

firentJal monitoring: Parental monitoring is report of parental monitoring of
youth activity as measured by 4 questions on the Youth Assessment Tool (questions
34-3 7, all reverse coded). Scores from 0 to 4 on each question after reversal could
result in a maximum of 16, indicating low parental monitoring and higher risk. The
Cronbach’s alpha for this scale was .89. Scores of -5 on any of the questions indicate
no adults at home and were not included in the calculated scores, excluding them
from analysis.

Parental support: Parental support is defined as perception by youth that
parents are there for them and care as measured by 2 questions on the Youth
Assessment Tool (questions 38 and 39, reverse coded). Scores from 0 to 4 on each
question could result in a maximum of 8, indicating low parental support and higher

risk, resulting in a Spearman’s rho of .66 with an alpha of < .01. Scores of -5 on

46

either question indicate no adults at home and were not included in the calculated
scores.

Parental approval of adolescent sexrgrj activifi and alcohol consumption: This
is defined as youth report of parents approval of teen sexual activity and alcohol
consumption as measured by two questions on the Youth Assessment Tool (questions
40 and 41). Scores range from 0 to 4 on each question with higher scores indicating
higher parental approval for these behaviors and higher risk.

Peer group: Peer group is report of friends getting youth into trouble, sexual
activity, drinking or drug use, and tobacco use as measured by 4 questions on the
Youth Assessment Tool (questions 42-45). Scores from 0 to 3 on each question could
result in a maximum of 12 with low scores for positive peer group and higher scores
with more negative peer group and higher risk. These items had a Cronbach’s alpha
of .90.

Youth’s sexualbeliefs: This is defined as youth’s belief that teens should not
have sexual intercourse as measured by the Youth Assessment Tool (question 46).
Scores range from 0 to 3 with lower scores indicating belief that teens should not
have sexual intercourse.

Rigarch Dem

The study was a cross-sectional survey, designed to examine the relationship
between perceived invulnerability to pregnancy and STDs and sexual activity. The
study also addressed perceived invulnerability in relation to other risk and protective

factors.

47

The Sample

The target population was youth involved with programming through the Kent
County Coalition on Adolescent Choices in Health (COACH) and Wedgwood
Christian Services. The sample was drawn mainly from youth participating in Life
Skills programs in Kent County, Michigan. Participants were all able to read and
write English. Subjects consisted of all youth attending Life Skills programs willing
to participate with parental consent. This resulted in a convenience sample of youth
from COACH programs in Kent County. While this decreases the generalizability of
the study, the groups included youth from different socio-economic and racial groups
making the findings more useful.

Data was collected during the regular Life Skills group time.
Parental/guardian consent was obtained before data collection (see Appendix B). The
group facilitator for each group explained the study and the need for parental consent
to each youth when handing out the consent forms. All willing youth who had
returned parental consent participated in the survey. Youth aged 18 and 19 were able
to sign their own consent forms (see Appendix C).

The socio-demographic make up of youth participating in COACH programs
is quite varied. The socio-economic and racial make up of youth attending past Life
Skills programs depended on the location of the group. Some sites are mainly low
income Afiican American, some highly Hispanic, and others with a higher proportion
of white students. COACH programming targets high-risk youth so usually includes a

high percentage of low-income youth, though some middle class youth also attend.

48

Using a standard alpha of .05 for a one-sided test with a beta of .2 for 80%
power, the total sample size required for a correlation coefficient with a modest effect
size of .3 is 67. Realizing that many factors in addition to perceived invulnerability
may contribute to an adolescent’s decisions about sexual activity, the larger the
sample the better chance of being able to detect even a small correlation between
perceived invulnerability and behavior regarding sexual activity. While youth may
not enjoy the time the data collection takes, the time was scheduled so they did not
miss any other group activities while filling out the survey.

Method of Data Collection

The investigator used a tool adapted from the Dane County Youth Assessment
2000 Survey (University of Wisconsin Board of Regents, 1999), along with the
invulnerability subscale of Lapsley’s New Personal Fable Scale (Lapsley, F itzGerald,
Rice, & Jackson, 1989), and additional questions developed by the researcher (see
Appendix A). The Dane County Youth Assessment 2000 survey includes basic
demographic information, questions on grades, living situation, substance use,
personal issues (including sexual activity), opinions about the community, their
family, and other views and opinions. The invulnerability subscale of Lapsley’s New
Personal Fable Scale focuses on the vulnerability piece of the personal fable concept
with some of the questions reverse coded.

Reliability and validity of measures.

Due to the nature of the survey tool, assessing reliability is challenging.
Because the factors being measured are likely to change over time, testing tool

stability would not be useful. Test equivalence is not feasible as no other tools were

49

found that test this range of risk and protective factors along with invulnerability, and
developing another tool for alternate form testing would be too time intensive for the
youth being asked to fill out two similar surveys. The 14 invulnerability subscale
items fiom the New Personal Fable Scale (Lapsley, F itzGerald, Rice, & Jackson,
1989) had a reliability of .79 for Greene, Kromar, and Walters (2000), though a
review of the literature was unsuccessful in finding additional reports for this scale.
For measures with multiple questions, Cronbach’s alpha was obtained.

Content validity of the tool was determined by a review of the literature. This
was also affirmed by researchers familiar with this field of study. Reviewers provided
feedback and suggestions about inclusion and exclusion of topics as well as overall
ease of use of the tool.

Study implementation.

Human subject application was submitted by the researcher to MSU Human
Subjects Committee, UCRIHS, and approval was obtained. The researcher submitted
appropriate documentation to the research cdmmittee at Wedgwood which reviewed
the study proposal and granted permission for data collection. Wedgwood is the
fiduciary agency for the Coalition on Adolescent Choices and Health (COACH) in
Kent County, which hosts the Life Skills programs and other groups. COACH is a
community coalition made up of concerned citizens and representatives from public
and private agencies involved with promoting positive youth development. The main
goal of COACH is to help youth make healthy choices regarding abstinence from

sexual activity, tobacco, alcohol, and other drugs. Wedgwood is supportive of

50

research that would help provide insight into these issues and was eager to help
facilitate this process.

The researcher copied all materials and provided information to group
facilitators on obtaining informed consent and data collection. Data collection began
fall of 2004 and was completed mid summer 2005.

Hymtheses

This study tested many hypotheses. Hypotheses addressing the main research
questions related to specific sexual behavior and the potential relationship of feelings
of invulnerability to that behavior. Keeping with an ecological focus, hypotheses
addressing individual as well as microsystem and macrosystem level risk and
protective factors and how they related to feelings of invulnerability were also
included.

Hypotheses relating to specific sexual behavior.

#1) Youth with higher perceived invulnerability will be more likely to be

sexually active.

#2) Youth with higher perceived invulnerability will have greater frequency of

sexual activity.

#3) Youth with higher perceived invulnerability will have a greater number of

lifetime partners.

#4) Youth with higher perceived invulnerability will more frequently have sex

while drunk or high.

#5) Youth with higher perceived invulnerability will have greater use of birth

control.

51

#6) Youth who have ever had an STD will have lower levels of perceived
invulnerability.

#7) Youth who have ever been pregnant or gotten someone pregnant will have
lower levels of perceived invulnerability.

#8) Youth with higher perceived invulnerability will be more likely to be
sexually active when other predictors of sexual activity are controlled.

#9) Perceived invulnerability will have less impact on the sexual behavior of
youth with few risk factors and more protective factors while having a greater
impact on those with multiple risk factors and fewer protective factors.
Hypotheses relating to individual risk factors.

#10) Younger youth will have higher levels of invulnerability.

#11) Youth who use alcohol or drugs will have higher levels of
invulnerability.

#12) Youth with a history of sexual abuse will have lower levels of
invulnerability.

#13) Youth with a history of physical abuse will have lower levels of
invulnerability.

#14) Youth who have witnessed physical abuse will have lower levels of
invulnerability.

#15) Males will have higher levels of invulnerability

#16) Minority youth will have higher levels of invulnerability.

#17) Youth with higher GPAs will have lower levels of invulnerability.

52

#18) Youth with strong religious beliefs will have lower levels of
invulnerability.

#19) Youth who believe that teens should not have sexual intercourse will
have lower levels of invulnerability.

Hypotheses relating to microsystem factors.

#20) Youth who live with both parents will have lower levels of
invulnerability.

#21) Youth whose parents have never been divorced or separated will have
lower levels of invulnerability.

#22) Youth with higher levels of parental monitoring will have lower levels of
invulnerability.

#23) Youth with higher levels of parental support will have lower levels of
invulnerability.

#24) Youth who have parents with higher levels of education will have lower
levels of invulnerability.

#25) Youth with parents who are more approving of adolescent sexual activity
will have higher levels of invulnerability.

#26) Youth with parents who are more approving of adolescent alcohol
consumption will have higher levels of invulnerability.

#27) Youth who report having friends who get into trouble, are sexually

active, drink or use drugs, or smoke will have higher levels of invulnerability.

53

Hypothesis relating to macrosystem/community factors.

#28) Youth reporting supportive communities will have lower levels of

invulnerability.
D_ata_AAalysis for Hypotheses

Frequencies were run on basic demographic information. For the research
question: “What is the relationship between youths’ perceived invulnerability to
pregnancy and STDs and sexual activity?” and the hypotheses that youth with higher
perceived invulnerability will be more likely to be sexually active, have greater
frequency of sexual activity, have a greater number of lifetime partners, have greater
use of birth control, and will more frequently have sex while drunk or high a
Spearman’s rho was used to look at the correlation between feelings of invulnerability
and sexual activity, frequency of sexual activity, number of lifetime partners, use of
birth control and frequency of sex while drunk or high for the study group. For the
hypotheses that youth who have ever had an STD and who have ever been or gotten
someone pregnant will have lower levels of perceived invulnerability an independent
samples t test were run.

For the second research question: “What is the relationship between youths’
perceived invulnerability to pregnancy and STDs and sexual activity when other
predictors of sexual activity are controlled?” and the hypothesis that youth with
higher perceived invulnerability will be more likely to be sexually active when other
predictors of sexual activity are controlled logistic regression was used. For the
hypothesis that perceived invulnerability will have less impact on adolescents with

few risk factors and more protective factors while having a greater impact on those

54

with multiple risk factors and few protective factors logistic regression was also be
used.

For variables measured using interval scales it was hoped a Pearson Product
Moment Correlation Coefficient (r) could be used to test the research questions about
how these individual risk and protective factors for adolescent sexual activity as well
as how factors at the mircrosystem and macrosystem levels relate to feelings of
invulnerability. The assumptions of a representative sample with a normal
distribution and a linear relationship between the variables were met for only the
hypothesis that youth reporting supportive communities will have lower levels of
invulnerability, however, so a Spearman’s rho was used for the remaining hypotheses
(e.g. hypotheses # 11, # 18, # 22, # 23, and # 27).

Hypotheses with variables that were not measured using an interval scale were
also tested using a Spearman’s rho (i.e. # 10, # 17, # 19, #24, # 25, and # 26) while
those hypotheses that had nominal level variable were tested using an independent
samples t test for equality of means. (Hypotheses # 12, # 13, # 14, # 15, # l6, # 20,
and # 21.)

Protection of Human Subjects

Human Subject approval was sought and obtained from MSU and Wedgwood.
Subject confidentiality was maintained and no individual subjects were identified in
the findings. Parental/guardian and participant consent were obtained for minors;
participant consent only was obtained for those 18 and 19 years of age. (See

Appendix B and C for copies of parental and participant consent forms.)

55

Limitations of the Study

In looking at the relationship between feelings of invulnerability and sexual
activity a limitation of this study is that it cannot establish causality. A longitudinal
study would better look at the relationship between perceived invulnerability as a
predictor of subsequent sexual activity. If a relationship is found, future longitudinal
studies could be designed that would be better able to determine if youth who feel
invulnerable to the potential negative consequences of sexual activity are more likely
to become sexually active.

Another limitation of the study is the use of a convenience sample. While it is
hoped that the sample will include youth from a variety of socio-economic and racial
groups, it is unlikely that the sample will end up with a proportion of youth from
these various groups that is similar to that of society. A concern also exists that youth
who are involved in COACH and Wedgwood programs may be somehow different
than youth who do not participate, and thus would not be included in the sample.
Even within the pool of potential participants from COACH and Wedgwood
programs, those who agree to participate and whose parents give permission may also
be different than those who do not want to participate or whose parents do not give
permission either because they are difficult to reach or simply do not want their child

to fill out the survey.

56

Surve 3

Chapter 4

Results

Signed consent forms for 75 youth were returned and 70 youth filled out

surveys. (Five of the youth were absent the day the surveys were completed.) Of the

70 surveys filled out by youth, 3 of them had two or more pages incomplete so were

excluded from the analysis resulting in 67 usable surveys.

Descriptive Statistics

Sixty-five point seven percent (44) females and 34.3% (23) males completed

usable surveys. Age of participants varied from 12 to 19. (See Table 1.) Results for

race/ethnicity are found in Table 2.

Table 1- Youth Age

 

 

 

 

 

 

 

 

 

 

 

Cumulative
Frequency Percent Valid Percent Percent
Valid 12 or younger 6 9.0 9.0 9.0
13 7 10.4 10.4 19.4
14 11 16.4 16.4 35.8
15 l 1 16.4 16.4 52.2
16 6 9.0 9.0 61.2
17 9 13.4 13.4 74.6
18 15 22.4 22.4 97.0
19 or older 2 3.0 3.0 100.0
Total 67 100.0 100.0
Table 2- Youth Race/ethnicity
Cumulative
Frequency Percent Valid Percent Percent
Valid Black/African-American 4 6.0 6.0 6.0
Hispanic 6 9.0 9.0 14.9
White (not Hispanic) 49 73.1 73.1 88.1
Mixed race 8 11.9 11.9 100.0
Total 67 100.0 100.0

 

 

 

 

 

 

57

 

Average grades for respondents varied from getting mostly As to getting

mostly Ds. (See Table 3.) Most youth lived in the Grand Rapids metropolitan area.

(See Table 4.). The majority of youth lived with two parents with living with mom

and step dad the second most common response. (See Table 5.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 3- Average Grades
Frequency Percent Valid Percent Cumulative Percent
Valid Mostly As 18 26.9 26.9 26.9
About half As & half BS 13 19.4 19.4 46.3
Mostly Bs 9 13.4 13.4 59.7
About half Bs & half Cs 12 17.9 17.9 77.6
Mostly Cs 9 13.4 13.4 91.0
About half Cs & half BS 4 6.0 6.0 97.0
Mostly Ds 1 1.5 1.5 98.5
Mostly below D 1 1.5 . 1.5 100.0
Total 67 100.0 100.0
Table 4- Where Youth Live
Valid Cumulative
Frequency Percent Percent Percent
Valid Grand Rapids metro 38 56.7 56.7 56.7
3323' Q'zlg‘fifmd’ 25 37.3 37.3 94.0
22$]?ng ggl‘fifllgafsgfig) 3 4.5 4.5 98.5
In country not on farm 1 1.5 1.5 100.0
Total 67 100.0 100.0
Table 5- With Whom Youth Lives
Frequency Percent Valid Percent Cumulative Percent
Valid Two parents 34 50.7 50.7 50.7
Mom and step dad 12 17.9 17.9 68.7
Dad and step mom 1 1.5 1.5 70.1
Mom only 10 14.9 14.9 85.1
Dad only 4 6.0 6.0 91.0
Group or foster home 3 4.5 4.5 95.5
Other relative 1 1.5 1.5 97.0
Alone or with friends 2 3.0 3.0 100.0
Total 67 100.0 100.0

 

 

 

 

 

 

58

 

 

 

The majority of moms had at least a high school education (see Table 6) as

did the majority for dads (see Table 7). Table 8 shows results for if parents had ever

divorced or separated.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 6- Morn's Education
Valid Cumulative
Frequency Percent Percent Percent
Valid Professional/graduate degree 7 10.4 11.3 11.3
Some school beyond 4 year college 5 7, 5 3,1 19,4
Graduated from 4 yr college I I 16.4 17.7 37.]
Graduated from a 2-year college or
technical school 2 3'0 3'2 40'3
Some college or technical school 13 19.4 21.0 61.3
High school 20 29.9 32.3 93.5
Elementary or junior high school 4 6.0 6.5 100.0
Total 62 92.5 100.0
Missing Don't know 5 7.5
Total 67 100.0
Table 7- Dad's Education
Valid Cumulative
Frequency Percent Percent Percent
Valid Professional/graduate degree 3 4,5 5,0 5.0
Some school beyond 4 year college 7 10.4 1 1.7 16.7
Graduated from 4 yr college 6 9,0 10,0 26,7
Graduated from a 2-year college or
technical school 3 4'5 5'0 3 1'7
Some college or technical school 1 1 16,4 18,3 50.0
High school 24 35.8 40.0 90.0
Elementary or junior high school 6 9.0 10.0 100.0
Total 60 89.6 100.0
Missing Don't know 7 10.4
Total 67 100.0
Table 8- Parents Divorced or Separated
Valid Cumulative
Frequency Percent Percent Percent
Valid Never divorced or separated 31 46.3 47.7 47.7
Never married 16 23.9 24.6 72.3
Divorced/separated more than 1 18 2 6.9 2.7.7 100.0
year
Total 65 97.0 100.0
Missing Missing 2 3.0
Total 67 100.0

 

59

 

 

 

Most youth were non-smokers. (See Table 9.) The majority of youth did not

drink at all. (See Table 10.) Most denied having five or more drinks at one time in the

last month. (See Table 11.) Fewer youth reported marijuana use than alcohol use.

(See Table 12.) Even less youth admitted to using other drugs. (See Table 13.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 9- Youth Smoking
Cumulative
Frequency Percent Valid Percent Percent
Valid Not at all 44 65.7 67.7 67.7
Once or twice 3 4.5 4.6 72.3
1 to 3 times a month I 1.5 1.5 73.8
Daily 17 25.4 26.2 100.0
Total 65 97.0 100.0
Missing Missing 2 3.0
Total 67 100.0
Table 10- Youth Drinking
Cumulative
Frequency Percent Valid Percent Percent
Valid Not at all 34 50.7 51.5 51.5
Once or twice 16 23.9 24.2 75.8
1 to 3 times a month 10 14.9 15.2 90.9
1-3 times a week 5 7.5 7.6 98.5
4-6 times a week 1 1.5 1.5 100.0
Total 66 98.5 100.0
Missing Missing 1 1.5
Total 67 100.0
Table 11- More than Five Drinks in the Last Month
Cumulative
Freryency Percent Valid Percent Percent
Valid Never 50 74.6 75.8 75.8
Yes, once 6 9.0 9.1 84.8
Yes, twice 3 4.5 4.5 89.4
Yes, 3-5 times 5 7.5 7.6 97.0
$52510 °’ mm 2 3.0 3.0 100.0
Total 66 98.5 100.0
Missing Missing 1 1.5
Total 67 100.0

 

60

 

 

Table 12- Youth Marijuana Use

 

 

 

 

 

 

 

 

 

 

Cumulative
Frecmency Percent Valid Percent Percent

Valid Not at all 43 64.2 65.2 65.2
Once or twice 11 16.4 16.7 81.8
1 to 3 times a month 2 3.0 3.0 84.8
1-3 times a week 4 6.0 6.1 90.9
4-6 times a week 3 4.5 4.5 95.5
Daily 3 4.5 4.5 100.0
Total 66 98.5 100.0

Missing Missing 1 1.5

Total 67 100.0

Table 13- Youth Use of Other Drugs
Cumulative
Frequency Percent Valid Percent Percent

Valid Not at all 59 88.1 89.4 89.4
Once or twice 3 4.5 4.5 93.9
1 to 3 times a month 4 6.0 6.1 100.0
Total 66 98.5 100.0

Missing Missing 1 1 .5

Total 67 100.0

 

 

 

 

 

 

 

 

The majority of youth strongly disagreed that they would not get pregnant (or

get a partner pregnant) with unprotected sex. (See Table 14.) The majority of youth

also strongly disagreed that they would not get a sexually transmitted disease (STD)

from unprotected sex. (See Table 15.) When asked in the reverse later in the

questionnaire a larger majority strongly agreed that they could get pregnant with

unprotected sex. (See Table 16). More youth also strongly agreed that they could get

an STD with unprotected sex. (See Table 17.)

Table 14- Youth Don't Fear Pregnancy with Unprotected Sex

 

 

 

 

 

 

 

Frequew Percent Valid Percent Cumulative Percent
Valid Strongly disagree 34 50.7 50.7 50.7
Disagree 7 10.4 10.4 61.2
Not sure 5 7.5 7.5 68.7
Agree 5 7.5 7.5 76.1
Strongly agree 16 23.9 23.9 100.0
Total 67 100.0 100.0

 

61

 

Table 15- Youth Don't Fear STDs with Unprotected Sex

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative Percent
Valid Strongly disagree 39 58.2 58.2 58.2
Disagree 6 9.0 9.0 67.2
Not sure 4 6.0 6.0 73.1
Agree 2 3.0 3.0 76.1
Strongly agree 16 23.9 23.9 100.0
Total 67 100.0 100.0
Table 16- Youth Think They Could Get Pregnant with Unprotected Sex
Frequency Percent Valid Percent Cumulative Percent
Valid Strongly agree 52 77.6 77.6 77.6
Agree 10 14.9 14.9 92.5
Not sure 2 3.0 3.0 95.5
Strongly disagree 3 4.5 4.5 100.0
Total 67 100.0 100.0
Table 17- 11‘ Youth Think They Could Get an STD with Unprotected Sex
Frequency Percent Valid Percent Cumulative Percent
Valid Strongly agree 50 74.6 74.6 74.6
Agree 1 1 16.4 16.4 91.0
Not sure 2 3.0 3.0 94.0
Strongly disagree 4 6.0 6.0 100.0
Total 67 100.0 100.0

 

Over half of the youth disagreed or strongly disagreed that their

religion/religious beliefs were important to them (see Table 18) while about 34%

agreed or strongly agreed that they were not very involved with religious activities

(see Table 19).

62

 

 

 

Table 18- Religion Important

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative Percent
Valid Strongly disagree 33 49.3 50.8 50.8
Disagree 12 17.9 18.5 69.2
Not sure 14 20.9 21.5 90.8
Agree 3 4.5 4.6 95.4
Strongly agree 3 4.5 4.6 100.0
Total 65 97.0 100.0
Missing no response 2 3.0
Total 67 100.0
Table 19- Not Very Involved in Religious Activity
Frequency Percent Valid Percent Cumulative Percent
Valid Strongly disagree 22 32.8 33.8 33.8
Disagree 1 1 16.4 16.9 50.8
Not sure 9 13.4 13.8 64.6
Agree 13 19.4 20.0 84.6
Strongly agree 10 14.9 15.4 100.0
Total 65 97.0 100.0
Missing No response 2 3.0
Total 67 100.0

 

When asking if youth had ever had sex how old they were the first time, over

half stated that they had not yet had sex. (See Table 20.) In addition to those youth

who had never had sex, many who had, had not had sex in the last 3 months. (See

Table 21.) For number of life time partners the same number (56.7%, 38) reporting

that they had never had sex reported no sexual partners. (See Table 22.) See Table 23

for youth report of frequency of sex when drunk. Sixty-one (91%) had never been

diagnosed with an STD while the remaining six (9%) had been diagnosed within the

last year. Table 23 shows results for birth control. The majority of youth (88.1%, 59)

had not been pregnant or gotten someone pregnant while 9% (6) had within the last

year, and 3% (2) had more than a year ago.

63

 

 

Table 20- 11’ Ever Had Sex Age 1st Voluntary Sex

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative Percent
Valid Never had sex 38 56.7 56.7 56.7
19 years old 1 1.5 1.5 58.2
17 years old 4 6.0 6.0 64.2
16 years old 5 7.5 7.5 71.6
15 years old 9 13.4 13.4 85.1
14 years old 8 11.9 11.9 97.0
13 years old 1 1.5 1.5 98.5
12 years old 1 1.5 1.5 100.0
Total 67 100.0 100.0
Table 21- How Often Youth Had Sex in the Last Three Months
Frequency Percent Valid Percent Cumulative Percent
Valid Have not had sex in
the last 3 months 43 64’2 64'2 64'2
1 time 6 9.0 9.0 73.1
2 times 2 3.0 3.0 76.1
3 times 3 4.5 4.5 80.6
4 times 1 1.5 1.5 82.1
5 times 2 3.0 3.0 85.1
6 or more times 10 14.9 14.9 100.0
Total 67 100.0 100.0
Table- 22 Number Lifetime Partners
Cumulative
Frequency Percent Valid Percent Percent
Valid No sexual partners 38 56.7 57.6 57.6
1 partner 8 11.9 12.1 69.7
2 partners 4 6.0 6.1 75.8
3 partners 7 10.4 10.6 86.4
4 Partners 5 7.5 7.6 93.9
5 partners 2 3.0 3.0 97.0
6 or more partners 2 3.0 3.0 100.0
Total 66 98.5 100.0
Missing System 1 1.5
Total 67 100.0

 

 

 

 

 

 

64

 

Table 23- How Often Youth Have Sex When Drunk

 

 

 

 

 

 

 

Cumulative
Frequency Percent Valid Percent Percent

Valid Never had sex 39 58.2 58.2 58.2
Never sex when drunk 15 22.4 22.4 80.6
Rarely sex when drunk 4 6.0 6.0 86.6
Sometimes sex when drunk 7 10,4 10,4 97.0
Most of the time 2 3.0 3.0 100.0
Total 67 100.0 100.0

 

 

Table 24- How Often Youth Use Birth Control

 

 

 

Frequency Percent Valid Percent Cumulative Percent

Valid Never had sex 39 58.2 58.2 58.2
Always 17 25.4 25.4 83.6
Most of the time 4 6.0 6.0 89.6
About half the time 1 1.5 1.5 91.0
Sometimes 2 3.0 3.0 94.0
Rarely 1 1.5 1.5 95.5
Never 3 4.5 4.5 100.0
Total 67 100.0 100.0

 

 

 

 

 

 

No youth were currently being sexually abused, though 6 (9%) admitted to
sexual abuse that had stopped, 60 (89.6%) stated they had never been sexually
abused, and 1 (1.5%) did not answer the question. The majority of youth (51, 76.1%)
had never been physically abused though 14 (20.9%) said they had been, but the
abuse had stopped and 2 did not respond to the question. Thirty (44.8%) had never

witnessed abuse while the remainder had in various settings. (See Table 25.)

Table 25- Witnossed Abuse or Beating

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative Percent
Valid Never 30 44.8 44.8 44.8
Yes in home 7 10.4 10.4 55.2
Yes in school 5 7.5 7.5 62.7
Yes in town 9 13.4 13.4 76.1
Yes in home and school 2 3.0 3.0 79.1
Yes in school and town 7 10.4 10.4 89.6
Yes in home and town 2 3.0 3.0 92.5
Yes in all three places 5 7,5 7.5 100.0
Total 67 100.0 100.0

 

 

65

When asked about their community, most strongly agreed or agreed that there
were neighbors who could help them with a problem. (See Table 26.) Results for if
adults in the community would tell parents if youth did something wrong are found in
Table 27 while Table 28 shows youth beliefs that people in their community know

and care about each other.

Table 26- Neighbors Could Help with a Problem

 

 

 

 

 

 

 

Fregency Percent Valid Percent Cumulative Percent
Valid Strongly agree 26 38.8 38.8 38.8
Agree 22 32.8 32.8 71.6
Disagree 10 14.9 14.9 86.6
Strongly disagree 9 13.4 13.4 100.0
Total 67 100.0 100.0

 

Table 27-Community Adult Would Tell Parent if Youth Did Something Wrong

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative Percent
Valid Strongly agree 22 32.8 32.8 32.8
Agree 27 40.3 40.3 73.1
Disagree 8 1 1.9 1 1.9 85.1
Strongly disagree 10 14.9 14.9 100.0
Total 67 100.0 100.0
Table 28- People in the Community Know and Care about Each Other
Frequency Percent Valid Percent Cumulative Percent
Valid Strongly agree 13 19.4 19.4 19.4
Agree 33 49.3 49.3 68.7
Disagree 1 1 16.4 16.4 85.1
Strongly disagree 10 14.9 14.9 100.0
Total 67 100.0 100.0

 

 

 

 

Most youth very often or always told an adult they lived with before going
out. (See Table 29.) The majority also very often or always talked about their plans
with fiiends with an adult in their home. (See Table 30.) Adults in the home very

often or always asked where most youth were going. (See Table 31.) When asked if

66

 

adults in their home usually knew what youth were doing after school most youth

stated they very often or always knew. (See Table 32.)

Table 29- Youth Tell Adults Before Going Out

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative Percent
Valid Very often/always 41 61.2 61.2 61.2
Often 9 13 .4 13 .4 74.6
Sometimes 7 10.4 10.4 85.1
Rarely 5 7.5 7.5 92.5
Never 5 7.5 7.5 100.0
Total 67 100.0 100.0
Table 30- Youth Talk with Adults about Plans
Frequency Percent Valid Percent Cumulative Percent
Valid Very often/always 39 58.2 58.2 58.2
Often 12 1 7 .9 17.9 76.1
Sometimes 6 9.0 9.0 85. l
Rarely 6 9.0 9.0 94.0
Never 4 6.0 6.0 100.0
Total 67 100.0 100.0
Table 31- Adults Ask When Youth Go Out
Frequency Percent Valid Percent Cumulative Percent
Valid Very often/always 42 62.7 64.6 64.6
Often I I 16.4 16.9 81.5
Sometimes 5 7.5 7.7 89.2
Rarely 4 6.0 6.2 95.4
Never 3 4.5 4.6 100.0
Total 65 97 .0 100.0
Missing No adults at home 2 3.0
Total 67 100.0
Table 32- Adult Knows What Youth Do after School
Frequency Percent Valid Percent Cumulative Percent
Valid Very often/always 42 62.7 64.6 64.6
Often 12 17.9 18.5 83.1
Sometimes 4 6.0 6.2 89.2
Rarely 3 4.5 4.6 93.8
Never 4 6.0 6.2 100.0
Total 65 97.0 100.0
Missing No adults at home 2 3.0
Total 67 100.0

 

 

 

 

 

67

 

 

 

 

The majority of youth stated that their parents were there when they needed

them. (See Table 33.) Even more felt their parents cared about them. (See Table 34.)

Table 33- Parent There When Needed

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative Percent
Valid Very often/always 43 64.2 64.2 64.2
Often 13 19.4 19.4 83.6
Sometimes 2 3.0 3.0 86.6
Rarely 5 7.5 7.5 94.0
Never 4 6.0 6.0 100.0
Total 67 100.0 100.0
Table 34- Parent Cares About Youth
Frequency Percent Valid Percent Cumulative Percent
Valid Very often/always 56 83.6 83.6 83.6
Often 5 7 .5 7 .5 91.0
Sometimes 2 3.0 3.0 94.0
Never 4 6.0 6.0 100.0
Total 67 100.0 100.0

 

 

 

Most youth strongly agreed that their parents think it is wrong for teens their

age to have sex. (See Table 35.). While still the majority, fewer strongly agreed that

their parents think it is wrong for teens their age to drink alcohol. (See Table 36.)

Table 35- Parents Say Teens Should Not Have Sex

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative Percent
Valid Strongly agree 42 62.7 62.7 62.7
Agree 9 13.4 13.4 76.1
Not sure 10 14.9 14.9 91.0
Disagree 2 3.0 3.0 94.0
Strongly disagree 4 6.0 6.0 100.0
Total 67 100.0 100.0
Table 36- Parents Say Teens Should Not Drink
Frequency Percent Valid Percent Cumulative Percent
Valid Strongly agree 39 58.2 58.2 58.2
Agree 1 1 16.4 16.4 74.6
Not sure 10 14.9 14.9 89.6
Disagree 2 3.0 3.0 92.5
Strongly disagree 5 7.5 7.5 100.0
Total 67 100.0 100.0

 

 

 

68

Table 37 shows if youth felt their fiiends help them stay out of trouble. Table
38 gives results if most of their friends do not have sexual intercourse. The majority
of youth reported that most of their friends do not drink or do drugs. (See Table 39.)

Most youth reported that most of their friends do not smoke. (See Table 40.)

Table 37- Youth Friends Help Them Stay Out of Trouble

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Frequency Percent Valid Percent Cumulative Percent
Valid Strongly agree 20 29.9 30.3 30.3
Agree 32 47.8 48.5 78.8
Disagree 10 14.9 15.2 93.9
Strongly disagree 4 6.0 6.1 100.0
Total 66 98.5 100.0
Missing No response 1 1.5
Total 67 100.0
Table 38- Most Friends Do Not Have Sex
Frequency Percent Valid Percent Cumulative Percent
Valid Strongly agree 2] 31.3 31.3 31.3
Agree 14 20.9 20.9 52.2
Disagree 1 5 22.4 22.4 74.6
Strongly disagree 17 25.4 25.4 100.0
Total 67 100.0 100.0
Table 39- Most Friends Do Not Drink of Do Drugs
Frequency Percent Valid Percent Cumulative Percent
Valid Strongly agree 22 32.8 32.8 32.8
Agree 14 20.9 20.9 53.7
Disagree 13 19.4 19.4 73.1
Strongly disagree 18 26.9 26.9 100.0
Total 67 100.0 100.0
Table 40- Most Friends Do Not Smoke
Frequency Percent Valid Percent Cumulative Percent
Valid Strongly agree 29 43.3 43.3 43.3
Agree 10 14.9 14.9 58.2
Disagree 14 20.9 20.9 79.1
Strongly disagree 14 20.9 20.9 100.0
Total 67 100.0 100.0

 

 

 

 

 

69

 

 

 

 

Youth were also asked their personal belief about if teenagers should not be
having sexual intercourse. Table 41 shows their responses.

Table 41- Youth Belief That Teens Should Not Have Sex

 

 

Frequency Percent Valid Percent Cumulative Percent
Valid Strongly agree 28 41.8 42.4 42.4
Agree 14 20.9 21.2 63 .6
Disagree 16 23 .9 24.2 87.9
Strongly disagree 8 11.9 12.1 100.0
Total 66 98.5 100.0
Missing No response 1 1.5
Total 67 100.0

 

 

 

 

 

 

 

Confltion between F eelirags of Invulnerafbilitv and Sexual Activity

Using a Spearman’s rho, the correlation coefficient for the first hypotheses
that youth with higher perceived invulnerability will be more likely to be sexually
active was not significant ((1 = .07) with an r of -.19 when looking just at sexual
vulnerability nor was it (a = .12) with an r of -.14 when looking just at the personal
fable scale. With the low reliability of the sexual vulnerability scale, but the strong
correlation of the first two questions with each other and the second two questions
with each other, the hypothesis was also tested with just the first two sexual
vulnerability questions resulting in a significant though weak negative correlation (r =
-.26, a = .02) and again using just the second two sexual vulnerability questions
showing no significant relationship (r = -.08, a = .26). This weak negative correlation
indicates that youth with higher perceived invulnerability are actually less likely to be
sexually active.

Using the same test the second hypothesis that youth with higher perceived
invulnerability will have greater frequency of sexual activity was not significant

either (a = .07) with an r of -.18 for sexual vulnerability nor when looking at the

70

personal fable scale (r = -.12, a = .16). When run with just the first two sexual
vulnerability questions there was a significant though weak negative correlation (r = -
.24, a = .03) indicating that youth with higher perceived invulnerability actually have
less frequency of sexual activity, but using just the second two sexual vulnerability
questions no significant relationship was found (r = -.08, a = .26).

The third hypothesis that youth with higher perceived invulnerability will
have a greater number of lifetime partners remained insignificant with an r of -.15 (a
= .11) for sexual vulnerability and an r of -.09 (a = .24) for the personal fable scale.
When rim with just the first two sexual vulnerability questions there was a significant
though weak negative correlation (r = -.24, a = .03) indicating that youth with higher
perceived invulnerability had fewer number of lifetime partners, but using just the
second two sexual vulnerability questions no significant relationship was found (r = -
.05, a = .33).

The fourth hypothesis that youth with higher perceived invulnerability will
more frequently have sex while drunk or high was significant ((1 = .034) using the
Spearman’s rho showing a small negative correlation of -.23 for sexual vulnerability
indicating that youth with higher perceived invulnerability have sex while drunk less
frequently, but was not significant (or = .15) with an r of -.13 for the personal fable
scale. When run with just the first two sexual vulnerability questions there was a
significant though weak negative correlation (r = -.26, a = .02) again indicating less
frequent sex while drunk for those with higher perceived invulnerability, but using
just the second two sexual vulnerability questions no significant relationship was

found (r = -.10, a = .20).

71

For the fifth hypotheses that youth with higher perceived invulnerability will
have greater use of birth control the results were not significant with an r of -.16 (a =
.10) for sexual vulnerability and an r of -.13 (a = .15) for the personal fable scale.
When run with just the first two sexual vulnerability questions there was a significant
though weak negative correlation (r = -.24, o. = .03) indicating that youth with higher
perceived invulnerability have less use of birth control, but using just the second two
sexual vulnerability questions no significant relationship was found (r = -.04, a =
.39).

Due to the low reliability of the personal fable scale a factor analysis was
done. Based on the factor analysis the five items with the highest loading on the first
factor were chosen from the original 14 item scale. These components included
questions 51, 53, 57, 59, and 60. (See Appendix A). When the first hypotheses that
youth with higher perceived invulnerability will be more likely to be sexually active
was tested using this new five item scale, the results were still not significant with an
alpha of .12 and an r of .15. The results remained insignificant for the hypothesis that
youth with higher levels of invulnerability would more frequently have sex (r = .19, a
= .11), have greater number of lifetime partners (r = .14, a = .12), would have greater '
use of birth control (r = .09, a = .24), and be more likely to have sex when drunk or
high (r = .14, a = .13).

Using an independent samples t test for the hypotheses that youth who have
ever had an STD will have lower levels of perceived invulnerability, for sexual
vulnerability those without a history of an STD had a mean score of 3.33 compared to

those with no history who averaged 3.56 indicating no significant difference (a = .90)

72

with a t value of .13 with 6.11 degrees of freedom (df). Using just the first two sexual
vulnerability questions those with a history an STD had a mean score of 1.67 while
those with no history had a mean score of 2.79 indicating no significant difference (a
= .38) with at value of .94 with 6.51 df. Using just the second two sexual
vulnerability questions those with a history an STD had a mean score of 1.67 while
those with no history had a mean score of .77 indicating no significant difference (a =
.53) with a t value of -.68 with 5.27 df. General vulnerability scores varied from an
average of 29 for those with a history of STD to 25.82 for those without but the
difference was still not statistically significant ((1 = .068) with a t of —2.06 and 9.27
df. When looking at only the five items with the highest loading form the personal
fable scale, the invulnerability score mean was 12.67 for those with a history of STD
compared to 10.54 for those without, remaining insignificant with an alpha of .25, at
of —1 .29 and 5.29 df.

Using the same test to look at the hypothesis that youth who have ever been or
had ever gotten someone pregnant will have lower levels of perceived invulnerability,
sexual vulnerability scores averaged 2.13 for those who had been or had gotten
someone pregnant while those who had never been or gotten someone pregnant
averaged 3.7, though the difference was not significant ((1 = .31, t = 1.1, and df =
9.31). Using just the first two sexual vulnerability questions showed a mean of 2.80
for those who had never been or gotten someone pregnant compared to 1.88 for those
who had (a = .50, t = 71, and df = 8.70) while using the second two sexual
vulnerability questions resulted in a mean of .93 for those with no history of

pregnancy and .25 for those who did have a history of pregnancy (a = .07, t = 1.92,

73

and df = 25.34). For general invulnerability the mean was 26.13 for those who had
been or had gotten someone pregnant compared to 26.1 for the rest which indicated
no significant difference (a = .99, t = -.01, and df = 12.73). The five factors with the
highest loading from the personal fable scale mean for those with a history of
pregnancy was 10.88 compared to 10.71 for those without a history still indicating no
significant difference (a = .88, t = -.16, and df = 11.21).

Before testing the hypothesis that youth with higher perceived invulnerability
will be more likely to be sexually active when other predictors of sexual activity are
controlled, Spearman’s rho correlations were run on these variables to make sure that
they were in fact correlated to sexual activity in this population. Average grades was
significantly correlated to sexual activity with an r of .53 (a < .01), as was parental
approval of teen sex (r = .51, a < .01), and parental approval of teen drinking (r = .29,
a = .02). Other significant results included peer group (r = .47, a < .01), drug use (r =
.60, a < .01), religion (r = .32, o. = .01), youth’s belief that teens shouldn’t have sex (r
= .62, a < .01), parents education (r = .48, a <.01), parental support (r = .38, a < .01),
supportive community (r = .23, a = .03), and parental monitoring (r = .36, a < .01),
all indicating higher likelihood of sexual activity for youth with high risk behavior.
For nominal level variables Chi-square tests were performed. Looking at with whom
youth live the results were also significant (value 18.39 with 7 degrees of freedom
and an alpha of .01) as they were for if parents were divorced (value 10.53, df = 2, a
< .01), if youth had witnessed abuse (value 12.00, df = 1, o. < .01), and if youth had
been physically abused (value 14.40, df = 1, a < .01) all indicating greater likelihood

of sexual activity for youth with risk factors while if youth had been sexually abused

74

was not significantly correlated to sexual activity (value .16 with 1 degree of freedom
and an alpha of .70).

Due to the relatively small sample size of 67 surveys only the four highest
scaled variable predictors of sexual activity were included in the initial logistic
regression to test the hypothesis that youth with higher perceived invulnerability will
be more likely to be sexually active when other predictors of sexual activity are
controlled. Using teens’ belief that teens should not have sex, drug use, GPA, and
parents’ approval of teen sex along with sexual vulnerability teens belief that teens
should not have sex and GPA were significant while the other variables were not.
(See Table 42.) The model Chi-square indicated a good fit with a value or 47.6 with 5
degrees of freedom and a significance level of < .01. Usingthe same four variables
but this time with only the first two sexual vulnerability questions, teens belief that
teens should not have sex and GPA remained significant while the other variables
were not with a model Chi-square indicating a good fit with a value or 47.47 with 5
degrees of freedom and a significance level of < .01. (See Table 43.) Again using the
same four variable but with the second two sexual vulnerability questions, belief that
teens should not have sex and GPA were still significant while the other variables
remained insignificant. The model Chi-square indicated a good fit with a value or

46.71 with 5 degrees of freedom and a significance level of < .01. (See Table 44.)

75

Table 42— Logistic Regression of Sexual Vulnerability and Top 4 Correlates

 

 

B S.E. Wald df Sig. Exp(B)

Step 1(a) I believe teens should not have sex 1,087 .450 5.824 1 .016 2.966
Drug use by teen .167 .109 2.363 1 .124 1.182

GPA .629 .296 4.502 1 .034 1.875

My parents think teen sex is wrong .697 .394 3.140 1 .076 2.008

Sexual vulnerability (all 4 questions) -,130 .128 1.035 1 .309 .878

Constant -3.553 1.077 10.875 1 .001 .029

 

 

 

 

 

 

 

 

 

a Variable(s) entered on step 1: teennosex, druguse, gpa, pamosex, sexvuln.

Table 43- Logistic Regression of 1" Two Sexual Vulnerability Questions and Top 4 Correlates

 

 

B S.E. Wald df Sig. Erqr(B)

Step 1(a) I believe teens should not have sex 1.052 .454 5.379 1 .020 2.863
Drug use by teen .171 .107 2.538 1 .111 1.186

GPA .606 .295 4.232 1 .040 1.833

My parents think teen sex is wrong .724 .399 3.282 1 .070 2.062

Sexual vulnerability (1" questions) -,143 .148 .936 1 .333 .867

Constant -3.542 1 .083 10.698 1 .001 .029

 

 

 

 

 

 

 

 

 

a Variable(s) entered on step 1: teennosex, druguse, gpa, pamosex, sexvullnd2.

Table 44- Logistic Regression 01'2”l Two Sexual Vulnerability Questions and Top 4 Correlates

 

 

B S.E. Wald df Sig. Exp(B)

Step 1(a) I believe teens should not have sex 1,140 .446 6.531 1 .011 3.127
Drug use by teen .197 .108 3.316 1 .069 1.217

GPA .627 .289 4.706 1 .030 1.871

My parents think teen sex is wrong .594 .374 2.520 1 .112 1.811

Sexual vulnerability (2"‘1 questions) -, 139 .284 .239 1 .625 .870

Constant -3.958 1.070 13.695 1 .000 .019

 

 

 

 

 

 

 

 

 

a Variable(s) entered on step 1: teennosex, druguse, gpa, pamosex, sexvul3and4.

When the same four variables were used with the personal fable scale teens
belief that teens should not have sex and GPA remained significant and drug use also
was significant. (See Table 45.) With 5 degrees of freedom the goodness if fit for a
model Chi-square was significant with the alpha < .01 and a value of 47.96. When run
with the five items with the highest loading from the personal fable scale, teen belief

that teens should not have sex and GPA were the only significant factors. The

76

goodness of fit was significant with an alpha of < .01, a model Chi-square value of

46.56, and 5 df. (See Table 46.)

Table 45- Logistic Regression of’ Personal Fable and Top 4 Correlates

 

 

 

 

 

 

 

 

 

 

 

 

B S.E. Wald df Sig. Exp(BL
Step 1(a) I believe teens should not have sex 1,144 .437 6.861 1 .009 3.138
Drug use by teen .216 .106 4.176 1 .041 1.241
GPA .649 .289 5.040 1 .025 1.914
My parents think teen sex is wrong .614 .370 2.755 1 .097 1.849
General vulnerability scale -.091 .077 1.401 1 .237 .913
Constant -1.880 1.973 .908 1 .341 .153
a Variable(s) entered on step 1: teennosex, druguse, gpa, pamosex, genvulnerability.
Table 46- Logistic Regression of' 5 Items with the Highest Loading on the
Personal Fable Scale and Top 4 Correlates
' B S.E. Wald df fig Exp(B)
Step 1(a) I believe teens should not have sex 1,128 .445 6.430 1 .011 3.089
Drug use by teen .198 .109 3.270 1 .071 1.218
GPA .594 .283 4.422 1 .035 1.812
My parents think teen sex is wrong .630 .373 2.853 1 .091 1.878
PM fighes‘ wading vulnerability .039 .141 .077 1 .782 1.040
questions

Constant -4.401 1.800 5.982 1 .014 .012

 

 

 

 

 

 

 

 

a Variable(s) entered on step 1: teennosex, druguse, gpa, pamosex, topfactrovul.

Further testing the hypothesis that youth with higher perceived invulnerability

will be more likely to be sexually active when other predictors of sexual activity are

controlled the top four nominal level variables converted to dichotomous variables

were used. These included not living with both biological parents, parents having

been divorced or separated, witnessing abuse, and having a history of abuse. Again

using logistic regression with these four dichotomous variables and sexual

vulnerability living with both biologic or adoptive parents and witnessing abuse were

significant while the other components were not. (See Table 47.) With 5 degrees of

 

 

freedom the goodness if fit for a model Chi-square was significant with the alpha of <

.01 and a value of 18.49. Rerunning the model with only the first two sexual

77

vulnerability questions living with both biologic or adoptive parents and witnessing

abuse were significant while the other components were not (see Table 48) with 5

degrees of freedom the goodness if fit for a model Chi-square was significant with the

alpha of < .01 and a value of 18.47. Next using only the second two sexual

vulnerability questions only witnessing abuse remained significant (see Table 49)

with 5 degrees of freedom the goodness if fit for a model Chi-square was significant

with the alpha of < .01 and a value of 18.06.

Table 47- Logistic Regression with Top 4 Dichotomous Variables and Sexual Vulnerability

 

 

 

 

 

 

 

 

 

B S.E. Wald df Sig Exp(B)

Step 1(a) Lives with two parents 1.215 .615 3.899 1 .048 3.370
Parents ever divorced or separated .030 .156 .036 1 .849 1.030

Has witnessed abuse, yes or no 1.384 .632 4.804 1 .028 3.992

Has ever been abused, yes or no -.052 .198 .070 1 .791 .949

Sexual vulnerability (4 questions scale) .06] .075 .656 1 .418 .941

Constant -1.525 .627 5.927 1 .015 .218

 

a Variable(s) entered on step 1: live2parents, everdivorce, witnessabusedic, everabusedic, sexvuln.

Table 48- Logistic Regression with Top 4 Dichotomous Variables

and 1" Two Sexual Vulnerability Questions

 

 

 

 

 

 

 

 

 

B S.E. Wald df Sig. Exp(B)

Step 1(a) Lives with two parents 1.266 .616 4.218 1 .040 3.547
Parents ever divorced or separated .018 .157 .013 1 .908 1.018

Has witnessed abuse. yes or no 1.315 .652 4.064 1 .044 3.725

Has ever been abused, yes or no -,050 .196 .066 1 .798 .951

Sexual vulnerability (1’t questions) -,080 .100 .651 1 .420 .923

Constant -1.507 .637 5.590 1 .018 .222

 

a Variable(s) entered on step 1: live2parents, everdivorce, witnessabusedic, everabusedic, sexvullnd2.

78

 

 

Table 49- Logistic Regression with Top 4 Dichotomous Variables
and 2"“ Two Sexual Vulnerability Questions

 

 

 

 

 

 

 

 

B S.E. Wald df Sig. Exp(B)

Step 1(a) Lives with two parents 1.183 .622 3.614 1 .057 3.265
Parents ever divorced or separated .048 .160 .089 1 .765 [.049

Has witnessed abuse, yes or no 1.539 .619 6.172 1 .013 4.660

Has ever been abused. yes or no -.064 .198 .105 1 .746 .938

Sexual vulnerability (questions 3and4) -,080 .163 .243 1 .622 .923

““5”“ -1.747 .548 1°"? 1 .001 .174

 

 

a Variable(s) entered on step 1: live2parents, everdivorce, witnessabusedic, everabusedic,

sexvul3and4.

Using the same four dichotomous variables with the personal fable scale

living with both biologic parents and witnessing abuse remained significant while no

other components were significant. (See Table 50). The model Chi-square was

significant (0. < .01) with a value of 19.42 and 5 df. Using the same four variables

with the five items with the highest loading from the personal fable scale showed that

only witnessing abuse was significant. (See Table 51.) The goodness of fit for this

model was significant ((1 < .01) with a Chi-square value of 17.87 with 5 df.

Table 50- Logistic Regression with Top 4 Dichotomous Variables and Personal Fable Scale

 

 

 

 

 

 

 

 

B S.E. Wald df Sig Exp(B)
Step 1(a) Lives with two parents 1.435 .646 4.935 1 .026 4.200
Parents ever divorced or separated .071 .163 .190 1 .663 1.074
Has witnessed abuse, yes or no 1.410 .622 5.141 1 .023 4.097
Has ever been abused. yes or no -.051 .199 .065 1 .799 .951
General vulnerability scale -.065 .052 1.556 1 .212 .937
Constant -. 182 1.377 .017 1 .895 .834

 

 

 

a Variable(s) entered on step 1: live2parents, everdivorce, witnessabusedic, everabusedic,
genvulnerability.

79

 

Table 51- Logistic Regression with Top 4 Dichotomous Variables and 5 Items with the
Highest Loading on the Personal Fable Scale

 

B S.E. Wald df Sig. Exp(B)
Step 1(a) Lives with two parents 1.208 .630 3.675 1 .055 3.347
Pam“ eve’d'vmed °' .038 .160 .057 1 .811 1.039

 

 

 

 

 

 

 

 

 

separated

Has witnessed abuse, yes 01' 110 1.503 .613 6.001 1 .014 4.493
Has ever been abused, yes or no -.071 .197 .131 1 .718 .931
Five highest loading vulnerability

questions .020 .083 .056 1 .8 13 1 .020
Constant -2.016 .979 4.244 1 .039 .133

 

a Variable(s) entered on step 1: live2parents, everdivorce, witnessabusedic, everabusedic,
topfacu'ovul.

In order to address the hypothesis that perceived invulnerability will have less
impact on the sexual behavior of youth with few risk factors and more protective
factors while having a greater impact on those with multiple risk factors and fewer
protective factors, the scaled variables were changed to dichotomous variables. Due
to the varying metrics across scales, this was necessary to prevent unequal weighting
for risk and protective factors.

For GPA, youth with average grades (i.e. half Bs and half Cs and better) were
classified as low risk while those getting mostly Cs or below were classified as being
high risk based on Lammers, Ireland, Resnick, and Blum’s (2000) findings that youth
with above average school performance were lower risk than those with average or
below school performance. Looking at family structure, the literature supports
(Mullan, Duncan, & Boisjoly, 2002; Talashek, Norr & Dancy, 2003) that living with
both biological parents vs. with one biological parent and a step parent, a single
parent, or any other arrangement is a protective factor as well as having parents who
have never been divorced or separated.

While little in the literature specified how much parental education would be

protective for youth, Kirby, Coyle, and Gould (2001) used the term “college

80

education” when looking at how overall community education impacted risk taking
behaviors in youth. With this in mind, a combined parental education score of seven
or lower (indicating on average that parents had some college or technical school or
had graduated from a 2-year college or technical school) was considered protective
while a combined score of eight or higher (indicating less parental education) was
considered a risk factor.

Changing drug use to a dichotomous variable was more challenging as even
low frequency use of harder drugs may be higher risk than higher fi'equency use of a
substance such as tobacco. With this in mind a youth, on average, smoking or using a
tobacco product three times a month or less (score of 2 or less), having had alcohol
twice or less ever (score of 1 or less), having never had marijuana (score of 0) or
other drugs (score of 0), and not having had more than four alcoholic drinks at one
time in the last month (score of 0) would be considered low risk with total drug scores
of three or less while scores of four and above would be considered a high risk.

For importance of religion/religious beliefs and religious activities youth with
combined scores of zero to one on the two item scale were considered low risk while
those with scores of two or higher were considered high risk. For history of sexual
abuse, physical abuse and witnessing physical abuse not having a history or
witnessing abuse was low risk while having been or witnessed abuse was high risk for
each of the three variables.

For the caring community scale scores of four and under were re-coded as low
risk meaning youth on average agreed with at least two of the three statements of

community caring while scores of five and above were considered high risk. Scores

81

of four and below on the parental monitoring scale indicating parents were often
aware of their youth’s activity for an average of at least three of the questions while
on average sometimes being aware for the remaining question were considered
protective. Scores of four or higher indicating less parental monitoring, were
considered a risk. On the parental support scale scores of three or less showing that
youth felt parents were often or very often there and cared about them indicated low
risk while scores of four and above indicated high risk.

Looking at the parental approval of teen sexual activity and alcohol use each
of these zero to four scores convert to zero for low risk strongly agreeing or agreeing
while one, high risk, would be represented by scores of two to four ranging from not
sure to strongly disagreeing. For peer group, combined scores of five and under
indicate low risk with generally agreeing that their friends help them stay out of
trouble and do not engage in risky behaviors while scores of 6 and above would
indicate at least some high risk behaviors in their friends. Lastly, for youth belief that
teens should not be having sex youth agreeing or strongly agreeing (0 or 1) were
coded as low risk (0) while youth disagreeing or strongly disagreeing (2 or 3) were
coded as high risk (1).

All 16 of these risk factors were totaled for a possible score of 0 to 16. To test
the hypothesis that perceived invulnerability will have less impact on the sexual
behavior of youth with few risk factors and more protective factors while having a
greater impact on those with multiple risk factors and fewer protective factors a
logistic regression was done entering total risk together with perceived vulnerability

to look for an interaction between these two variables and sexual activity. This test

82

was first run with sexual vulnerability. The results were not significant (see Table 52)
and the model Chi-square indicated a poor goodness of fit with a value of 1.51 with 1
df and an alpha of .22. When run with just the first two sexual vulnerability questions
the results remained insignificant (see Table 53) with the model Chi-square indicating
a poor goodness of fit with a value of .42 with 1 (If and an alpha of .56. Then run with
just the second two sexual vulnerability questions the results remained insignificant
(see Table 54) with the model Chi-square indicating a poor goodness of fit with a
value of 2.03 with 1 df and an alpha of .15. General vulnerability based on the
personal fable scale was run next. Results from this test were significant as was the
goodness of fit with the model Chi-square value of 22.08 with 1 df and an alpha of <
.01 indicating support for they hypothesis. (See Table 55.) When running the test with
the five items with the highest loading from the personal vulnerability scale the
results were also significant, again supporting the hypothesis (see Table 56) with the

model Chi-square indicating a good fit with a value of 21.42, 1 df and an alpha of

 

 

<.01.
Table 52- Logistic Regression with Total Risk and Sexual Vulnerability
' B 5.13. Wald df Sig. Exp(B)
Step 1(a) Total risk factors by sexual 017 014 1 40 5 1 236 1 017
vulnerability ' ' ' ' '
Constant -.499 .340 2.1 50 1 . 143 .607

 

 

 

 

 

 

 

 

 

a Variable(s) entered on step 1: riskfactorstotaldic * sexvuln .

Table 53- Logistic Regression with Total Risk and 1" Two Sexual Vulnerability Questions

 

 

 

 

 

 

 

 

 

 

B S.E. Wald df Sj. Exp(B)

Step 1(a) Total risk factors by lst sexual
vulnerability questions .012 .018 .418 1 .518 1.012
Constant -.397 .336 1.398 1 .237 .673

 

a Variable(s) entered on step 1: riskfactorstotaldic " sexvullnd2 .

83

Table 54- Logistic Regression with Total Risk and 2"" Two Sexual Vulnerability Questions

 

 

 

 

 

 

B $.13. Wald df s15. Exp(B)
Step 1(a) Total risk factors by 2'“I sexual
vulnerability questions .049 .050 .982 I .322 1.050
Constant -.414 .307 1.816 I . I 78 .661

 

 

 

 

a Variable(s) entered on step 1: riskfactorstotaldic "‘ sexvul3and4 .

Table 55- Logistic Regression with Total Risk and Personal Fable Scale

 

 

 

 

B S.E. Wald df Sig Exp(B)

Step 1(a) General vulnerability scale by 12.77
total risk factors .018 .005 9 1 .000 1.018
““5““ -2573 .740 ”'1? 1 .001 .076

 

 

 

 

 

 

 

 

a Variable(s) entered on step 1: genvulnerability * riskfactorstotaldic .

Table 56- Logistic Regression with Total Risk and 5 Items with the
Highest Loading on the Personal Fable Scale

 

 

 

B S.E. Wald df Sig. Exp(B)

Step Total risk factors by five highest 12.05
1(a) loading vulnerability questions '040 '01 l 3 l '00] 1040
““5“” -2.349 .686 ”'73 1 .001 .095

 

 

 

 

 

 

 

 

a Variable(s) entered on step 1: riskfactorstotaldic . topfactrovul .
Correlation of Factors Impacting Sexual Activity and Feelings of Invulnerability
Testing for the following hypothesis were all done using a Spearman’s rho as
the samples were not normally distributed or were not interval level variables. For the
hypothesis that younger youth will have higher levels of invulnerability using the
personal fable scale the r of -.05 was not significant ((1 = .35) nor was it when looking
just at feelings of invulnerability to the potential consequences of sexual activity (a =
.09) with an r of -.21. It remained insignificant (r = -.18, 01 = .07) when run with just
the first two sexual vulnerability questions but was significant using the second two
sexual vulnerability questions (r = -.26, a = .02). Running this correlation with the

adjusted five items with the highest loading from the of the personal fable scale from

84

the factor analysis, the results were also significant with an r of .21 and an alpha of
.05.

The r of <01 was not significant (01 = .49) for the hypothesis that youth who
use alcohol or drugs will have higher levels of invulnerably but was (a = .01, r = -.28)
when looking just at sexual vulnerability indicating a negative correlation between
alcohol or drug use and feelings of invulnerability. When looking at just the first two
sexual vulnerability questions the negative relationship was a little stronger (r = -.32,
a < .01) but was not significant using just the second two questions (r = -.06, a = .31).
Then using the five items with the highest loading from the personal fable scale the
results were again significant, this time with a positive correlation of .21 and an alpha
of .05.

For the hypothesis that youth with strong religious beliefs will have lower
levels of invulnerability using the personal fable scale the r was .03 with an a = .40
and remained insignificant even when just looking at sexual vulnerability (a = .28, r =
.07), just the first two sexual vulnerability questions (r = .02, a = .42), just the second
two sexual vulnerability questions (r = .20, a = .051), as well as the five items with
the highest loading from the personal fable scale (r = .10, a = .21).

For the hypothesis that youth with higher levels of parental monitoring will
have lower levels of invulnerability looking at the personal fable scale the r was .09
(a = .23) and remained insignificant even when just looking at sexual vulnerability (r
< .-.01, o. = .49), just the first two sexual vulnerability questions (r = .04, a = .37),
just the second two sexual vulnerability questions (r = -.04, a = .39), and the five

items with the highest loading from the personal fable scale (a = .21, r = .10).

85

An insignificant r value of .05 (a = .36) was obtained for the hypothesis that
youth with higher levels of parental support will have lower levels of invulnerability
using the personal fable scale as it was for sexual invulnerability (r = .-.18, a = .07),
just the first two sexual vulnerability questions (r = -.12, a = .17), just the second two
sexual vulnerability questions (r = -.15, a = .12), and the five items with the highest
loading from the personal fable scale (r = .12, a = .16).

Using the personal fable scale an r of -.05 (o. = .34) resulted for the hypothesis
that youth who report having friends who get into trouble, are sexually active, drink
or use drugs, or smoke will have lower levels of perceived invulnerability. Using
sexual invulnerability the results remained insignificant (a = .06) with an r of .-.20
but a weak negative significant relationship was found using just the first two sexual
vulnerability questions (r = -.2 l , o. = .05), but was not using just the second two
sexual vulnerability questions (r = -.15, a = .11), while the five items with the highest
loading from the personal fable scale resulted in an insignificant r of .16 and an alpha
of .10.

The hypothesis that youth reporting supportive communities will have lower
levels of invulnerability was tested using the Pearson Product Moment Correlation as
the results were fairly normally distributed. The results were not significant when
using the personal fable scale with an r of .08 ((1 =26). Looking at sexual
vulnerability resulted in an r of -.14 and an alpha of .13, using just the first two
sexual vulnerability questions resulted in an r of -.18 and an alpha of .08, using just

the second two sexual vulnerability questions resulted in an r of -.01 and an alpha of

86

.46, while the five items with the highest loading from the personal fable scale had an
r of .12 and an alpha of.l8.

Using a Spearman’s rho for the hypothesis that youth with higher GPAs will
have lower levels of invulnerability the correlation coefficient of r = .01 was not
significant ((1 = .46). Looking at sexual vulnerability the r = -.06 was not significant
either (a = .33) nor was it for just the first two sexual vulnerability questions with an r
of -.11 with an alpha of .20 or just the second two sexual vulnerability questions (r
=.46, a = .39). It was, though, for the five items with the highest loading from the
personal fable scale (r =.20, a = .01).

Using the personal fable scale, the Spearman’s rho r value of .06 was not
significant ((1 = .34) for the hypothesis that youth who have parents with higher levels
of education will have lower levels of invulnerability. When using the sexual
vulnerability scale the r was -.13 with an alpha of .17, using just the first two sexual
vulnerability questions the r was -.10 with an alpha of .24, using just the second two
sexual vulnerability questions the r was -.13 with an alpha of .17, and with the five
items with the highest loading from the personal fable scale the r was significant (or =
.03) with a weak correlation value of .24.

For the hypothesis that youth with parents that are more approving of
adolescent sexual activity will have higher levels of invulnerability, the results
indicated an r of -.04 (a = .3 8) for the personal fable scale. Using the sexual
vulnerability scale resulted in an r of .02 with an alpha of .43. Using just the first two
sexual vulnerability questions showed an r of .06 with an alpha of .33 while just the

second two sexual vulnerability questions resulted in an r of -.09 and an alpha of .24.

87

With the five items with the highest loading from the personal fable scale, the results
remained insignificant with an r of -.11 with an alpha of .19.

When looking at the hypothesis that youth whose parents were more
approving of teen drinking will have higher levels of invulnerability using the
personal fable scale the r was .14 (a = .13). When using the sexual vulnerability scale
the r was -.11 (a = .18). With just the first two sexual vulnerability questions the r
was -.10 (a = .45) and using just the second two sexual vulnerability questions the r
of -.24 was significant (6 = .02), but using the five factor scale the r of .13 was no
longer significant with an alpha of .14.

For the hypothesis that youth who believe that teens should not have sexual
intercourse will have lower levels of invulnerability using the personal fable scale the
Spearman’s rho was -.03 which was not significant ((1 = .39) nor was the r of -.08 (a
= .25) for the sexual vulnerability scale, just the first two questions of the sexual
vulnerability scale (r = .09, a = .25), just the second two questions of the sexual
vulnerability scale (r = .10, a = .22), or the five items with the highest loading from
the personal fable scale (r = .10, o. = .22).

For the hypothesis that youth with a history of sexual abuse will have lower
levels of invulnerability, the mean personal fable score was 25.96 for those who had
never been abused and 26.93 for those with a history of sexual abuse. An independent
samples t test confirmed that there was no significant difference between the means
of the two groups with a t of -.58, 24 degrees of freedom (df) and an alpha of .57.
Looking at only the sexual vulnerability scale the mean for those never having been

sexually abused was 3.6 compared to 3.5 for those who had. The independent samples

88

t test also confirmed that there was no significant difference between the means with
a t of .08, 7.47 degrees of freedom and an alpha of .94, again not supporting the
hypothesis. Using only the first two sexual vulnerability questions the mean for those
never having been sexually abused was 2.82 compared to 1.83 but at test again
confirmed that there was no significant difference between them with a t of .71, 64 df,
and an alpha of .48. Then with just the second two sexual vulnerability questions the
mean for those never having been sexually abused was .78 compared to 1.67 for those
with a history of sexual abuse with attest confirming no significant difference with a
t of -.67, 5.28 df, and an alpha of .53. When using only the five factor vulnerability
scale the mean for those who had been abused was 10.80 compared to 10.83 for the
others and again did not support the hypothesis with a t of -.02, 5.7 degrees of
freedom and an alpha of .99.

The mean personal fable vulnerability score for youth with a history of
physical abuse was 26.93 compared to 25.96 for those without. This did not support
the hypothesis that youth who had been abused would have lower levels of
invulnerability, with a t of -.58, 23.77 df and an alpha of .57. Looking at sexual
vulnerability the mean for those with a history of abuse was 2.64 while it was 3.92 for
the rest. The t of 1.13 with 25 degrees of freedom showed an insignificant alpha of
.27 also not supporting the hypothesis. Using just the first two sexual vulnerability
questions the mean for those who had never been abused was 2.9 compared to 2.29
for those with a history of abuse with a t of .65, 22.06 df, and an insignificant alpha of
.52. For the second two sexual vulnerability questions the mean for those with no

history of abuse was 1.02 compared to .36 with a t value of 1.18, 63 df, and an

89

insignificant alpha of .24. The hypothesis was not supported using just the five factor
vulnerability scale either, which resulted in means of 10.43 for those with no history

of abuse compared to a mean of 11.93 for those with a history and at value of —1 .49,
22.85 degrees of freedom and an alpha of .15.

For the hypothesis that youth who have witnessed physical abuse will have
lower levels of invulnerability, when looking just at the personal fable vulnerability
scale youth who had witnessed abuse had a mean vulnerability score of 25.57 while
those who had not had a mean score of 26.77 for at of .79 with 60.86 df and was not
significant with an alpha of .43. Using the sexual vulnerability scale showed means of
2.49 for those witnessing abuse compared to 4.83 for those not witnessing abuse. The
t value of 2.35 with 61 degrees of freedom did support the (hypothesis that youth who
had witnessed abuse would have lower invulnerability scores with an alpha of .02.
Using just the first two sexual vulnerability questions showed a mean of four for
those never having witnessed abuse while those who had showed a mean of 1.62. The
t value of 3.19 with 65 (If was also significant with an alpha of < .01. With just the
second two sexual vulnerability questions means for those never having witnessed
abuse were .83, very close to those never having witnessed abuse at .87 with a t of -
.07, 64.54 df, and an alpha of .94. Finally, testing the same hypothesis using the five
factor personal fable invulnerability scale the mean for those witnessing abuse was
11.11 while those not witnessing abuse had a mean of 10.27 resulting in an
insignificant (a = .34) t of -.95 with 63.28 degrees of freedom.

Males had a mean personal fable vulnerability score of 26.78 while females

averaged 25.75. For the hypothesis that males will have higher levels of

90

invulnerability, an independent samples t test resulted in a t of .68 with 37.92 df and
was not significant with an alpha of .54. Looking at sexual vulnerability males had a
mean score of 5.78 compared to 2.36 for females resulting in at of 3.02 with 32.37 df
and a significant alpha of < .01. With just the first two sexual vulnerability questions
the mean for males was 4.61 compared to 1.68 for females resulting in a t of 3.87
with 65 degrees of freedom and an alpha of less than .01. Using just the second two
sexual vulnerability questions the mean for males was 1.17 compared to .68 for
females with a t of .90, 31.22 df, and an insignificant alpha of .37. Using the five
factor personal fable scale showed a mean of 10.74 for males and 10.73 for females
with results insignificant (a = .99) for a t of .01 with 38.14 df.

The mean personal fable vulnerability score for minority youth was 25.22
while white youth had a mean score of 26.43. The results of the independent samples
1 test did not support the hypothesis that minority youth will have higher levels of
invulnerability, with a t of .75, 33.68 df and an alpha of .46. Looking just at sexual
vulnerability the mean for minority youth was 2.67 compared to 3.86 for white youth
but the results remained insignificant with an alpha of .27 for at value of 1.11 with
35.42 df. Using just the first two sexual vulnerability scale questions showed similar
results with mean for minority youth 2.92 compared to 2.06 for white youth with at
of .99, 31.84 df, and an alpha of .33. The second two sexual vulnerability questions
were also quite close with means of .61 for minority youth and .94 for white youth
resulting in a t of .86 with 58.6 df and an alpha of .40. The five factor personal fable
scale had similar results with a mean of 3.13 for minority and 3.70 for white youth

with a t value of 1.86 with 35.57 df and an alpha of .07.

91

For the hypothesis that youth who live with both parents will have lower
levels of invulnerability the results were not significant. The mean personal fable
vulnerability score for those living with both parents was 25.15 with those not living
with both parents was 27.09 with the results of the independent samples t test
showing a t of -.1.31 with 64.58 df and an alpha of .19. The mean score on the sexual
vulnerability scale for youth living with both parents was 4.38 and 2.67 for the others
resulting in a t value of 1.7 with 60.29 df with an alpha of .09. Using just the first two
sexual vulnerability questions the mean for youth living with both parents was 3.24
compared to 2.12 for those not living with both parents resulting in at of 1.42, 64.45
df and an alpha of .16. With the second two sexual vulnerability questions the mean
for youth living with both parents was 2.11 compared to 1.50 those not living with
both parents resulting in at of 1.35, 59.74 df, and an alpha of .18. The mean for youth
living with two parents on the 5 item personal fable scale was 9.76 compared to 11.73
for the others with at value of —2.29, 60.63 df and a significant alpha of .026
supporting this hypothesis.

The mean personal fable score for youth whose parents had never been
divorced or separated was 25.35 while the mean for the rest was 27 .09 with the
independent samples t-test showing no significant difference with at of-l.15, 62.73
(if and an alpha of .26. Means for youth whose parents had never been divorced or
separated on the sexual vulnerability scale was 4.35 compared to 2.76 for the others
but was still not significant (t = 1.52, df = 53.56, and 01 = .13). Using the first two
sexual vulnerability questions the mean for those whose parents had never divorced

or separated was 3.16 compared to 2.18 for those whose parents had with a t of 1.24,

92

59.93 df, and an alpha of .22. With the second two sexual vulnerability questions the
mean for those whose parents had never divorced or separated was 1.12 compared to
.59 for those whose parents had not resulting in a t of 1.31 with 63 df and an alpha of
.19. Using the top 5 factors the mean for those whose parents had never divorced or
separated was 9.84 with the others at 11.56, but again was not significant (t = -1.97,

df= 61.93, and a = .05).

93

Chapter 5
Discussion of Results

This study used an ecological perspective to help better understand adolescent
sexual activity. More specifically, it looked at the little explored individual factor of
perceived invulnerability, in relation to risk factors associated with significant
contexts of sexual activity and other individual factors.

Youth Demographics

It is helpful to look at how the youth in this study compare to other youth in
the area and nation wide. While a convenience sample was used, with some potential
self selection bias, in most categories the sample appears to be fairly representative.

One area where the results were somewhat atypical was in regards to gender.
Almost twice as many female youth as male youth filled out surveys. It was expected
that this convenience sample would have fairly equal numbers of males and females,
though it is possible that the Life Skills groups surveyed had a higher percentage of
female youth or it could be that the girls were better about taking the consent forms
home for parental signatures and remembering to return them on time.

The age span of respondents was not as evenly representative as it could have
been either as it was somewhat bimodal with over 30% 14 or 15 and over 30% 17 or
18. (See Table 1.) This distribution was likely due to the large number of Life Skills
classes taught to 8th and 9th grade students who are typically 14 or 15 years of age as
well as the number of classes taught to 12th grade students who are usually 17 or 18

years of age.

94

More representatively, nearly three quarters of youth completing surveys were
White. This mirrors the national average of 75.1%, is slightly below the Michigan
percentage of 80.2 and slightly above the Grand Rapids percentage of 67.3 (US
Census, 2000a). Because the majority of youth were from the greater Grand Rapids
region this percentage reflects fairly well the overall demographics for the area.
Considering that the Life Skills programs try to target high risk youth, it was expected
that the sample might include more minority youth, though there are many high risk
White youth in the study area as well. White youth also may have had better return
rates for their consent forms, though this is purely speculation.

As expected, the majority of youth were from the greater Grand Rapids area
where most of the Life Skills groups were held. A few groups were also held in
smaller cities and outlying areas in the more rural part of Kent County accounting for
the small number of youth living in a small town or in the country.

Aside from the overrepresentation of female youth, overall youth
characteristics were fairly representative of youth in the study area and even youth
nation wide. While still limited as a convenience sample with potential self-selection
bias, this may help some with generalizing to other similar populations.
Relationships between Feelings of Invulnegbility and Sexual Activifl

When looking at feelings of invulnerability to the consequences of sexual
activity, over a quarter of youth felt they would not get pregnant or get a partner
pregnant or get an STD with unprotected sex. When the questions were asked with
reverse coding later in the survey only around 5% felt strongly that they would not

get pregnant or get a partner pregnant or get an STD with unprotected sex. The

95

variation in the two sets of responses might indicate that the youth did not read the
questions very carefully or had difficulty with the reverse coding and may also
explain the relatively low Cronbach’s alpha of .75 for these four items. A
crosstabulation run with each of the two questions that were negatively worded
against the same positively worded questions indicated that 13 of the 67 respondents
may have misunderstood the question as they gave contradictory responses. Since the
correlation between just the first two questions using a Spearman’s rho was .84 (a <
.01) and between the second two questions was .76 (a <.01) it was felt best to analyze
each hypothesis not only with the full four question sexual vulnerability scale, but
each of the two sub-scales in turn as well even though the correlation between the two
sub-scales using a Spearman’s rho was only .36 (a < .01).

The low Cronbach’s alpha of .56 for the personal fable scale was also of
concern. While this personal fable scale was designed specifically to get at the
vulnerability piece of the personal fable, some of the questions, for example those
about feelings, did not seem to relate as well to the idea of risk taking behaviors. The
concern about the low reliability of the scale in this study led to a factor analysis.
Using the five items with the highest loading on the first factor from the factor
analysis resulted in a Cronbach’s alpha of .65 and was used in hypothesis testing,
along with the original personal fable scale.

About 43% of the youth had ever had sex. This is slightly lower than the
national average of 45.6% for high school students (Dillard, 2002). Since this sample
included some younger youth, not yet in high school, this would explain the slightly

higher percentage of youth who had not yet had sex.

96

Fourteen and 15 years of age were the most common responses when youth

who were sexually active reported when they first had sex. (See Table 20.) To see

how this compared to the actual age of the youth, a cross-tabulation was done. (See

Table 57.) While the lower rate of sexual activity in the 16 year olds is surprising, the

overall trend that older youth are more likely to be sexually experienced was expected

and again reflects national averages (Dillard, 2002).

Table 57- Youth Age and Chronological Age of' 1“ Sex Cross Tabulation

 

 

 

 

Count
Chronological age of 1" sex Total
never 12 13 14 15 16 17 18
had years years years years years years years
sex old old old old old old old
”um '2 °’ 6 0 0 0 0 0 0 0 6
age younger
13 7 0 0 0 0 O 0 0 7
14 10 0 1 0 0 0 O 0 11
15 6 1 0 4 0 0 O O 11
16 5 0 0 1 0 0 0 0 6
l7 1 0 0 2 3 2 l 0 9
18 3 0 0 1 6 2 2 1 15
19 or older 0 0 0 0 0 1 1 0 2
Total 38 1 1 8 9 5 4 1 67

 

 

 

 

 

 

 

 

 

 

 

From a risk perspective over half of the youth who had reported having had

sex were currently sexually active having had sex at least three times in the last three

months. Over half of the sexually active youth reported three or more life time

partners with the percentage of youth reporting four or more life time partners very

close to national averages (Dillard, 2002). Over half of those who reported having

had sex had never had sex while drunk and only about 21% of sexually active youth

had ever been diagnosed with an STD. Almost 60% of sexually active youth reported

consistent use of birth control and nearly 30% had been or had gotten their partner

97

 

 

pregnant. Considering the high risk nature of the study population, these results are
not surprising.

Youth with higher perceived invulnerability will be more likely to be sexually
active.

This hypothesis was not supported. The only significant relationship found
was a weak negative correlation with sexual activity and feelings of sexual
vulnerability using only the first two, negatively worded and reverse coded,
vulnerability questions. If youth read the questions correctly and did feel invulnerable
to the potential consequences of unprotected sexual activity it appears these youth
were less likely to have been sexually active. Conversely, youth who felt more
vulnerable were more likely to have had sex. It is unclear if the sample size was too
small to pick up any other significant relationships or if no other relationships exist. It
is possible that, while feelings of invulnerability might contribute to adolescent sexual
activity as postulated, once an adolescent has engaged in sexual activity feelings of
invulnerability may decrease as the youth might better realize the risks, especially if
the youth has experienced any of the potential negative consequences. A longitudinal
study that could follow youth from pre adolescence or early adolescence, before they
are sexually active, throughout adolescence would be better able to detect any
potential relationship to feelings of invulnerability and possible changes over time
especially in relation to sexual activity. If, as speculated, youth with higher levels of
invulnerability are more likely to engage in sexual activity, following them over time

would allow detection of this relationship.

98

Youth with higher perceived invulnerability will have greater fiequency of
sexual activity.

This hypothesis was not supported either with the only significant relationship
again coming from use of the first two sexual vulnerability scale questions with a
weak negative correlation. Similar to the findings of Boyer et al. (2000), this suggests
that youth may be able to judge their increased risk with increased sexual activity
even though it may not impact behavior, as youth with greater frequency of sex had
lower levels of invulnerability than youth with less frequency on this subscale

Youth with higher perceived invulnerability will have a greater number of
lifetime partners.

As before, using just the first two sexual vulnerability questions resulted in a
weak negative but significant correlation. All of the other vulnerability scales
produced even weaker correlations, none of which were significant. If youth
answering the negatively worded sexual vulnerability questions did indeed
understand the questions correctly, the most likely explanation again supports the
findings of Boyer et al. (2000) that youth may well be able to judge risk and that
youth who have had more partners may realize their increased vulnerability even if
this ability to judge risks and realize vulnerability does not lead them to change their
behavior.

Youth with higher perceived invulnerability will more frequently have sex
while drunk or high.

Two significant relationships were found between sexual activity and feelings

of invulnerability. The correlations between youth levels of perceived invulnerability

99

to the consequences of sexual activity and frequency of having sex while drunk or
high resulted in a small negative correlation that does not support the hypothesis that
youth with higher perceived invulnerability will more frequently have sex while
drunk or high. Testing again using just the first two sexual vulnerability questions
resulted in a slightly stronger, though still weak, negative correlation. Since
correlations cannot show causation but only indicate the presence of a relationship, it
is possible that youth who engage in sex while drunk or high feel more vulnerable,
even though it may not change their behavior. This is similar to findings by Cohen
and Bruce (1997) who found no correlation between perceived probability of STD
incident and actual behavior in a group of college students as well as results from
Williams et a1. (2003) showing that female college studentswere unlikely to take
precautions to reduce risk even when they acknowledged having sex with a high risk
partner. It may also be that youth who do not engage in sex while drunk or high may
feel less vulnerable as previously speculated for number of partners and frequency of
sexual activity since they do not engage in this high risk behavior.

Youth with higher perceived invulnerability will have greater use of birth
control.

The only significant result found when testing this hypothesis was a weak
negative correlation using the first two sexual vulnerability questions. This result,
assuming youth read the negatively worded sexual vulnerability questions correctly,
would indicate that youth with greater birth control use actually feel somewhat more
vulnerable. It seems counterintuitive that youth using birth control would feel more

vulnerable than youth not using birth control. Youth not using birth control may not

100

be sexually active though, and therefore, may feel less vulnerable, even with the
questions specifically saying “if I have unprotected sex”. Using a Spearman’s rho a
very high correlation of .94 (a < .01) was found between birth control use and sexual
activity as youth who are not sexually active generally have no need for birth control.
Almost 75% of sexually active youth stated they used birth control most of the time
or always with only about 14% saying they rarely or never did. Since correlations can
not determine cause and effect it may be that the nearly 75% of sexually active youth
who are good about using birth control took precautions because they felt more
vulnerable to the possible negative consequences of sexual activity. Being able to
look at longitudinal data regarding feelings of invulnerability, sexual activity, and
birth control use would be very helpful in better determining relationships and even
possible causation.

Youth who have ever had an STD will have lower levels of perceived
invulnerability.

Looking at youth who had a history of an STD there were no significant
differences between their mean invulnerability scores compared to the scores of those
who had never had an STD. While it was expected that youth, having had an STD,
would have lower levels of invulnerability, it may be that those youth who had a
history of an STD may not feel more vulnerable because they may think they are
protected if they are using birth control. Using a Spearman’s rho to look for a
correlation between history of STD and birth control use (with many forms also
offering some protection against STDs) there was a significant relationship with an r

of .35 and an alpha of <01. This provides some support for the explanation that these

101

youth may think that the use of protection may offer a small measure of safety against
STDs.

Youth who have ever been pregnant or gotten someone pregnant will have
lower levels of perceived invulnerability.

Youth who had been or had ever gotten anyone pregnant had mean
invulnerability scores that were not significantly different than those who had not.
Similar to youth with a history of an STD, however, the relationship between history
of pregnancy and birth control use was significant (a < .01) with an r of .36 also
providing support for why these youth may not have had decreased feelings of
invulnerability.

Youth with higher perceived invulnerability will be more likely to be sexually
active when other predictors of sexual activity are controlled.

Spearman Rho correlations for scaled predictors of sexual activity and sexual
activity in youth were all significant with r values ranging from .29 for parental
approval of teen drinking to .62 for youth’s belief that teens should not have sex.
These correlations reflect previous studies (Beal, Ausiello, & Perrin, 2001; Blum &
Rinehard, 1997; Kalil & Kunz, 1999; Kirby, 2002a; Luster & Small, 1994; Perkins et
al., 1998; and Wu et al., 2003) showing how risk factors relate to sexual activity in
youth.

Testing the hypothesis that youth with higher perceived invulnerability will be
more likely to be sexually active when other predictors of sexual activity are
controlled was difficult due to the small sample size. For this reason number of

variables in the logistic regression was limited to five. The top four variables, as

102

determined by correlation to sexual activity, were used with each of the vulnerability
scales in turn. The hypothesis was not supported. It may be that the small sample size
prohibited detection of a relationship, or it may be that feelings of invulnerability are
not correlated to sexual activity as speculated.

As some of the risk factors were nominal level data, chi-square tests were
performed to look for a difference in sexual activity for youth with and without these
risk factors. Nominal level risk factors associated with sexual activity included not
living with both parents, if parents were ever divorced or separated, if youth had
witnessed abuse, and if youth had been physically abused as indicated in previous
studies (Kirby, 2002a; Mulan et al., 2002; Perkins et al., 1998; and Talashek et al.,
2003). History of sexual abuse was not significantly correlated to sexual activity in
this population even through it was found to be correlated with sexual activity in
other studies (Kirby, 2002a; Perkins et al., 1998). The small percentage (9%) of youth
who had been sexually abused may have made it harder to detect a relationship in this
study.

When the four significant nominal level variables and sexual activity were
entered into a logistic regression with each of the vulnerability scales in turn, the
hypothesis that youth with higher perceived invulnerability would be more likely to
be sexually active when other predictors of sexual activity were controlled was still

not supported.

103

Perceived invulnerability will have less impact on the sexual behavior of
youth with few risk factors and more protective factors while having a greater impact
on those with multiple risk factors and fewer protective factors.

To test the hypothesis that perceived invulnerability will have less impact on
the sexual behavior of youth with few risk factors and more protective factors while
having a greater impact on those with multiple risk factors and fewer protective
factors each of the scaled variables was changed to a dichotomous variable in order to
calculate a total risk score without unequal weighting of any of the variables. When
the calculated risk scores were entered into a logistic regression with the sexual
invulnerability scale as well as the two separate sexual vulnerability subscales the
results were not significant.

When the total risk scores were entered into the model with the personal fable
scale as well as the five items with the highest loading on the first factor, the results
for both were significant. These support the hypothesis indicating that feelings of
invulnerability have a greater impact on the sexual activity of youth with more risk
factors and fewer protective factors. So, while the study was unable to find a
relationship between feelings of invulnerability and sexual activity when controlling
for the most highly correlated risk factors for sexual activity, it was able to detect a
relationship between feelings of invulnerability and sexual activity when controlling
these factors in youth with higher levels of risk. The small, but significant, impact of
feelings of invulnerability on higher risk youth lends supports to the work of
Hammen (2003) who felt that high risk youth could be impacted more by protective

factors than lower risk youth. In this case, though, high risk youth were impacted

104

more by feelings of invulnerability than low risk youth. Sameroff, Gutman, and Peck
(2003) suggest that protective factors have less impact on lower risk youth while
having greater impact on higher risk youth emphasizing the constellation of risk in
the lives of youth. This supports using a human ecology framework to help better
understand the many influences impacting adolescent sexual activity. The factors
associate with adolescent sexual activity found at the individual, microsystem, and
macrosystem levels in this study were also noted in previous works (Beal, Ausiello,
& Perrin, 2001; Blum & Rinehard, 1997; Kalil & Kunz, 1999; Kirby, 2002a; Luster
& Small, 1994; Mulan et al., 2002; Perkins et al., 1998; Talashek et al., 2003; and Wu
et al., 2003), while adding the little studied individual factor of perceived
invulnerability.

The finding that feelings of invulnerability have a greater impact on the sexual
activity of youth with more risk factors and fewer protective factors may be the first
to be documented on this topic. Some of the previously documented factors
associated with adolescent sexual activity may be impacted by intervention programs.
Interventions to help youth by tutoring, to improve school success and grades;
interventions to decrease teen drug use; interventions to involve youth in religious
activities; and interventions to educate parents to improve parental support and
monitoring are all examples that might help decrease risk factors and subsequently
decrease adolescent sexual activity. Other factors associated with adolescent sexual
activity though, are difficult or impossible to impact with basic intervention
programs. Youth programs can’t change a youth’s history of abuse or the fact that the

youth’s parents divorced and may or may not be able to change parental and youth

105

approval for teen sexual activity and drinking. At a macrosystem level, it may be
possible to impact cultural and societal norms that contribute to single parent homes,
abuse, low educational attainment, and views on the acceptability of adolescent
sexual activity and drinking. Changing these cultural and societal norms would
require changes in the media and how sexual activity is portrayed. This type of
systemic change needs strong grassroots community and/or high level governmental
support and possibly even regulation. A combined bottom up and top down approach
might be able to overcome some of the many challenges to this type of change.

The possibility for impacting adolescent sexual activity by decreasing the risk
factors found in this and previous studies needs further exploration. The added
finding that feelings of invulnerability in youth with high levels of measured risk also
impacts adolescent sexual activity is a new finding that need further substantiation.
While other risk factors associated with adolescent sexual activity have been well
documented, the literature contains little on feelings of invulnerability and how these
feelings relate to sexual activity. Knowing that, in this sample, a correlation was
found between feelings of invulnerability and high risk youth, this concept needs
further study. If feelings of invulnerability are found to correlate to adolescent sexual
activity in other samples, interventions to help counter these feelings, such as
interactive computer games or virtual reality simulations (Knoppers, 2003a) might be

helpful and should be investigated further.

106

Relationship between Factors Impfiting Sexual Activitiand Feelings of
Invulnerability

Younger youth will have higher levels of invulnerability.

The literature suggests that younger youth have higher levels of
invulnerability (Buis & Thompson, 1989; Greene, Kromar, & Walters, 2000; Lapsley,
1990; and Vartanian, 2000). There was no significant correlation between age and
any of the invulnerability scales except the second two sexual vulnerability questions
which showed a weak negative correlation and the five items with the highest loading
on the first factor which showed a weak positive correlation. So, for those two
positively worded sexual vulnerability questions younger youth felt slightly more
vulnerable than older youth supporting the thought that some younger youth had low
levels of invulnerability while some older youth had high levels. Younger youth
attending Life Skills classes may have realized greater levels of vulnerability to the
consequences of sexual activity through course content yet still demonstrated slightly
greater feelings of general invulnerability. Older youth, may have participated in high
risk sexual behaviors and not experienced any negative outcomes buoying their
feelings of invulnerability just on that scale.

Youth who use alcohol or drugs will have higher levels of invulnerability.

It was also hypothesized that youth who use alcohol or drugs would have
higher levels of invulnerability. Drug use included tobacco, alcohol, marijuana, and
other drugs. The majority of youth in this sample were non-smokers though over a
quarter of them smoked daily. While disappointing that so many youth smoke, the

percentage is not at all unexpected and fairly reflective of rates of smoking reported

107

by Burns and Johnston (1999). Similarly reported rates of youth drinking and drug
use were about what was expected, as Life Skills groups target higher risk youth.

Even with a fairly representative sample, when testing this hypothesis using
the personal fable scale no significant relationship was found. When testing with the
sexual vulnerability scale there was a weak negative correlation. The correlation was
a little stronger, though still weak for the first two sexual vulnerability questions but
was no longer found to be significant for the second two sexual vulnerability

questions alone. This negative correlation suggests that youth who use alcohol or

1:

drugs may actually realize these behaviors may put them at risk for the negative
consequences of sexual activity resulting in lower reported levels of invulnerability
on this scale. This unexpected finding may relate to program content in Life Skills
courses that emphasize the potential negative consequences of sexual activity and the
relationship to alcohol or drug use.

Interestingly enough, using the five items with the highest loading on the
personal fable scale resulted in a small significant positive relationship suggesting
that youth who use alcohol or drugs do have slightly higher levels of general
invulnerability when looking only at the more sensitive top five personal fable scale
questions. This supports the basic premise of the personal fable, that youth do not
think the risks, such as those involved in drinking or drug use, apply to them (Greene,
Rubin, & Kromar, 2002). The correlation between alcohol and drug use and the more
sensitive personal fable scale affirms the view of egocentric adolescents not believing
anything bad could happen to them in a general sense (Buis & Thompson, 1989)

while the negative correlation with the sexual vulnerability scale and the one subscale

108

shows that, at least in this population, youth can perceive some risks. This ability of
youth to perceive at least some risk, in this case related to the negative consequences
of sexual activity, indicates that interventions to decrease feelings of invulnerability
might be useful.
Youth with a history of sexual abuse will have lower levels of invulnerability.
Looking at the hypothesis that youth with a history of sexual abuse would
have lower levels of invulnerability there was no significant difference using any of

the vulnerability scales. Few of the youth surveyed reported a history of sexual abuse

' ub. .l‘bwfla

so detecting a relationship with the small number of subjects in the category may ‘4‘
have been difficult. A larger sample size with proportionately more youth with a
history of abuse might allow detection of a relationship, if one exists. Trying to get a
more inclusive sample might also help as it is possible, for whatever reason, that
youth with a history of sexual abuse may have been less likely to return signed
consent forms to participate in the survey.

Youth with a history of physical abuse will have lower levels of
invulnerability.

The results for history of physical abuse for this population are lower than a
study reported by Straus and Runyan (1997) and lower than expected rates if use of
corporal punishment in this age group is considered abusive but higher than rates
reported to Child Protective Service agencies (Straus & Runyan, 1997). Similar to
sexual abuse, youth who had a history of physical abuse were not significantly

different in levels of invulnerability compared to those with out a history of abuse.

109

 

Youth who have witnessed physical abuse will have lower levels of
invulnerability.

Over 55% of youth surveyed had witnessed physical abuse. The high
percentage of youth who had witnessed abuse likely reflects the communities where
these youth live. For the hypothesis that youth who had witnessed physical abuse
would have lower levels of invulnerability there was no significant difference using
the personal fable scale or the five items with the highest loading on the first factor
but there was using the sexual vulnerability scale and the first two sexual
vulnerability questions subscale. It is unclear why witnessing abuse would allow
youth to realize their vulnerability to the consequences of sexual activity but not
impact their general feelings of invulnerability. It is possible that feelings of
vulnerability resulting from witnessing abuse were more readily translated into the
more concrete feelings of vulnerability to the negative consequences of sexual
activity verses the more abstract general feelings of vulnerability measured by the
personal fable scale and the more sensitive five item scale. Significant results for only
one of the scales may also reflect having a small sample size that can detect some but
not all relationships. Additional research with larger sample sizes is needed to see if
relationships could be detected for general vulnerability as well.

Males will have higher levels of invulnerability.

Similarly males did not vary significantly from females for feelings of
invulnerability as measured by the personal fable scale and the five items with the
highest loading on the first factor but did with the sexual vulnerability scale and the

first two sexual vulnerability questions subscale. So, males felt less vulnerable to the

110

Consequences of sexual activity than females, but did not vary significantly from
females for general feelings of invulnerability. The higher levels of feelings of
invulnerability to the consequences of sexual activity are not surprising when
considering that adolescent males tend to be high in sensation seeking behaviors,
which include sexual activity (Green, Kromar, & Walters, 2000) while, generally
feeling somewhat immune to the potential negative consequences. Further research
into this relationship is needed as interventions to help counter these feelings of
invulnerability may be useful. The ability to pick up a difference in feelings of
invulnerability using the sexual vulnerability scale and one of the two subscales but
not the personal fable scales may also be related to having a small sample size,
especially considering that about half as many males completed surveys as females.
Additional studies with larger sample sizes are needed to see if a difference could be
detected between males and females.

Minority youth will have higher levels of invulnerability.

Minority youth did not vary significantly in their levels of vulnerability on any
of the vulnerability scales as compared to White youth. While the sample included
approximately 75% White youth, mirroring the national average (US Census, 2000a)
the higher risk nature of these white youth make it more difficult to generalize these
findings to other populations. The lack of any significant difference between White
and minority youth in this study may reflect that the entire study population was
considered somewhat high risk while differences between minority and White youth

may often reflect 'risks levels verses actual race. Also, since only about 25% of the

111

sample was minority and the overall sample size was small, a relationship may exist,
but could not be detected.

Youth with higher GPAs will have lower levels of invulnerability.

For the hypothesis that youth with higher GPAs would have lower levels of
invulnerability no significant relationship was found using the personal fable scale or
the sexual vulnerability scale. The five items with the highest loading personal fable
scale showed a weak relationship with an r of .20 and an alpha of <01 . The top five

factors may do a better job of getting at the vulnerability part of the personal fable

1' -'. 1W . a!

[.2

scale than the full 14 question scale, allowing the relationship to be detected even
with the smaller sample size. Donnelly, Eburne, and Eadie (1999) believe that
educating students about risks to help overcome perceived invulnerability can impact
behavior. A surprising result in the sample population was the high level of reported
grades, with over one quarter stating they got mostly As and over half getting mostly
Bs or greater. (See Table 3.) This may reflect some self-selection bias as stronger
students may be more responsible in following through to gain parental consent and
return consent forms. It is possible that this larger number of youth with higher GPAs
picked up better on general societal messages about risk and vulnerability. It is
unclear, however, why youth with better grades did not feel more vulnerable to the
consequences of sexual activity. Even so, educating students about risks may help
decrease feelings of invulnerability and may even impact behavior.

Youth with strong religious beliefs will have lower levels of invulnerability.

Youth with strong religious beliefs did not demonstrate lower levels of

invulnerability using any of the vulnerability scales. While it was thought that more

112

religious youth who typically engage in sexual activity at later ages than youth who
are less religious would feel more vulnerable, it was not the case in this sample. Only
about 10% of youth reported that their religion was important to them while nearly
35% said they were very involved in religious activities. (See Tables 18 and 19.) At
first this may seem somewhat contradictory and there is the possibility that youth did
not read the questions carefully as one of them was reverse coded. It may, however,
just reflect that while youth can be involved in many religious activities they may
view them as just that, activities, without feeling that the religious aspect of the
activity is important. This could reflect parental encouragement of certain activities
that the youth may not view with the same level of religious importance or may also
simply reflect that religious groups in the study area offer many activities as a means
of outreach to youth who may not be fully committed to that faith. While religiosity
does correlate to sexual activity, it was not one of the stronger correlations and so
may not have demonstrated a strong enough relationship to be detected with the
sample size, especially considering the small percentage of youth that reported that
religion was important to them.

Youth who believe that teens should not have sexual intercourse will have
lower levels of invulnerability.

Nearly 70% of youth felt teenagers should not be having sexual intercourse
even though only about 57% had never engage in sexual intercourse themselves. So,
some sexually experienced youth still thought that teenagers should not be having
sex. While it was hoped that youth would answer honestly, it is possible that they

responded how they thought they should, not necessarily how they actually felt. This

113

 

l ‘.—'T.. Ayn-.1 an
2

may also be reflective of their regrets over having had sex, their sense of being
special and that it was all right for them to have sex but not other teens, or may be
related to the Life Skills programs most attended that promote abstinence.

When testing the hypothesis that youth who believe teenagers should not be
having sexual intercourse would have higher levels of vulnerability no significant
difference was found between vulnerability levels of the youth who felt that teenagers
should not be having sex and those who thought it was all right. Since so many youth
felt that teens shouldn’t be having sex it may have made it more difficult to pick up a
difference between them and the youth who thought it was all right considering the
small sample size.

Youth who live with both parents will have lower levels of invulnerability.

Most youth lived with two parents while living with mom and step dad or
mom only were the second and third most common response. (See Table 5.) Fewer
two parent homes might have been expected considering the high risk nature of the
sample though results closely mirror Kent County demographics (US Census, 2000b)
and these youth may have been better about returning signed consent forms.

For the hypothesis that youth who live with both parents would have lower
levels of invulnerability the results were not significant using the personal fable scale
or the sexual vulnerability scale or subscale but were using the top five factor scale.
These results show that youth who live with both parents had lower levels of
invulnerability for the five items with the highest loading on the first factor though no
significant difference was detected using the other scales. A larger sample size might

have detected relationships using the other scales as well.

114

The lower invulnerability levels using the more sensitive top five factor scale
for youth living with both parents support much past research showing the benefits of
two parent households (Cooksey, Mott, & Neubauer, 2002; Lamers, Ireland, Resnick,
& Blum, 2000; and Miller, 2002). To see if living with both parents in this sample of
youth related to factors found to be protective against adolescent sexual activity found
in previous studies, Spearman’s rho correlations were run. Results showed a
significant ((1 < .01) r of .53 for average grades, an r of .46 with an alpha less than .01
for drug use, an r of .41 with an alpha less than .01 for witnessing abuse, an r of .33
with an alpha of .01 for parents education, an r of .36 with an alpha of less than .01
for parent monitoring, an r of .30 with an alpha of .01 for parental support, an r of
.29 with an alpha of .01 for religion, and an r of .26 with an alpha of .02 for peer
group. In this sample, percentage of youth living with two parents was similar to US
Census (2000b) data and while it is unclear if this population mirrored national
averages in other categories the positive correlations with other risk factors offer
some support for the protection afforded to youth living with both parents.

While youth programs can try to support parents so that youth can continue to
enjoy the benefits of living with both parents, there is little youth workers can do
about this risk factor for the many youth already living in single parent homes.
Interventions to encourage both parents to build strong relationships to each other
during pregnancy and throughout the child’s developmental years may also be helpful
but would require much more intensive programming and possibly years before
seeing any potential benefit, especially in the area of adolescent health. Working to

change the culture of our society to encourage stable two parent homes for the benefit

115

 

 

of our youth and future society might be possible through grassroots community work
as well as policy changes at all levels of government.

Youth whose parents have never been divorced or separated will have lower
levels of invulnerability.

When looking at youth whose parents had never been divorced or separated
there was no significant difference in vulnerability scores using any of the scales.
Again, a larger sample size may have been able to detect a difference especially
considering the very high correlation between youth living with two parents and those
with a history of divorce or separation. Using a Spearman’s rho resulted in an r of .88
with an alpha less than .01 showing these two factors to be highly correlated, as
expected.

Youth with higher levels of parental monitoring will have lower levels of
invulnerability.

Overall most youth reported fairly strong parental monitoring indicating
parental involvement in the lives of these youth. For a high risk population this may
seem unexpected, though some self selection bias may be reflected here as youth with
more supportive parental relationships may have been more motivated and had more
encouragement from parents to return their signed consent forms.

Youth with higher levels of parental monitoring were hypothesized to have
lower levels of invulnerability but no significant relationship was found for any of the
scales. While a larger sample size might have been able to detect a weaker
relationship it may also be true that levels of parental monitoring do not impact youth

feelings of invulnerability at all.

116

Youth with higher levels of parental support will have lower levels of
invulnerability.

Most youth surveyed also reported fairly strong parental support. Similar to
parental monitoring, having supportive parents was not related to lower levels of
invulnerability for any of the scales. This may indicate that parental influence does
not impact youth perceptions of invulnerability or it may be the relationship was too
weak to detect with this sample size.

Youth who have parents with higher levels of education will have lower levels
of invulnerability.

Only a small percentage of parents had not completed at least high school.
(See Tables 6 and 7.) The largest percentage were marked for high school education,
though many parents also had at least some college. This may again reflect some self
selection bias if youth whose parents were more highly education were more likely to
return signed consent forms.

There was no significant relationship detected for youth who had parents with
higher levels of educations using the personal fable scale or any of the sexual
vulnerability scales but, similar to youth with higher GPAs, there was using the five
items with the highest loading on the first factor ( r = .24, a = .03). This, as suspected
for youth with higher GPAs, may reflect the greater sensitivity of the five factor scale
but remains unclear as to why these youth would not feel more vulnerable to the
consequences of sexual activity. Reasons why youth whose parents had higher levels
of education might have lower levels of vulnerability may be similar to those for

youth with higher GPAs. To look for a correlation between youth average grades and

117

parental education a Spearman’s rho was run resulting in an r of .57 and an alpha of
.01 showing a moderately strong correlation. Youth whose parents had higher levels
of education were more likely to get better grades. This may reflect some degree of
inherited intelligence as some level of intelligence is generally needed to complete
higher levels of education or could be more related to the environmental impact
parents who value education have on their children’s grade attainment. Youth in both
categories, those whose parents had higher levels of education and those with better
grades, were more likely to have lower levels of invulnerability on the more sensitive
five factors personal fable scale. It is possible, considering the slightly stronger
correlation for youth whose parents have higher levels of education as compared to
youth with higher grades, that more highly educated parents are instructing their
children more about potential risks that might impact their feelings of invulnerability.
Since correlations cannot prove causality, further investigation of these relationships
is needed. Better understanding of why youth whose parents are more highly
educated have lower levels of invulnerability might lead to interventions to help
lower feelings of invulnerability in all youth.

Youth with parents who are more approving of adolescent sexual activity will
have higher levels of invulnerability.

Over 75% of youth agreed or strongly agreed that their parents think it is
wrong for teens their age to have sex, still leaving close to 25% that weren’t sure or
disagreed. In working with high risk populations it might be expected that there
would be less parental disapproval for adolescent sexual activity, though even parents

who may have engaged in sexual activity as adolescents might still not want their

118

children to do so. It is also possible, again, that more youth whose parents
disapproved of adolescent sexual activity returned their signed consent forms.

It was hypothesized that youth with parents more approving of adolescent
sexual activity would have higher levels of invulnerability, but this was not supported
using any of the scales. Just as no significant relationships were found between youth
who believe that teens should not have sexual intercourse and any of the vulnerability
scales. This suggests that feelings of invulnerability are not impacted by personal
beliefs of the youth or their parents or if a relationship exists it was too small to detect
with this sample size.

Youth with parents who are more approving of adolescent alcohol
consumption will have higher levels of invulnerability. _

Fewer youth strongly agreed that their parents think its wrong for teens their
age to drink alcohol than those who felt their parents thought it was wrong for
adolescents to engage in sexual activity, though the total for strongly agreed and
agreed was still close to 75%. It was hypothesized that youth with parents more
approving of adolescent drinking would have higher levels of invulnerability, but this
was not supported using any of the scales except the second two sexual vulnerability
questions subscale which showed a weak negative correlation. It is unclear why youth
whose parents are more approving of adolescent alcohol consumption would have
lower levels of invulnerability to the consequences of sexual activity but not general
feelings of invulnerability. There was a .41 correlation (a < .01) between adolescent
drinking and parents approval of adolescent drinking which may help explain the

similar results for each of these groups.

119

 

 

Youth who report having friends who get into trouble, are sexually active,
drink or use drugs, or smoke will have higher levels of invulnerability.

Similar percentages of youth (over 75%) agreed or strongly agreed that their
friends helped them stay out of trouble though only a little over half agreed or
strongly agreed that their fi'iends do not have sexual intercourse and do not drink or
do drugs with a slightly higher percent (58%) agreeing that their friends don’t smoke.
This reflects a fairly positive peer group for general behavior and a rather mixed
group for sexual activity, drinking, doing drugs, and smoking which fits expectations
for the study population.

The hypothesis that youth with a more negative peer group will have higher
levels of invulnerability was not supported, though, with the only significant
relationship a weak negative correlation using just the first two sexual vulnerability
questions. This negative correlation suggests that youth with fi'iends that are more
likely to get into trouble, be sexually active, drink, use drugs, or smoke actually have
slightly lower levels of invulnerability to the potential consequences of sexual
activity. While believed the impact of a negative peer group on these youth would
buoy feelings of invulnerability, the opposite may actually be true. Youth with peers
who are more likely to get in trouble, drink or do drugs, or have sex may actually
realize they are more at risk.

Youth reporting supportive communities will have lower levels of
invulnerability.

When looking at the hypothesis that youth reporting supportive communities

will have lower levels of invulnerability there was no significant correlation for any

120

of the scales. Approximately 70% of youth felt that there were neighbors who could
help them with a problem, that adults in the community would probably tell their
parents if they were doing something wrong, and that people in their community
know and care about each other. This suggests a fairly supportive community, yet,
considering that over half of youth had observed abuse, might indicate that youth live
in high risk areas but with caring neighbors.

Conclusion

Limitations.

A major limitation of this study was the sample size. While large enough to
detect moderate and some weaker correlations it may not have been large enough to
detect all of the relationships that may have been present and it was too small to fully
test those hypotheses requiring logistic regression. Additional research with larger
sample sizes is needed to further explore the concept of invulnerability as it relates to
sexual activity.

Another limitation was that, as a cross sectional study, it could not
demonstrate causality. Ideally, longitudinal studies that could track changes in
perceived invulnerability through adolescents as it relates to sexual activity should be
attempted.

A possible confounding factor in this study was the use of youth who were
participating in Life Skills programs. While the intention was to survey youth in the
early portion of the program, the difficulty in obtaining the signed consent forms
meant that some groups of youth were surveyed in the middle or near the end of the

programs. It is unclear if program content affected the way they answered any of the

121

survey questions. If using youth in similar programs in the future, surveys should be
done as close to the beginning of the program as possible.

Another weakness of this study was the poor reliability of the sexual
vulnerability scale and the personal fable scale. The scores from these scales were
used to test every hypothesis in this study. The poor reliability of the scales makes it
difficult to know if unsupported hypotheses are incorrect or simply a result of tools
that do not adequately measure the concept of feelings of invulnerability in
adolescents.

The sexual vulnerability scale consisted of two negatively worded reverse
coded questions placed early in the survey and two positively worded and coded
questions placed later in the survey. (See Appendix A.) Thetwo sets of questions
each asked if the youth felt they could get pregnant or get an STD with unprotected
sex, with only the wording varying between the sets. Perhaps additional questions
should have been added to firrther assess feelings of invulnerability to the
consequences of sexual activity. Considering the high percentage of sexually active
youth who used birth control, maybe asking youth about feelings of vulnerability in
relation to birth control use may have been useful. A sexual vulnerability scale with
more questions would have allowed a wider range of responses and might have been
more sensitive, especially if the reverse coding was a problem for youth. Before
further studies are undertaken, this scale will likely need to be revised and tested to
address these concerns. Hopefully a revised scale will result in higher reliability and a
better opportunity to truly test feelings of invulnerability to the consequences of

sexual activity in youth.

122

The reliability for the personal fable scale was even lower than that of the
sexual vulnerability scale. While the factor analysis allowed hypothesis testing using
the five items with the highest loading, further work on the tool is needed. It is
unclear why the tool, which had a reliability of .79 for Greene, Kromar, and Walters
(2000) had a reliability of only .56 in this sample. When reading through the 14
questions of this scale (see Appendix A) the top five questions (51, 53, 57, 68, and
60) from the factor analysis along with questions 56 and 58 seem to best get at the
concept of vulnerability as addressed in this study. Questions 48 and 50 do seem to
address risk somewhat while questions 47, 49, 54, and 55 seem to address general
feelings without getting at the vulnerability piece very well. While the concept of
personal fable and feelings of invulnerability in adolescents‘has been around for
decades, little research has been done and few tools exist to look at how these feelings
impact adolescent sexual activity. Modifying this tool or developing a similar tool
with a higher level of reliability is needed to allow future research to better measure
the concept of feelings of invulnerability in adolescents.

Lastly, the sampling technique used limits generalizability of the study results.
In addition to using a convenience sample of youth attending Life Skills programs
there was also a potential self-selection bias. While this sample mirrored national
averages in the percentage of minority youth and numbers living in two parent
households as well as general levels of sexual activity, it is unclear if the youth
returning signed consent forms varied significantly from youth who did not and how
well this group might represent youth nation wide. Future studies should strive to get

larger more representative samples.

123

 

Implications for future research, theory, and social policy/interventions.

Overall the findings support a human ecological perspective for understanding
factors that impact adolescent sexual activity at the individual, microsystem, and
macrosystem levels and support past research regarding correlates of sexual activity
in adolescents. Feelings of invulnerability were more likely to impact the sexual
behavior of youth with more risk factors. Further research is needed to better
understand the interaction between feelings of invulnerability and adolescent
behavior, especially as it relates to sexual activity. Considering some of the
unexpected results, longitudinal studies might be able to get a sense of causal
relationships to better guide interventions. Future studies should continue to assess
and attempt to control other factors impacting adolescent sexual activity in order to
view feelings of invulnerability in the context of the adolescent’s environment.

Research continues to support the ecological perspective to understanding
adolescent behavior. Risk factors and the way they interact remain crucial in planning
interventions to improve adolescent outcomes. Viewing adolescents in the context of
their environment and working to impact various aspects of their environment to
decrease risk remains important. The multiple factors impacting adolescent sexual
activity found in this and previous studies further support Bronfenbrenner’s (1989)
premise that adolescents are inextricably embedded in their environment. How the
adolescent personal fable and feelings of invulnerability interact within the adolescent
environment remains unclear. No simple relationships were found, with results based
on feelings of invulnerability to the potential consequences of sexual activity

sometimes contradicting those found when testing only general feelings of

124

invulnerability. This study brings to light a previously uninvestigated area of
adolescent development that needs further research

Considering the study findings that general feelings of invulnerability in
higher risk youth had a greater impact on sexual activity than for youth with lower
levels of invulnerabilty, those working with adolescents should continue to promote
protective factors while working to decrease risk factors, focusing on higher risk
youth if funding is limited. Understanding that multiple factors impact adolescent
behavior helps explain the ability for programs that focus on sexual abstinence and
programs that focus on community service to both have positive outcomes for
decreasing sexual activity in youth. Public policy needs to continue to support a wide
range of programs that take an ecological approach to positive youth development
and to decreasing adolescent sexual activity.

Many of the factors that impact adolescent sexual activity are at the individual
and microsystem levels. Yet, some of these factors cannot be easily impacted by
standard intervention programs. Factors at the macrosystem level and greater societal
norms that impact many microsystem factors need broad level system changes. For
example, evidence supports the benefit of two parent households for youth (Cooksey,
Mott, & Neubauer, 2002; Lamers, Ireland, Resnick, & Blum, 2000; and Miller, 2002)
yet out of wedlock births are increasingly common and carry little social stigma to
discourage them. The acceptance of adolescent sexual activity as inevitable if not
undesirable is another. While feelings of invulnerability may impact, in part,
adolescent sexual activity, the human ecological perspective, viewing the entire

system and the many factors impacting adolescent development and behavior, must

125

not be overlooked. This perspective should guide future research looking at the
environmental factors impacting adolescent sexual activity and potential intervention
at all levels that could help decrease adolescent sexual activity.

This study took an important first step in trying to determine if feelings of
invulnerability contribute to adolescent sexual activity. The cross-section survey
design of this study made it difficult to interpret some of the unexpected results.
Findings that indicated youth could perceive some risk and actually felt somewhat
vulnerable to the consequences of sexual activity need to be put in context. Did youth
have feelings of invulnerability initially and, only afier engaging in sexual activity,
realize their vulnerability? Or did some youth feel vulnerable, yet still engage in high
risk behavior? The answer to theses questions makes a difference in planning
interventions to decrease adolescent sexual activity. If helping adolescents realize
their vulnerability does not impact behavior are there additional interventions that
can?

While the small sample size may have limited the number of relationships
detected the results did show that perceived invulnerability had more impact on the
sexual behavior of youth with more risk factors than it did for youth with fewer risk
factors. This supports an ecological perspective of how interrelated these risk factors
may be. Future studies with larger samples may be able to detect additional
relationships that were not found in this study. Further exploration of the relationship
of perceived invulnerability and sexual activity in adolescents to better determine the
nature and strength of this relationship in other samples is desired. Keeping an

ecological perspective allows continued consideration of the multiple factors

126

impacting adolescent sexual activity. Helping higher risk youth identify feelings of
invulnerability to the potential consequences of sexual activity may be a useful
intervention for decreasing adolescent sexual activity especially if used in conjunction
with interventions aimed at other risk factors. Any attempted interventions to try and

decrease feelings of invulnerability should be carefully studied.

127

APPENDIX A
Knoppers’ Youth Assessment

This Questionnaire is being used to help identify youth beliefs, attitudes,
opinions, problems, and strengths.

The results will help in planning future programs and services for young
people in Kent County. All your answers are confidential, so please answer the
questions thoughtfully and honestly. If you cannot answer a question honestly, please
leave it blank.

Your participation is voluntary. You don’t have to answer any questions you
don’t want to.

0 Please follow the directions mark each answer carefully
o Raise your hand or ask for help if you need it.

Adaptations of the Dane County Youth Assessment 2000*

About Yourself
Mark the answer which best describes you.

1. Are you male or female? Male Female

2. To what racial or ethnic group do you belong?

__ Native American Indian

_ Black/African-American
Hispanic

_ Hmong

__ Asian (Not Hmong)

__ White (Not Hispanic)

__ Mixed race (for example, part Hispanic and part White)
If Mixed, which races?

Other, please list

 

 

3. How old are you?

__ 12 or younger __ 16
__13 __17
__14 __18

15 __19orolder

128

llI

 

4. What is the average grade you usually get in your courses at school?

__ Mostly As

__ About half As & half 85

__ Mostly Bs

__ About half Bs & half Cs

__ Mostly Cs

__ About half Cs & half Ds

__ Mostly Ds

__ Mostly below D

Your Living Situation
Answer these questions about the adults you live with.

1"“Wfi

5. Which best describes where you live the majority of the time?
__ In the Grand Rapids metropolitan area
_ In a smaller city (like Rockford, Lowell or Walker)
__ In a small town or village (like Kent City, Cedar Springs, Sparta)
_ In the country, not on a working farm

On a working farm

6. Whom do you live with most of the time?

__ Two parents (biological or adoptive)

__ Mother and stepfather
Father and stepmother

__ With mother only

__ With father only

__ Half the time with my mother, half the time with my father (shared custody)
With parent and another adult (non-relative)

__ Group home or foster home

_ With a relative (aunt, uncle, grandparents, etc.)

I live alone or with friends

129

7. How much education did your mother/stepmother complete? Give your best guess

if not sure.

Elementary or junior high school

High school

Some college or technical school

Graduated from a 2-year college or technical school

Graduated from a 4-year college

Some school beyond 4-year college

Professional or graduate degree (Ph.D, M.D., M.A., law degree, etc)
Don’t know

8. How much education did your father/stepfather complete? Give your best guess if

not sure.

Elementary or junior high school
High school

Some college or technical school

Graduated from a 2-year college or technical school

Graduated from a 4-year college

Some school beyond 4-year college

Professional or graduate degree (Ph.D, M.D., M.A., law degree, etc)

Don’t know

9. Are your biologic (or adoptive if you are adopted) parents divorced or separated?

Never divorced or separated

Never married

Divorced/separated within the last year

Divorced/separated more than one year ago

130

Substance Use Past Year
Please indicate how often you have used the following during the past year
1 to 3 l to 3 4 to 6

Not Once Times Times Times
At All or Twice a Month a Week a Week Daily
0 1 2 3 4 5

10. Tobacco (smoking cigarettes,
cigars, pipe, snuff, chewing) 0 1 2 3 4 5

11. Beer/wine/wine coolers,
hard liquor 0 1 2 3 4 5

12. Marijuana 0 l 2 3 4 5

13. Other drugs (inhalants,

hallucinogens, cocaine/crack,

stimulants, steroids, another persons

prescription drugs, etc.) 0 l 2 3 4 5

14. During the past month, have you had 5 or more alcoholic drinks at one time? (A

“drink” is a glass of wine or beer, a bottle or can of beer, a shot of liquor, or a mixed
drink.)
Never

Yes, once
__ Yes, twice

Yes, 3 to 5 times

Yes, 6 to 9 times

Yes, 10 or more times

131

Personal Issues
Please indicate how much you agree or disagree with the following statements.

Strongly Not Strongly
Agree Agree Sure Disagree Disagree
0 1 2 3 4
15. I will not get pregnant
(or get my partner pregnant)
if I have unprotected
sexual intercourse 0 1 2 3 4

16. I will not get a sexual

transmitted disease (such as

Chlamydia, gonorrhea, HIV/AIDS,

herpes, or htunan papilloma virus

that causes genital warts)

if I have unprotected

sexual intercourse 0 1 2 3 4

17. My religion/religious
beliefs are important to me 0 l 2 3 4

Which best describes you.

18. If you have ever been sexually active, how old were you the first time you
voluntarily had sexual intercourse?

I have never been sexually active __ 15 years old
_ 11 years old or younger _ 16 years old
__ 12 years old __ 17 years old
__ 13 years old _ 18 years old
__ 14 years old __ 19 years old or older

19. How ofien have you had sexual intercourse in the past 3 months?
__ I have not been sexually active in the last 3 months
1 time
2 times
3 times
4 times
5 times

6 or more times

132

20. How many sexual partners have you had in your lifetime?

__ I have not had any sexual partners

__ 1 sexual partner

__ 2 sexual partners

__ 3 sexual partners

__ 4 sexual partners

__ 5 sexual partners

6 or more sexual partners

21. If you have ever been sexually active, how often were you drunk or high while
having sexual intercourse?
I have not had sexual intercourse

Never
__ Rarely

Sometimes
__ About half the time
_ Most of the time
__ Always

22. Have you ever been diagnosed or treated for a sexually transmitted disease
(including HIV/AIDS, human papilloma virus/HPV/genital warts, Herpes, syphilis,
gonorrhea, or Chlamydia)?

No

Yes, within the past year

Yes, more than a year ago

23. If you have had sexual intercourse, how ofien do you and/or your partner use
some form of birth control?
I have not had sexual intercourse

Never
__ Rarely

Sometimes
__ About half the time
__ Most of the time

Always

133

I-

 

 

24. Have you ever been pregnant or made someone pregnant?
__ No
_ Yes, within the past year
__ Yes, more than a year ago

Please indicate how much you agree or disagree with the following statements.

Strongly Not Strongly
Agree Agree Sure Disagree Disagree
0 1 2 3 4

25. If I have unprotected sexual intercourse I
could get pregnant (or get my partner pregnant) 0 1 2 3 4

26. IfI have unprotected sexual intercourse I
could get a sexual transmitted disease (such as
chlamydia, gonorrhea, HIV/AIDS, herpes, or

human papilloma virus that causes genital warts) 0 l 2 3 4
27. 1 am not very involved with religious activities 0 i 1 2 3 4
Which best describes you.

28. Have you ever been sexually abused by an adults? (Sexual abuse is when someone
in your family or another person does sexual things to you or makes you do sexual
things to them.)

No

I am currently being sexually abused

I was sexually abused, but the abuse has stopped

29. Have you ever been physically abused by an adults (e. g., beat up, hit with an
object, kicked, or some other form of physical force)?
No

I am currently being physically abused
I was physically abused, but the abuse has stopped

30. Have you ever witnessed someone being beaten or physically abused?

Never Yes, in my home and at school
Yes, in my home Yes, in my school and town
Yes, in my school Yes, in all three of these places

Yes, in my town

134

Opinions About Your Community
Please read each statement carefully and decide whether you agree or disagree.

Strongly Strongly
Agree Agree Disagree Disagree
0 l 2 3
31. If I had a problem, there are neighbors
whom I could count on to help me. 0 1 2 3

32. If I were doing something wrong, adults in

my community would probably tell my parent(s). 0 1 2 3
33. People in my community know and

care about each other. 0 1 2 3
Your Family

Indicate how much of the following are true about the adults who live with you (e. g.,
your parents(s), stepparent, foster parent(s), or other guardian).
Very No adults
Never Rarely Sometimes Ofien Often at Home
0 1 2 3 4 5

34. I tell them whom I’m going
to be with before I go out. 0 1 2 3 4 5

35. I talk to them about the plans
I have with my friends. 0 1 2 3 4 5

36. When I go out they ask me
where I’m going 0 l 2 3 4 5

37. They usually know what
I am doing afler school. 0 1 2 3 4 5

Indicate how much of the following are true for you. (Answers these questions about
either your parent(s) or the adults who live with you).
Very No adults
Never Rarely Sometimes Often Often at Home
0 1 2 3 4 5
38. My parent(s) are there when I need them. 0 1 2 3 4 5

39. My parent(s) care about me. 0 1 2 3 4 5

135

Please indicate how much you agree or disagree with the following statements.

Strongly Not Strongly
Agree Agree Sure Disagree Disagree
0 1 2 3 4

40. My parent(s) think it is wrong for
teens my age to have sexual intercourse. 0 1 2 3 4

41. My parent(s) think it is wrong for
teens my age to drink alcohol. 0 1 2 3 4

Your Friends and Views
Please indicate how much you agree or disagree with the following statements.

Strongly Strongly
Agree Agree Disagree Disagree
0 1 2 3
42. My friends help me to stay out of trouble 0 1 2 3

43. Most of my fi'iends do not have sexual intercourse 0 1 2 3

44. Most of my friends do not drink or do drugs

O
_.
N
b)

45. Most of my friends do not
smoke cigarettes or chew tobacco 0 1 2 3

46. I believe teenagers should not

be having sexual intercourse. 0 1 2 3

Please continue to the last page and thank you for your help!

136

Vulnerability Questions from The New Personal Fable Scale **
Please indicate how much you agree or disagree with the following statements.

Strongly Not Strongly
Agree Agree Sure Disagree Disagree
0 1 2 3 4

47. Nothing seems to really bother me. 0 l 2 3 4
48. I don't believe in taking chances. 0 1 2 3 4
49. I am a fragile person. 0 1 2 3 4
50. I believe in knowing how something

will turn out before I try it. 0 1 2 3 4
51. I believe in taking risks. 0 1 2 3 4
52. There are times when I think that 0 1 2 3 4

I'm indestructible.
53. I can get away with things 0 l 2 3 4
that other people can't.

54. It is impossible for people to hurt 0 1 2 3 4

my feelings.
55. My feelings are easily hurt. O 1 2 3 4
56. Special problems, like using drugs

or becoming pregnant could

never happen to me. 0 1 2 3 4
57. I enjoy taking risks. 0 1 2 3 4
58. It is easy to take risks because I

never get hurt. 0 1 2 3 4
59. I don't take chances because

I usually get in trouble. 0 l 2 3 4
60. I am not afraid to do dangerous things. 0 1 2 3 4

" Adaptation of the Dane County Youth Assessment used with permission of Steven Small, University of
Wisconsin I""Vulnerability Subscale of the New Personal Fable Scale fiom Lapsley, D. K., FitzGerald, D. P., Rice,
K. G., & Jackson, S. (1989). Separation-individualtion and the “new look” at the imaginary audience and personal
fable: A test of an integrative model. Journal of Adolescent Research, 4, 483-505.

137

‘——'—-.
4. . a ..

APPENDIX B

Permission for The Youth Assessment Survey on
Perceptions of Invulnerability and Adolescent Sexual Activity

Dear Parent(s)/guardian(s),

This is an invitation for your child to participate in a study to look at some of
the factors that influence adolescent sexual activity. Your child’s participation in this
study will help us better understand these factors so that we can improve
interventions to postpone sexual activity in adolescents.

Your child’s participation in this study is voluntary and limited to one survey
that should take only about 10-15 minutes to complete. If your child feels
uncomfortable with any question on the survey the question may be left blank and
your child may choose to stop filling out the survey at any time. To better understand
the many factors that may impact adolescent sexual activity the survey includes
sensitive topics including questions about drug use and abuse. Your child’s name will
not be on the survey. All answers are confidential. Results will be reported for groups
of participants only. Your child’s privacy will be protected to the maximum extent
allowable by law.

If you have any questions about this study please contact the investigators,
Sherry Knoppers at 887-0875 or sknopper@grcc.edu or Lawrence Schiamberg at
(517) 432-8293 or schiambe@msu.edu. If you have questions or concerns regarding
the rights of your child as a study participant, or are dissatisfied at any time with any
aspect of this study, you may contact — anonymously, if you wish - Peter Vasilenko,
Ph.D., Chair of the University Committee on Research Involving Human Subjects
(UCRIHS) by phone: (517) 355-2180, fax: (517) 432-4503, e-mail: ucrihs@msu.edu,
or regular mail: 202 Olds Hall, East Lansing, MI 48824.

Your signature below indicates your voluntary agreement to allow your child
to participate. Your child’s signature indicates your child’s willingness to participate
as well.

 

I give permission for to
participate
parent/guardian’s name child’s name

 

In the Youth Assessment Survey

 

 

parent/guardian’s signature today’s date

I am willing to participate in this survey
child/youth’s name today’s date

 

138

APPENDIX C

Permission for The Youth Assessment Survey on
Perceptions of Invulnerability and Adolescent Sexual Activity

Dear Participant

This is an invitation for you to participate in a study to look at some of the
factors that influence adolescent sexual activity. Your participation in this study will
help us better understand these factors so that we can improve interventions to
postpone sexual activity in adolescents.

Your participation in this study is voluntary and limited to one survey that
should take only about 10-15 minutes to complete. If you feel uncomfortable with any
question on the survey the question may be left blank and you may choose to stop
filling out the survey at any time. To better understand the many factors that may
impact adolescent sexual activity the survey includes sensitive topics including
questions about drug use and abuse. Your name will not be on the survey. All
answers are confidential. Results will be reported for groups of participants only.
Your privacy will be protected to the maximum extent allowable by law.

If you have any questions about this study please contact the investigators,
Sherry Knoppers at 887-0875 or sknopper@grcc.edu or Lawrence Schiamberg at
(517) 432-8293 or schiambe@msu.edu. If you have questions or concerns regarding
your rights as a study participant, or are dissatisfied at any time with any aspect of
this study, you may contact — anonymously, if you wish — Peter Vasilenko, Ph.D.,
Chair of the University Committee on Research Involving Human Subjects
(UCRIHS) by phone: (517) 355-2180, fax: (517) 432-4503, e-mail: ucrihs@msu.edu,
or regular mail: 202 Olds Hall, East Lansing, MI 48824.

Your signature below indicates your voluntary agreement to to participate.

 

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

I am willing to participate in this survey
name today’s date

 

 

139

REFERENCES

Aarons, S. J., Jenkins, R. R., Raine, T. R., El-Khorazaty, M. N, Woodward, K. M.,
Williams, R. L., Clark, M. C., & Wingrove B. K. (2000). Postponing sexual
intercourse among urban junior high school students- a randomized controlled
evaluation. Journal of Adolescent Health, 27, 236-247.

Agha, S. (2002). An evaluation of the effectiveness of a peer sexual health
intervention among secondary-school students in Zambia. AIDS Education
and Prevention, 14, 269-281.

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

Alan Guttmacher Institute. (1994). Sex and America 's Teenagers. New York: Author.
Altman-Palm, N. & Tremblay, C. H. (1998). The effects of parental
involvement laws and the AIDS epidemic on the pregnancy and abortion rates
of minors. Social Science Quarterly, 79, 846-862.

Anderson, N. L. R., Nyamathi, A., McAvoy, J. A., Conde, F., & Casey, C. (2001).
Perceptions about risk for HIV/AIDS among adolescents in juvenile detention.
Western Journal of Nursing Research, 23, 336-359.

Arnold, E. M., Smith, T. E., Harrison, D. F ., & Springer, D. W. (1999). The effects
of an abstinence-based sex education program on middle school students’
knowledge and beliefs. Research on Social Work Practice, 9(1), 10-24

Aten, M. J ., Siegel, I). M., Enaharo, M., & Auinger, P. (2002). Keeping middle
school students abstinent: Outcomes of a primary prevention intervention.
Journal of Adolescent Health, 31, 70-8.

Bailey W., Young, M., Knickerbocker, C., & Doan, T. (2002). A cautionary tale
about conducting research on abstinence education: How do state abstinence
coordinators define “sexual activity?” American Journal of Health Education,
33, 290-296.

Bauman, K. E. & Udry, J. R. (1981). Subjective expected utility and adolescent
sexual behavior. Adolescence, 16, 527-535.

Baumrind, D. (1987). A developmental perspective on adolescent risk taking in
contemporary America. In W. Damon (Ed.), New directions for child
development: Adolescent health and social behavior, 3 7, 93-126. San
Francisco: Jossey-Bass.

140

Beal, A. C., Ausiello, J ., & Perrin, J. M. (2001). Social influences on health -risk
behaviors among minority middle school students. Journal of Adolescent
Health, 28, 474-480.

Bearman, P. S. & Bruckner, H. (2001). Promising the future: Virginity pledges and
First intercourse. American Journal of Sociology, 106, 859-912.

Benson, PL. (2003). Developmental assets and asset-building communities:
Conceptual and empirical foundations. In Lerner, R. M. and Benson, P. L.
(Eds.), Developmental assets and asset building communities (pp. 19-43).
New York: Kluwer.

Benson, P. L. (1997). All kids are our kids: What communities must do to raise

caring and responsible children and adolescents. San Francisco: J ossey-Bass
Publishers.

Blake, L. M., Simkin, L., Ledsky, R., Perkins, C. & Calabrese, J. M. (2001). Effects
of a parent-child communications intervention on young adolescents’ risk for
early onset of sexual intercourse. Family Planning Perspectives, 33, 52-61.

Blinn-Pike, L. (1999). Why abstinent adolescents report they have not had sex:
Understanding sexually resilient youth. Family Relations, 48, 295-301.

Blum, R. W., McNeely, C., & Nonnemaker, J. (2001). Vulnerability, risk, and
protection. Journal of Adolescent Health, 31, 28-39.

Blum, R. W. & Rinehart, P. M. (1997). Reducing the risk: Connections that make a
dzflerence in the lives of youth. Minneapolis: University of Minnesota
Division of General Pediatrics and Adolescent Health.

Boyer, C. B., Shafer, M., Wibbelsman, C. J ., Seeberg, D., Teitle, E., & Lovell, N.
(2000). Associations of sociodemographic, psychosocial, and behavioral
factors with sexual risk and sexually transmitted diseases in teen clinic
patients. Journal of Adolescent Health, 27, 102-111.

Brewster, K. L., Cooksey, E. C., Guilkey, D. K., & Rindfuss, R. R. (1998). The
changing impact of religion on the sexual and contraceptive behavior of

adolescent women in the United Sates. Journal of Marriage & the Family,
60, 493-504.

Bronfenbrenner, U. (1989). Ecological systems theory. Annals of child development,
6, 187-249.

Buis, J. M., & Thompson, D. N. (1989). Imaginary audience and personal fable: a
brief review. Adolescence, 24, 773-781.

141

Burns, D. M. & Johnston, L. D. (1999). Overview of recent changes in adolescent
smoking behavior. In D. M. Burns and R. H. Amacher (Eds), Changing
adolescent smoking prevalence: Where it is and why. US. Department of
Health and Human Services, Public Health Service, National Institutes of
Health, and National Cancer Institute Smoking and Tobacco Control
Monograph 14 (pp. 1-8).

Cauce, A. M., Stewart, A., Rodriguez, M. D., Cochran, B., & Ginzler, J. (2003).
Overcoming the odds? Adolescent development in the context of urban
poverty. In S. S. Luthar (Ed.), Resilience and vulnerability: Adaptation in the
context of childhood adversities (pp. 343-364). New York: Cambridge
University Press.

Chapin, J. (2001). It won’t happen to me: The role of optimistic bias in African
American teens’ risky sexual practices. The Howard Journal of
Communications, 12, 49-59.

Cohen, L. J. & Bruce, K. E. (1997). Sex and morality: Real risk and perceived
vulnerability. The Journal of Sex Research, 34, 279-291.

Collins, J. L., Robin, L., & Wooley, S. F. (2002). Programs-that-work; CDC’s guide
to effective programs that reduce health-risk behavior of youth. The Journal
of School Health, 72, 93-99.

Cooksey, E. C., Mott, F. L. & Neubauer, S. A. (2002). Friendships and early
relationships: Links to sexual initiation among American adolescents born to

young mothers. Perspectives on Sexual and Reproductive Health, 3 4, 118-
1 26.

Coyle, K., Basen-Engquist, K., Kirby, D., Parcel, G., Banspach, 8., Collins, J .,
Baumler, E., Carvajal, S. & Harrist, R. (2001). Safer choices: Reducing teen
pregnancy, HIV, and STDs. Public Health Reports, 116, S82.

Davis, C. M., Yarber, W. L., Bauserrnan, R., Schreer, G., & Davis, S. L. (1998).
Handbook of sexuality-related measures. Thousand Oaks, California: Sage
Publications.

Denner, J ., Kirby, D., Coyle, K., & Brindis, C. (2001). The protective role of social
capital and cultural norms in Latino communities: A study of adolescent
births. Hispanic Journal of Behavior Sciences, 23, 3-21.

DiClemente, R. J ., Wingood, G. M., Sionan, C., Crosby, R ., Harrington, K., Davies,
S., Hook, E. W., & Oh, M. K . (2002). Association of adolescents' history of
sexually transmitted disease (STD) and their current high-risk behavior and
STD status. Sexually Transmitted Diseases, 29, 503-509.

142

 

 

Diggs, J. R., Wallis, H., Mohn, J. K., & Jones, K. (2001). A perspective on the
medicalimplications of the virginity pledge among teens. Harrisburg, PA: The
Physicians Consortium.

Dillard, K. (2002). Adolescent sexual behavior. 1: Demographics. Washington, DC:
Advocates for Youth.

Dittus, P. J. & Jaccard, J. (2000). Adolescents’ perceptions of maternal disapproval of
sex: Relationship to sexual outcomes. Journal of Adolescent Health, 26, 268-
278.

Doniger, A. S., Adams, E., Utter, C. A., & Riley, J. S. (2001). Impact evaluation of
the “Not Me, Not Now” abstinence-oriented, adolescent pregnancy prevention
communications program. Journal of Health Communication, 5, 45-60.

Donnelly, J ., Eburne, N., & Eadie, C. (1999). Dispelling myths about HIV
transmission: An exploration of risk appraisal and self-responsibility. The
International Electronic Journal of Health Education, 2, 73-75.

Doswell, W. M. & Braxter, B. (2002). Risk-taking behaviors in early adolescent
minority women: Implications for research and practice. Journal of Obstetric,
Gynecologic, and Neonatal Nursing, 31, 454-461.

East, P. & Kieman, E. A. (2001). Risk among youths who have multiple sisters who
were adolescent parents. Family Planning Perspectives, 33, 75-80.

East, P., Kieman, E., & Chavez, G. (2003). An evaluation of Califomia’s adolescent
sibling pregnancy prevention program. Perspectives on Sexual &
Reproductive Health, 35, 62-70.

Elkind, D. (1979). The child and society: Essays in applied child development.
New York: Oxford University Press.

F ergusson, D. M. & Horwood, L. J. (2003). Resilience to childhood adversity:
Results of a 21-year study. In S. S. Luthar (Ed.), Resilience and vulnerability:
Adaptation in the context of childhood adversities (pp. 130-155). New York:
Cambridge University Press.

Frankenberger, K. D. (2000). Adolescent egocentrism: A comparison among
adolescents and adults. Journal of Adolescence, 23, 343-354.

Gorrnan-Smith, D. & Tolan, P. H. (2003). Positive adaptation among youth exposed
to community violence. In S. S. Luthar (Ed.), Resilience and vulnerability:
Adaptation in the context of childhood adversities (pp. 392-413). New York:
Cambridge University Press.

143

 

 

Gillmore, M. R., Archibald, M. E., Morrison, D. M., Wilson, A., Wells, E. A., Hoppe,
M. J ., Nahom, D., & Murowchick, E. (2002). Teen sexual behavior:

Applicability of the theory of reasoned action. Journal of Marriage and
Family, 64, 885-897.

Goldfarb, E. 8., Duncan, D. F ., Young, M., Eadie, C., & Castiglia, D., (1999).
Evaluation of an abstinence-based curriculum for early adolescents: First year

changes in sex attitudes, knowledge, and behavior. North American Journal of
Psychology, 1, 243-254.

Goossens, L., Beyers, W., Emmen, M., & van Aken, M. A. (2002). The imaginary
audience and personal fable: Factor analyses and concurrent validity of the
“New Look” measures. Journal of Research on Adolescence, 12, 193-215.

Greene, K., Kromar, M. & Walters, L. H. (2000). Targeting adolescent risk-taking
behaviors: The contributions of egocentrism and sensation-seeking. Journal of
Adolescence, 23, 439-461.

Greene, K., Rubin, D. R., & Kromar, M. (2002). Elaboration in processing adolescent
health messages: The impact of egocentrism and sensation seeking on
message processing. Journal of Communication, 52, 812-831.

Haglund, K. (2003). Sexually abstinent African American adolescent females’
descriptions of abstinence. Journal of Nursing Scholarship, 35, 231-236.

Halpern, C. T., Joyner, K., Udry, J. R., & Suchindran, C. (2000). Smart teens don’t
have sex (or kiss much either). Journal of Adolesdent Health, 26, 213-225.

Halpern, C. T., Udry, J. R., & Suchindran, C. (1997). Testosterone predicts initiation
of coitus in adolescent females. Psychosomatic Medicine, 5 9, 161-171.

Hammen, C. (2003). Risk and protective factors for children of depressed parents. In
S. S. Luthar (Ed.), Resilience and vulnerability: Adaptation in the context of
childhood adversities (pp. 50-75). New York: Cambridge University Press

Hawkins, M. 1., Davis, M., Eady, C., Rausch S. Donnelly, J. & Young, M. (2002).
Meanings of abstinence and sexual activity for rural youth. American Journal
of Health Education, 33, 140-147. '

Hetherington, E. M. & Elmore, A. M. (2003). Risk and resilience in children coping
with their parents’ divorce and remarriage. In S. S. Luthar (Ed.), Resilience
and vulnerability: Adaptation in the context of childhood adversities (pp. 182-
212). New York: Cambridge University Press.

Hulton, L. J. (2001). Adolescent sexual decision-making: An integrative review. The
'Online Journal of Knowledge Synthesis for Nursing, 8, document number 4.

144

Hutchinson, M. K. (1999). Individual, family, and relationship predictors of young

women’s sexual risk perceptions. Journal of Obstetric, Gynecologic, and
Neonatal Nursing, 28, 60-67.

Jack, M. S. (1989). Personal fable: A potential explanation for risk-taking behavior in
adolescents. Journal of Pediatric Nursing, 4, 334-338.

Jemmott, J. B., Jemmott, L. S. & Fong, G. T. (1998). Abstinence and safer sex HIV
risk-reduction interventions for African American adolescents. Journal of the
American Medical Association, 279, 1529-1536.

Kalil, A. & Kunz, J. (1999). First births among unmarried adolescent girls: Risk and
protective factors. Social Work Research, 23, 197-208.

Kaiser Family Foundation. (2003). Sex education in the United States: Policy and
politics. American Journal of Health Education, 33, 282-289.

Kalmuss, D., Davidson, A. C., Cohall, A., Laraque, A., & Cassell, D. C. (2003).
Preventing sexual risk behaviors and pregnancy among teenagers: Linking

research and programs. Perspectives on Sexual & Reproductive Health, 35,
87-93.

Kann, L., Kinchen, S., Williams, B., Ross, J ., Lowry, R., Grunbaum, J ., Kolbe, J. L.,
& State and Local YRBSS Coordinators. (2000). Youth risk behavior

surveillance- United States, 1999. In: CDC Surveillance Summaries, June 9,
2000. MMWR 2000:49 (No. SS-S): 1-32.

Karofsky, P. S., Zeng, L, & Kosorok, M. R. (2001). Relationship between adolescent-
parental communication and initiation of first intercourse by adolescents.
Journal of Adolescent Health, 28, 41-45.

Kerr, M. H., Beck, K., Shattuck, T. D., Kattar, C., & Uriburu, D. (2003). F amily
involvement, problem and prosocial behavior outcomes of Latino youth.
American Journal of Health Behavior, 27, 555-565.

Kirby, D. (2001a). Emerging answers: Research findings on programs to reduce teen

pregnancy (Summary). Washington, DC: National Campaign to Prevent Teen
Pregnancy.

Kirby, D. (2001b). Understanding what works and what doesn’t in reducing
adolescent risk-taking. Family Planning Perspectives, 33, 276-281.

Kirby, D. (2002a). Antecedents of adolescent initiation of sex, contraceptive use, and
pregnancy. American Journal of Health Behavior, 26, 473-485.

145

fl. " T".

 

Kirby, D. (2002b). Effective approaches to reducing adolescent unprotected sex,
pregnancy, and childbearing. Journal of Sex Research, 39, 51-57.

Kirby, D. (2002c). The impact of schools and school programs upon adolescent
sexual behavior. The Journal of Sex Research, 39, 27-33.

Kirby, D., Coyle, K., & Gould, J. B. (2001). Manifestations of poverty and birthrates
among young teenagers in California zip code areas. Family Planning

Perspectives, 33, 63-69.

Kirby, D, Korpi, M., Barth, R. P., & Cagampang, H. H. (1997). The
impact of the Postponing Sexual Involvement Curriculum among youths in
California. Family Planning Perspectives, 29, 100-108.

Klaczynski, P. A. & Fauth, J. M. (1996). Intellectual ability, rationality, and
intuitiveness as predictors of warranted and unwarranted optimism for future
life events. Journal of Youth & Adolescence, 25, 755-773.

Knoppers, S. M. (2003a). Adolescent personal fable. In J. R. Miller, R. M. Lerner, L.
B. Schiamberg, & P. M. Anderson’s (Eds), The encyclopedia of human
ecology: Vol. 1 (pp. 22-25). Denver, CO: ABC CLIO.

Knoppers, S. M. (2003b). Sexual abstinence. In J. R. Miller, R. M. Lerner, L B.
Schiamberg, & P. M. Anderson’s (Eds), The encyclopedia of human ecology:
Vol. 2 (pp. 638-641). Denver, CO: ABC CLIO.

Kowaleski, L. (2000). Staying out of trouble: Community resources and problem
behavior among high-risk adolescents. Journal of Marriage & the Family, 62,
449-464.

Koshar, J. H. (2001). Teen pregnancy 2001: Still no easy answers. Pediatric
Nursing, 27, 505-509.

Lammers, C. Ireland, M., Resnick, M., & Blum, R. (2000). Influences on adolescents’
decision to postpone onset of sexual intercourse: A survival analysis of

virginity among youths ages 13 to 18 years. Journal of Adolescent Health, 26,
42-48.

Landry, D. J ., Kaeser, L., & Richards, C. L. (1999). Abstinence promotion and the
provision of information about contraception in public school district
sexuality education policies. Family Planning Perspectives, 31, 280-286.

Lapsley, D. K. (1990). Continuity and discontinuity in adolescent social cognitive
development. In R. Montemayor, G. R. Adams, and T. P. Gullotta (Eds),
From childhood to adolescence: A transitional period? (pp. 183-204).
Newbury Park: Sage Publications.

146

Lapsley, D. K., FitzGerald, D. P., Rice, K. G., & Jackson, S. (1989). Separation-
individualtion and the “new look” at the imaginary audience and personal

fable: A test of an integrative model. Journal of Adolescent Research, 4, 483-
505.

Lerner, R.M. (2003), Developmental assets and asset-building communities: A view
of the issues. In Lerner, R. M. and Benson, P. L. (Eds), Developmental assets
and asset-building communities (pp. 3-18). New York: Kluwer.

Levine, J. A., Pollack, H., & Comfort, M. E. (2001). Academic and behavioral
outcomes among the children of young mothers. Journal of Marriage & the
Family, 63, 355-369.

Lieberman, L. D., Gray, H., Wier, M. Fiorentino, R., & Maloney, P. (2000). Long-
term outcomes of an abstinence-based, small-group pregnancy prevention
program in New York City schools. Family Planning Perspectives, 32, 237-
245.

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

Luthar, S S. & Zelazo, L. B. (2003). Research on resilience: An integrative review. In
S. S. Luthar (Ed), Resilience and vulnerability: Adaptation in the context of
childhood adversities (pp. 510-549). New York: Cambridge University Press.

Manlove, J ., Terry, E., Gitelson, L., Romano, A. R., & Russell, S. (2000). Explaining
demographic trends in teenage fertility, 1980-1995. Family Planning
Perspectives, 32, 166-175.

Martyn, K. K. & Martin, R. (2003). Adolescent sexual risk assessment. Journal of
Midwifery & Women ’s Health, 48, 213-219.

Masten, A. S. & Powell, J. L. (2003). A resilience framework for research, policy,
and practice. In S. S. Luthar (Ed.), Resilience and vulnerability: Adaptation in
the context of childhood adversities (pp. 1-25). New York: Cambridge
University Press.

Merrick, J. C. (1996). Kids, sex, and contraceptives: Dilemmas in a liberal society.
Politics and the Life Sciences, 15, 281-282.

Miller, B. C. (2002). Family influences on adolescent sexual and contraceptive
behavior. Journal of Sex Research, 3 9, 22-26.

147

Miller, B. C., Norton, M. C., Fan, X., & Christopherson, C. R. (1998). Pubertal
development, parental communication, and sexual values in relation to
adolescent sexual behavior. Journal of Early Adolescence, 18, 27-52.

Mohn, J. K., Tingle, L. R., & Finger, R. (2003). An analysis of the causes of the
decline in non-marital birth and pregnancy rates for teens from 1991 to 1995.
Adolescent & Family Health, 3, 39-47.

Morrison-Beedy, D., Carey, M. P., & Aronowitz, T. (2003). Psychosocial correlates
of HIV risk behavior in adolescent girls. Journal of Obstetric, Gynecologic,
and Neonatal Nursing, 32, 94-101.

Mullan, K. H., Duncan, G. J ., & Boisjoly, H. (2002). Evaluating the role of “nothing
to lose” attitudes on risky behavior in adolescence. Social Forces, 80, 1005-
1039.

Nahom, D., Wells, E. A., Gillmore, M. R. (2001). Differences by gender and sexual
experience in adolescent sexual behavior: Implications for education and HIV
prevention. The Journal of School Health, 71, 153-158.

Nangle, D. W., & Hansen, D. J. (1998). Adolescent heterosocial competence
revisited: Implications of an expanded conceptualization for the prevention of
high-risk sexual interactions. Education and Treatment of Children, 21, 431-
446.

The National Campaign to Prevent Teen Pregnancy. (2001). Halfway there: A
prescription for continued progress in preventing teen pregnancy.
Washington, DC: Author.

O’Donnell, L., Myint, U. A., O’Donnell, C. R. & Stueve, A. (2003). Long-term
influence of sexual norms and attitudes on timing of sexual initiation among
urban minority youth. The Journal of School Health, 73, 68-75.

O’Donnell, L., O’Donnell, C. R., & Stueve, A. (2001). Early sexual initiation and
subsequent sex-related risks among urban minority youth: The reach for
health study. Family Planning Perspectives, 33, 268-275.

O’Donnell, L., Stueve, A., O’Donnell, C., Duran, R., Doval, A. S., Wilson, R. F.,
Haber, D, Perry, E., & Pleck, J. H. (2002). Long-term reductions in sexual
initiation and sexual activity among urban middle schoolers in the Reach for
Health Service Learning Program. Journal of Adolescent Health, 31, 93-100.

Office of the Surgeon General (2001). The surgeon general’s call to action to promote
sexual health and responsible sexual behavior. Page 13. [Online] Retrieved

September 27, 2003. Available:
http://www.surgeongeneral.gov/library/sexualhealth/default.htm.

148

Out, J. W. & Lafreniere, K. D. (2001). Baby think it over: Using role-play to prevent
teen pregnancy. Adolescence, 36, 571-582.

Owens, E. B. & Shaw, D. S. (2003). Poverty and early childhood adjustment. In S. S.
Luthar (Ed.), Resilience and vulnerability: Adaptation in the context of
childhood adversities (pp. 267-292). New York: Cambridge University Press.

Paradise, J. E., Cote, J ., Minsky, S., Lourenco, A., & Howland, J. (2001). Personal
values and sexually experienced urban adolescent girls. Journal of Adolescent
Health, 28, 404-409.

Parsons, J. T., Perry, N. H., Bimbi, D. & Borkowski, T. (2000). Perceptions of the
benefits and costs associated with condom use and unprotected sex among late
adolescent college students. Journal of Adolescence, 23, 377-391.

Parsons, J. T., Siegel, A. W., & Cousins, J. H. (1997). Late adolescent risk-taking:
Effects of perceived benefits and perceived risks on behavioral intentions and
behavioral change. Journal of Adolescence, 20, 381-392.

Perkins, D. F., Luster, T., Villarruel, F. A. & Small, S. (1998). An
ecological, risk-factor examination of adolescents’ sexual activity in three
ethnic groups. Journal of Marriage & the Family, 60, 660-673.

Pete, J. M. & DeSantis, L. (1990). Sexual decision making in young black adolescent
females. Adolescence, 25, 145-154.

Philliber, S., Kaye, J. W., Herrling, S., & West, E. (2002). Preventing pregnancy and
improving health care access among teenagers: An evaluation of the

Children’s Aid Society-Carrera program. Perspectives on Sexual and
Reproductive Health, 34, 244-251.

The Physicians Consortium. (2003). New study shows that abstinence, not condom
use, is the major reason for the decline in the birth and pregnancy rates
among teens. Harrisburg, PA: Author.

The Physicians Consortium. (2002). Sexual messages in government-promoted
programs and today ’s youth culture. Harrisburg, PA: Author.

Porter, C. P., Oakley, D. J ., Guthrie, B. J ., & Killion, C. (1999). Early adolescents’
sexual behaviors. Issues in Comprehensive Pediatric Nursing, 22, 129-142.

Quandrel, M. J ., Fishchhoff, B. & Davis, W. (1993). Adolescent (in)vulnerability.
American Psychologist, 48, 102-116.

Rector, R. (2002). The effectiveness of abstinence education programs in reducing
sexual activity among youth. Washington, DC: Heritage Foundation.

149

Rector, R. E., Johnson, K. A., & Noyes, L. R. (2003). Sexually active teenagers are
More likely to be depressed and to attempt suicide. Center for Data Analysis
Report #03-04.

Roberts, D. F. (1993). Adolescents and the mass media: From “Leave It to Beaver”
to “Beverly Hills 90210”. In R. Takanishi (Ed.), Adolescence in the 1990’s
(pp. 171-186). New York: Teachers College Press.

Sameroff, A., Gutman, L. M., & Peck, S. C. (2003). Adaptation among youth facing
multiple risks: Prospective research findings. In S. S. Luthar (Ed.), Resilience
and vulnerability: Adaptation in the context of childhood adversities (pp. 354-
391). New York: Cambridge University Press.

Satcher, D. (2001). The Surgeon General’s call to action to promote sexual health and
responsible sexual behavior. American Journal of Health Education, 32, 356-
368.

Scales, P. & Leffert, N. (1999). Developmental assets: A synthesis of the
scientific research on adolescent development. Minneapolis: The Search
Institute.

Seidman, E. & Pedersen, S. (2003). Holistic contextual perspectives on risk,
protection, and competence among low-income urban adolescents. In S. S.
Luthar (Ed.), Resilience and vulnerability: Adaptation in the context of
childhood adversities (pp. 318-342). New York: Cambridge University Press.

Small, S. A., & Luster, T. (1994). Adolescent sexual activity: An ecological, risk-
factor approach. Journal of Marriage and the Family, 5 6, 181-192.

Smith, E. A., & Udry, J. R. (1985). Coital and non-coital sexual behaviors of white
and black adolescents. American Journal of Public Health, 75, 1200-1203.

Spear, C., Young, M., & Denny, G. (1997). Field testing of an abstinence-

based sexuality education program for upper elementary school students.
Journal of Health Education, 28, 335-344.

Standler, 1., Wolchik, S., Davis, C., Haine, R., & Ayers, T. (2003). Correlational and
experimental study of resilience in children of divorce and parentally bereaved
children. In S. S. Luthar (Ed.), Resilience and vulnerability: Adaptation in the
context of childhood adversities (pp. 213-240). New York: Cambridge

University Press.

150

Straus, M. A. & Runyan, D. I. (1997). Physical abuse. In S. B Friedman, M. M.
Fisher, S. K. Schonberg, and E. M. Alderman (Eds), Comprehensive
adolescent health care (2“ ed) (pp. 723-728). St Louis, MO: Mosby-Year
Book Inc.

Strong, B., DeVault, C. & Sayad, B. W. (1996). Core concepts in human sexuality.
Toronto: Mayfield Publishing Company.

Sutton, 8., McVey, D., & Glanz, A. (1999). A comparative test of the theory of
reasoned action and the theory of planned behavior in the prediction of
condom use intentions in a national sample of English young people. Health

Psychology, 18, 72-81.

Talashek, M. L., Norr, K. F ., & Dancy, B. L. (2003). Building teen power for sexual g
health. Journal of Transcultural Nursing, 14, 207-216. \

Taris, T. W. & Semin, G. R. (1999). Sexual efficacy of adolescents, perrnissiveness,
and the transition toward nonvirginity: From intention to behavior. Journal of
Applied Social Psychology, 29, 903-917.

Taylor, C.S., Lerner, R M., von Eye, A, Balsarro, A. B., Doivling, E. M., Anderson, P.A.,
Bobek, D. L., & Bjelobrk D. (2002).Stability of attributes of positive functioning
and of developmental assets among Afiican American adolescent male gang
and community based organization members. In Lerner, R. M., Taylor, C. S.
and von Eye, A. (Eds) New directions for youth development, pathways to
positive development among diverse youth, No. 95, (pp. 35-56). Indianapolis,
IN: Jossey-Bass Publishers.

Topolski, T. D., Patrick, D. L., Edwards, T. C., Huebner, C. E., Connell, F. A., &
Mount, K. K. (2001). Quality of life and health-risk behaviors among
adolescents. Journal of A dolescent Health, 29, 426-435.

University of Wisconsin Board of Regents (1999). Dane County youth assessment
2000.

US Census (2000a). Michigan quick facts fi'om the US Census Bureau. Retrieved
October 11, 2005, http://quickfacts.census.gove/qfd/states.html.

US Census (2000b). Profile of general demographic characteristics: 2000, Kent
County Michigan. Retrieved October 11, 2005,
http://factfinder.census.gov/servlet.

Vartanian, L. R. (2000). Revisiting the imaginary audience and personal fable

constructs of adolescent egocentrism: A conceptual review. Adolescence, 35,
639-661.

151

Ward, L. M., & Rivadeneyra, R. (1999). Contributions of entertainment television to
adolescents’ sexual attitudes and expectations: The role of viewing amount
versus viewer involvement. The Journal of Sex Research, 36, 237-249.

Werner, E. E. & Smith, R. (2001). Journeys fiom childhood to midlife: Risk,
resilience, and recovery. Ithicah, NY: Cornell University Press.

Williams, S. S., Norris, A. E., & Bedor, M. M. (2003). Sexual relationships, condom
use, and concerns about pregnancy, HIV/AIDS, and other sexually transmitted
diseases. Clinical Nurse Specialist, 17, 89-94.

Wu, Y., Stanton, B., Galbraith, J ., Kaljee, L., Cottrell, L., Li, X., Harris, C. V.,
D’Alessandri, D., & Burns, J. M. (2003). Sustaining and broadening
intervention impact: A longitudinal randomized trial of three adolescent risk
reduction approaches. Pediatrics, 111, 32-38.

Yarber, W. L., Milhausen, R., Crosby, R. A., & DiClemente, R. J. (2002). Selected
risk and protective factors associated with two or more lifetime sexual
intercourse partners and non-condom use during last coitus among US. rural
high school students. American Journal of Health Education, 33 , 206-216.

Yates, T. M., Egeland, L. B., & Soufe, A. (2003). Rethinking resilience: A
developmental process perspective. In S. S. Luthar (Ed.), Resilience and
vulnerability: Adaptation in the context of childhood adversities (pp. 243-
266). New York: Cambridge University Press.

152

      

MICHIGAN STATE U

WWII/llIll/ll!!!llIll/lllllllll/l/llll/I

1293 02845 0629